TTC McNicoll Bus Garage TPAP Air Quality Assessment Toronto, ON

Transcription

1 TTC McNicoll Bus Garage TPAP Air Quality Assessment Toronto, ON Novus Reference No Version No. 1 (Final) December 3, 2014 NOVUS PROJECT TEAM: Scientist: Project Manager: Jenny Vesely, B.Eng., EIT Scott Shayko, Hon. B. Comm, B.Sc. Air Quality Sound, Vibration & EMI/RFI Sustainable Water Wind & Climate Novus Environmental Inc. 150 Research Lane, Suite 105, Guelph, Ontario, Canada N1G 4T2 info@novusenv.com tel fax

2 This page intentionally left blank for 2-sided printing purposes

3 McNicoll Garage Air Quality Assessment December 3, 2014 Table of Contents 1.0 Introduction Project Description Contaminants of Concern Emissions from Buses and Motor Vehicles Emissions from Heating Equipment and Standby Diesel Generator Fugitive Emissions Applicable Guidelines... 3 Guideline D Ambient Air Quality Criteria Background (Ambient Conditions) Overview Selection of Relevant Ambient Monitoring Stations Selection of Worst-Case Monitoring Station Detailed Analysis of Selected Worst-Case Monitoring Stations Summary of Background Conditions Assessment Approach General Approach Location of Sensitive Receptors within the Study Area Facility Operations and Exhaust Parameters Bus Operations Comfort Heating Equipment and Standby Diesel Generator Paint Booth and Shop Areas Liquid Storage Tanks Employee Parking Lot Meteorological Data Emission Rates Vehicle Emission Rates (Buses and Employee Parking Lot) Heating Equipment and Standby Generator Emission Rates Paint Booth and Shop Areas Liquid Storage Tanks Modelling Methods Air Dispersion Modelling Using AERMOD Assessment of Negligibility for Contaminants in the Paint Booth and Shop Areas Results Novus Environmental i

4 McNicoll Garage Air Quality Assessment December 3, Combined Results for All Emission Sources, Not Including the Paint Booth and Shop Areas Results for the Paint Booth and Shop Areas Conclusions References List of Tables Table 1: Contaminants of Interest... 2 Table 2: Guideline D-6 Potential Influence Areas and Recommended Minimum Setback Distances for Industrial Land Uses... 4 Table 3: Applicable Contaminant Guidelines... 5 Table 4: Relevant MOECC and NAPS Monitoring Station Information... 8 Table 5: Comparison of Background Concentrations Table 6: Summary of Background NO Table 7: Summary of Background CO Table 8: Summary of Background PM Table 9: Summary of Background PM Table 10: Summary of Background Acetaldehyde Table 11: Summary of Background Acrolein Table 12: Summary of Background Benzene Table 13: Summary of Background 1,3-Butadiene Table 14: Summary of Background Formaldehyde Table 15: Predicted Hourly Bus Movements at the McNicoll Facility Table 16: Liquid Storage Tank Specifications Table 17: Schedule for Employees Arriving and Leaving the Parking Lot Table 18: MOVES Input Parameters Table 19: MOVES Output Emission Factors for Diesel Transit Buses for Table 20: Re-Suspended Particulate Matter Emission Factors Table 21: TANKS Model Emission Rates Table 22: Assessment of Negligibility for Liquid Storage Tanks Table 23: Worst-Case Predicted Concentrations as a Percentage of the Guideline Novus Environmental ii

5 McNicoll Garage Air Quality Assessment December 3, 2014 List of Figures Figure 1: Project Site... 1 Figure 2: Effect of Trans-boundary Air Pollution (MOECC, 2005)... 6 Figure 3: Typical Wind Direction during a Smog Episode... 7 Figure 4: Relevant MOECC and NAPS Monitoring Stations... 8 Figure 5: Summary of Background Conditions Figure 6: Receptor Locations Figure 7: Path for Buses Entering and Leaving the Facility Figure 8: Wind Frequency Diagram for Pearson International Airport List of Appendices Appendix A: Heating Equipment Specifications Appendix B: Paint Booth and Shop Area Contaminant Assessment Appendix C: Contour Plots for each contaminant Novus Environmental iii

6 McNicoll Garage Air Quality Assessment December 3, 2014 This page left blank for 2-sided printing purposes Novus Environmental iv

7 McNicoll Garage Air Quality Assessment December 3, Introduction Novus Environmental Inc. (Novus) was retained by URS Canada Inc. (URS) to conduct an air quality assessment for the proposed McNicoll Bus Garage located in the City of Toronto, Ontario. The focus of the assessment was to predict impacts at the nearby air-sensitive receptors from bus emissions as well as other stationary emission sources onsite. 1.1 Project Description The project includes the construction of a new bus storage and maintenance facility for the Toronto Transit Commission (TTC). The proposed facility is located on McNicoll Avenue, just east of Kennedy Road in the City of Toronto. The new facility will be used to house buses when they are not in use, and for general maintenance and repair on the buses. The majority of emissions will be due to idling buses prior to going into service. Emissions from natural gasfired heating equipment and standby generators, paint booth and shop areas and fugitive emissions from liquid storage tanks and employee parking lot were also considered. Figure 1 shows the project site, with the proposed building shown in blue and the employee parking lot shown in orange. Directly west of the proposed site is the Mon Sheong retirement home, and further west exists residential dwellings. North and west of the site exists industrial lands. Figure 1: Project Site Novus Environmental 1

8 McNicoll Garage Air Quality Assessment December 3, Contaminants of Concern 2.1 Emissions from Buses and Motor Vehicles The contaminants of interest from motor vehicles have largely been determined by scientists and engineers with United States and Canadian government agencies such as the U.S. Environmental Protection Agency (EPA), the Ontario Ministry of the Environment and Climate Change (MOECC), Environment Canada (EC), Health Canada (HC), and the Ontario Ministry of Transportation (MTO). These contaminants are primarily emitted due to fuel combustion, brake wear, tire wear, the breakdown of dust on the roadway. The contaminants of interest from motor vehicles are categorized as Criteria Air Contaminants (CACs) and Volatile Organic Compounds (VOCs). The contaminants emitted during fuel combustion include all of the CACs and VOCs, and the contaminants emitted from brake wear, tire wear, and breakdown of road dust include the particulates. A summary of these contaminants are provided in the following table. Table 1: Contaminants of Interest Criteria Air Contaminants (CACs) Nitrogen Dioxide (NO 2) Carbon Monoxide (CO) Fine Particulate Matter (PM 2.5) (<2.5 microns in diameter) Coarse Particulate matter (PM 10) (<10 microns in diameter) Volatile Organic Compounds (VOCs) Acetaldehyde Acrolein Benzene 1,3-Butadiene Formaldehyde These contaminants have been selected for this assessment due to their potential effect on human health or the environment and based on our experience represent the contaminants that are most likely to exceed government criteria for a facility of this nature. 2.2 Emissions from Heating Equipment and Standby Diesel Generator The main concern associated with boiler and generator exhaust due to the combustion of natural gas or diesel, is oxides of nitrogen (NOx), specifically nitrogen dioxide (NO2) in relation to human health. For this assessment, NO2 was assessed as the contaminant of concern from the natural gas-fired heating equipment. Novus Environmental 2

9 McNicoll Garage Air Quality Assessment December 3, Fugitive Emissions Fugitive emissions onsite were considered from re-suspended particulate matter from buses driving onsite, from the paint booth and shop space, from the storage tanks and vehicles in the parking lot. Contaminants of concern from the paint booth include several chemicals, including VOCs, contained in products used for painting and touching up the buses. It should be noted that the TTC will be using water-based paint on the buses, reducing the fugitive VOC emissions from the facility. The main concern for emission from the shop spaces is particulate matter from maintenance activities and products used. These areas will have fume extraction arms, downdraft exhaust welding tables, portable fume exhaust systems and a wall-mounted dust collector. It is assumed that this equipment will be used when needed, and all dust will be collected through the dust collector and not exhausted through the stacks. The touch-up paint shop will also have filter banks, using Fiberglass Paint Arrestor Pads for removal of paints, lacquer and enamels. The storage tanks will contain diesel fuel and various vehicle oils and fluids. The main concern for fugitive emissions from the storage tanks is evaporation of VOCs from the various products into the headspace of the tank, which vents to the atmosphere. The most volatile component present in any of the tanks is benzene, contained in the diesel fuel tanks. Given benzene s high vapour pressure and conservatively low standard under O.Reg 419/05, benzene was assessed as a worst-case contaminant emission scenario from the diesel tanks. Propylene glycol and isopropyl alcohol emissions were also assessed as criteria contaminants from the coolant and windshield fluids. 2.4 Applicable Guidelines There are several Provincial guidelines which have been considered in this assessment. Guideline D-6 The D-series of guidelines were developed by the Ontario Ministry of the Environment and Climate Change (MOECC) in 1995 as a means to assess recommended separation distances and other control measures for land use planning proposals in an effort to prevent or minimize adverse effects from the encroachment of incompatible land uses where a facility either exists or is proposed. The guideline specifically addresses issues of odour, dust, noise and litter. Guideline D-6 Compatibility Between Industrial Facilities and Sensitive Land Uses, addresses industrial land uses similar to the proposed bus facility. From the Guideline s synopsis, Guideline D-6 is intended to be applied in the land use planning process to prevent or minimize future land use problems due to the encroachment of sensitive land uses and Novus Environmental 3

10 McNicoll Garage Air Quality Assessment December 3, 2014 industrial land uses on one another. As the proposed project does not require a land use planning assessment (neither an Official Plan Amendment nor a Zoning By-law Amendment is required), Guideline D-6 does not strictly apply; regardless, it still can be used to consider what would generally be considered acceptable. Guideline D-6 defines an Area of Influence and a Recommended Minimum Setback distance for three classes of industrial operation: light, medium, and heavy industrial uses. These distances are determined by industry class and are shown in Table 2. Table 2: Guideline D-6 Potential Influence Areas and Recommended Minimum Setback Distances for Industrial Land Uses Industry Classification Area of Influence Recommended Setback Distance Class I Light Industrial 70 m 20 m Class II Medium Industrial 300 m 70 m Class III Heavy Industrial 1000 m 300 m Based on the size of the facility and the nature of the use, the proposed McNicoll bus facility is consistent with a Class 2 industry, with an Area of Influence of 300 m, and a Recommended Minimum Setback Distance of 70 m. Guideline D-6 recommends that detailed assessments be conducted where sensitive land uses are located within the Area of Influence of the industrial facility. There are several sensitive receptors within the Area of Influence. The closest sensitive use is the Mon Sheong residential development/ long term care facility. The detailed analyses presented in the subsequent sections of the report meet this requirement of Guideline D-6. Guideline D-6 also provides a Recommended Minimum Setback Distance of 70 m for Class 2 facilities. The distances between the Mon Sheong facility and the McNicoll facility are: Property line to property line 23 m Mon Sheong Building to closest on-site bus route 30 m While the Mon Sheong facility lies within the Recommended Minimum Setback Distance from the proposed McNicoll bus facility, Guideline D-6 is clear that the Minimum Setback Distance is a recommendation only. Section 4.10 of the Guideline allows for development to occur within the minimum setback for redevelopment, infilling and mixed use areas. This project would qualify as redevelopment. In such cases, Section 4.10 of the Guideline requires that a detailed assessment be conducted to show that the relevant air quality guidelines are met. The detailed analyses presented in the subsequent sections of the report show that this is the case. Thus, the minimum setback requirements of Guideline D-6 have been addressed. Novus Environmental 4

11 McNicoll Garage Air Quality Assessment December 3, 2014 Ambient Air Quality Criteria In order to assess the impact of the project, the predicted effects at sensitive receptors were predicted using detailed dispersion modelling, and compared to published guidelines. Relevant agencies and organizations in Ontario and their applicable contaminant guidelines are: MOECC Ambient Air Quality Criteria (AAQC) Canadian Council of Ministers of the Environment (CCME) Canada Wide Standards (CWSs) Within the guidelines, the threshold value for each contaminant and its applicable averaging period was used to assess the maximum predicted effect at sensitive receptors derived from computer simulations. The applicable averaging periods for the contaminants of interest are based on 1-, 8- and 24-hour acute (short-term) exposures. The threshold values and averaging periods used in this assessment for the main contaminants of concern are presented in Table 3. It should be noted that the CWS for PM2.5 is not based on the maximum threshold value. Instead, it is based on the annual 98 th percentile value, averaged over three consecutive years. Guidelines for the chemicals contained in the various products used onsite in the paint booth and shop areas are not presented in Table 3, but instead are presented in Appendix B. Table 3: Applicable Contaminant Guidelines Type Criteria Air Contaminants (CACs) Volatile Organic Compounds (VOCs) Pollutant Averaging Period Guideline (µg/m 3 ) Source NO 2 1 hr 400 AAQC 24 hr 200 AAQC CO 1 hr 36,200 AAQC 8 hr 15,700 AAQC PM hr 27 * AAQC (CWS) PM hr 50 Interim AAQC Acetaldehyde 24 hr 500 AAQC Acrolein 1 hr 4.5 Environmental 24 hr 0.4 Registry Benzene 24 hr 2.3 Environmental Registry 1,3-Butadiene 24 hr 10 Environmental Registry Formaldehyde 24 hr 65 AAQC * The CWS is based on the annual 98 th percentile concentration, averaged over three consecutive years. The standard becomes 27 in year Novus Environmental 5

12 McNicoll Garage Air Quality Assessment December 3, Background (Ambient Conditions) 3.1 Overview Background (ambient) conditions are contaminant concentrations that are exclusive of emissions from the proposed project infrastructure. These emissions are typically the result of trans-boundary (macro-scale), regional (meso-scale), and local (micro-scale) emission sources and result due to both primary and secondary formation. Primary contaminants are emitted directly by the source and secondary contaminants are formed by complex chemical reactions in the atmosphere. Secondary pollution is generally formed over great distances in the presence of sunlight and heat and most noticeably results in the formation of fine particulate matter (PM2.5) and ground-level ozone (O3), also considered smog. In Ontario, a significant amount of smog originates from emission sources in the United States which is the major contributor during smog events, usually occurring in the summer season (MOECC, 2005). During smog episodes, the U.S. contribution to PM2.5 can be as much as 90 percent near the southwest U.S. border and approximately 50 percent in the Greater Toronto Area (GTA). The effect of U.S. air pollution on Ontario on a high PM2.5 day and on an average PM2.5 spring/summer day is illustrated in the following figure. High PM2.5 Days Average PM2.5 of Spring/Summer Season US + Background US + Background Figure 2: Effect of Trans-boundary Air Pollution (MOECC, 2005) Air pollution is strongly influenced by weather systems (i.e., meteorology) that typically move out of central Canada into the mid-west of the U.S. then eastward to the Atlantic coast. This weather system generally produces winds with a southerly component that travel over major emission sources in the U.S. and result in the transport of pollution into Ontario. This phenomenon is demonstrated in the following figure and is based on a computer model run from the Weather Research and Forecasting (WRF) Model. Novus Environmental 6

13 McNicoll Garage Air Quality Assessment December 3, 2014 Figure 3: Typical Wind Direction during a Smog Episode As discussed above, understanding the composition of background air pollution and its influences is important in determining the potential impacts of a project, considering that the majority of the combined concentrations are typically due to existing elevated background levels. In this assessment, background conditions were characterized utilizing existing ambient monitoring data from MOECC and NAPS (National Air Pollution Surveillance) Network stations and added to the modelled predictions in order to conservatively estimate the combined concentration. 3.2 Selection of Relevant Ambient Monitoring Stations A review of MOECC and NAPS ambient monitoring stations in Ontario was undertaken to identify the monitoring stations that are in relevant proximity to the study area and that would be representative of background contaminant concentrations in the study area. Four MOECC (Toronto East, Toronto North, Toronto West and Toronto Downtown) and six NAPS (Toronto Downtown, Etobicoke South, Etobicoke North, Newmarket, Egbert and Windsor) stations were determined to be representative. The locations of the relevant ambient monitoring stations in relation to the study area are shown in Figure 4 and their station information can be found in Table 4. It should be understood that the selection of the Egbert and Windsor stations is due to the fact that formaldehyde and acetaldehyde have only been recently measured at the Egbert and Windsor stations and acrolein has only been recently measured at the Windsor station. It is likely that acrolein concentrations from Windsor result in conservative background concentrations in the study area due to the large amount of industrial activity in the Windsor area. Note that the Egbert and Windsor stations are not shown in the figure due to their distance from the study area. Novus Environmental 7

14 McNicoll Garage Air Quality Assessment December 3, 2014 Figure 4: Relevant MOECC and NAPS Monitoring Stations Table 4: Relevant MOECC and NAPS Monitoring Station Information City/Town Station ID Location Operator Contaminants Toronto East Kennedy Rd./Lawrence Ave MOECC NO2 PM2.5 Toronto North Hendon Ave./Yonge St. MOECC NO2 PM2.5 Toronto West Resources Rd. MOECC CO Toronto Downtown University Ave. W. MOECC CO Toronto Downtown College St NAPS Benzene 1,3-Butadiene Etobicoke South Kipling Ave NAPS Benzene 1,3-Butadiene Etobicoke North Elmcrest Road NAPS Benzene 1,3-Butadiene Newmarket Eagle St. NAPS Benzene 1,3-Butadiene Egbert Simcoe RR56/Murphy Rd. NAPS Formaldehyde Acetaldehyde Windsor College Ave./Prince Rd. NAPS Formaldehyde Acetaldehyde Acrolein Since the study area is surrounded by many monitoring stations, a comparison was performed for the available data on a contaminant basis, to determine the worst-case representative background concentration (see Section 3.3). Selecting the worst-case ambient data will result in a conservative combined assessment. Novus Environmental 8

15 McNicoll Garage Air Quality Assessment December 3, Selection of Worst-Case Monitoring Station The most recent five years of ambient monitoring data publically available from the selected stations were statistically summarized for the desired averaging periods, 1, 8 and 24-hr. For the CACs, data was available for the years and for the VOCs, data was available for at all stations except for Egbert, at which measurements were no longer recorded after For the contaminants with hourly monitoring data (NO2, CO and PM2.5), the station with the highest maximum value over the 5-year period for each contaminant and averaging period was selected to represent background concentrations in the study area. Using the maximum concentration is a very conservative assumption because it represents an absolute worst-case background scenario, which likely only occurred for one hour or one day over the five-year period. For this reason, it is often suggested that the 90 th percentile background concentration be selected to represent a reasonable worst-case scenario. However, in order to build conservatism into the results, the maximum background concentration was selected. Ambient VOC data is not monitored hourly, but is typically measured every six days. To combine this dataset with the hourly modelled concentrations, each measured 6-day value was applied to all hours between measurement dates, when there were 6 days between measurements. When there was greater than six days between measurements, the 90 th percentile measured value for the year in question was applied for those days in order to determine combined concentrations. This method is conservative in determining combined impacts as it assumed the 10 th percentile highest concentrations whenever data was not available. Table 5 shows a comparison of the relevant stations for each contaminant of interest, and the selection of the worst-case station. Novus Environmental 9

16 McNicoll Garage Air Quality Assessment December 3, 2014 Table 5: Comparison of Background Concentrations Note: PM 10 is not measured in Ontario; therefore, background concentrations were estimated by applying a PM 2.5/PM 10 ratio of 0.54 (Lall et al., 2004). Contaminant Worst-Case Station Contaminant Worst-Case Station NO 2 (1-hr) Toronto East 1,3-Butadiene Etobicoke South NO 2 (24-hr) Toronto North Benzene Etobicoke North CO (1-hr) Toronto West Formaldehyde Egbert CO (8-hr) Toronto West Acrolein Windsor PM 2.5 (24-hr) Toronto East Acetaldehyde Egbert PM 2.5 (3-yr) Toronto North PM 10 Toronto East Novus Environmental 10

17 McNicoll Garage Air Quality Assessment December 3, Detailed Analysis of Selected Worst-Case Monitoring Stations Year 2009 to 2013 hourly ambient monitoring data, the most recent 5 years publically available for CACs from nearby monitoring stations, was statistically summarized for the desired averaging period; 1-hour, 8-hour or 24-hour averaging periods were used. VOC data was available for the years , except at the Egbert station where measurements were stopped after VOCs are typically measured in Ontario on a 6-day basis. Where data was present every 6 days, the measured concentration was applied to all hours in that period. Where there was a greater than 6-day gap in the data, the maximum concentration for the given year was used to supplement the dataset. A detailed statistical analysis of the selected worst-case background monitoring station for each of the contaminants is presented below. The statistical analysis was summarized for average, 90 th percentile and maximum concentration. Each site was summarized on a yearly basis and for the five-year period. Where measurements exceeded the guideline, frequency analysis was performed. Novus Environmental 11

18 McNicoll Garage Air Quality Assessment December 3, 2014 Table 6: Summary of Background NO2 Statistical Analysis Five-Year Summary Statistic % of MOECC Guideline Maximum 39% 90 th Percentile 14% Average 7% Conclusion: A review of five years of ambient monitoring data from the Toronto East Station indicated that background concentrations are well below the MOECC guideline on a 1- hour basis. Statistic % of MOECC Guideline Maximum 45% 90 th Percentile 24% Average 14% Conclusion: A review of five years of ambient monitoring data from the Toronto North Station indicated that background concentrations are well below the MOECC guideline on a 24- hour basis. Novus Environmental 12

19 McNicoll Garage Air Quality Assessment December 3, 2014 Table 7: Summary of Background CO Statistical Analysis Five-Year Summary Statistic % of MOECC Guideline Maximum 6% 90 th Percentile 1% Average <1% Conclusion: A review of five years of ambient monitoring data from the Toronto West Station indicated that background concentrations are well below the MOECC guideline on a 1- hour basis. Statistic % of MOECC Guideline Maximum 12% 90 th Percentile 3% Average 2% Conclusion: A review of five years of ambient monitoring data from the Toronto West Station indicated that background concentrations are well below the MOECC guideline on an 8- hour basis. Novus Environmental 13

20 McNicoll Garage Air Quality Assessment December 3, 2014 Table 8: Summary of Background PM2.5 Statistical Analysis Five-Year Summary Statistic % of CWS Guideline Maximum 133% 98 th Percentile 76% 90 th Percentile 47% Average 25% Conclusion: A review of five years of ambient monitoring data from the Toronto East Station indicated that the maximum background concentration exceeded the CWS on a 24-hour basis. However, the guideline for PM 2.5 is based on the 98 th percentile value averaged over three consecutive years. Therefore, the highest 3-year average of 20.5 µg/m 3 was below the guideline. Frequency analysis was still conducted in order to show the number of days the background exceeded the guideline (see below). Number of Days Number of Days > Measured CWS Guideline 1, Conclusion: Frequency analysis determined that 24- hour concentrations exceeded the CWS on an infrequent basis. Measured concentrations exceeded the guideline 5 days over the 5-year period. This means that the background concentration exceeded the guideline less than 1% of the time over the 5-year period. Novus Environmental 14

21 McNicoll Garage Air Quality Assessment December 3, 2014 Table 9: Summary of Background PM10 Statistical Analysis Five-Year Summary Statistic % of MOECC Guideline Maximum 133% 90 th Percentile 47% Average 25% Note: PM 10 is not monitored in Ontario; therefore, background concentrations were estimated by applying a PM 2.5/PM 10 ratio of Lall et al. (2004) 1150 Conclusion: A review of five years of PM 10 data calculated from PM 2.5 ambient monitoring data from the Toronto East Station indicated that the estimated maximum background concentration exceeded the MOECC guideline on a 24-hour basis. Therefore, frequency analysis was conducted to determine the number of days the estimated background exceeded the MOECC guideline (see below). Number of Days Number of Days > Measured MOECC Guideline 1, Conclusion: Frequency analysis determined that 24- hour concentrations exceeded the MOECC guideline on an infrequent basis. Estimated concentrations exceeded the MOECC guideline 5 days over the 5 year period, with 4 days occurring in This means that the background concentration exceeded the MOECC guideline less than 1% of the time over the 5 year period. Novus Environmental 15

22 McNicoll Garage Air Quality Assessment December 3, 2014 Table 10: Summary of Background Acetaldehyde Statistical Analysis Five-Year Summary Statistic % of MOECC Guideline Maximum <1% 90 th Percentile <1% Average <1% Conclusion: A review of five years of ambient monitoring data from the Egbert Station indicated that the maximum background concentration was well below the MOECC guideline. Table 11: Summary of Background Acrolein Statistical Analysis Five-Year Summary Statistic % of MOECC Guideline Maximum 32% 90 th Percentile 19% Average 15% Conclusion: A review of five years of ambient monitoring data from the Windsor Station indicated that the maximum background concentration was well below the MOECC guideline. Novus Environmental 16

23 McNicoll Garage Air Quality Assessment December 3, 2014 Table 12: Summary of Background Benzene Statistical Analysis Five-Year Summary Statistic % of MOECC Guideline Maximum 99% 90 th Percentile 41% Average 27% Conclusion: A review of five years of ambient monitoring data from the Etobicoke North Station indicated that the maximum background concentration was slightly below the MOECC guideline. Table 13: Summary of Background 1,3-Butadiene Statistical Analysis Five-Year Summary Statistic % of MOECC Guideline Maximum 4% 90 th Percentile 1% Average <1% Conclusion: A review of five years of ambient monitoring data from the Etobicoke South Station indicated that the maximum background concentration was well below the MOECC guideline. Novus Environmental 17

24 McNicoll Garage Air Quality Assessment December 3, 2014 Table 14: Summary of Background Formaldehyde Statistical Analysis Five-Year Summary Statistic % of MOECC Guideline Maximum 13% 90 th Percentile 8% Average 5% Conclusion: A review of five years of ambient monitoring data from the Egbert Station indicated that the maximum background concentration was well below the MOECC guideline. 3.5 Summary of Background Conditions Based on a review of the most recent ambient monitoring dataset, all contaminants were below their respective MOECC criteria with the exception of PM10. PM10 concentrations were calculated based on their relationship to PM2.5. It should be noted that even though the maximum concentration of PM2.5 exceeded the CWS, the guideline for PM2.5 is based on an average annual 98 th percentile concentration, averaged over three consecutive years. Therefore, it was determined that the maximum rolling 98 th percentile average was 20.5 µg/m 3, which is less than the guideline. A summary of the background concentrations as a percentage of their respective MOECC guidelines or CWS is presented in the following figure. Also presented is the number of days that the monitoring data was above the MOECC guideline or CWS. Novus Environmental 18

25 McNicoll Garage Air Quality Assessment December 3, 2014 Figure 5: Summary of Background Conditions 4.0 Assessment Approach 4.1 General Approach In order to estimate the worst-case impacts resulting from emissions from the McNicoll Bus Garage the following were conducted: Emission rates were estimated based on U.S. EPA and MOECC published values; Air dispersion modelling was conducted; and Maximum model results were combined with maximum background concentrations to provide conservative predictions of worst-case impacts. Novus Environmental 19

26 McNicoll Garage Air Quality Assessment December 3, Location of Sensitive Receptors within the Study Area Land uses which are defined as sensitive receptors for evaluating potential air quality effects are: Health care facilities; Senior citizens residences or long-term care facilities; Child care facilities; Educational facilities; Places of worship; and Residential dwellings. The nearest existing sensitive receptor is the Mon Sheong residence/long-term care facility, located just west of the facility, approximately 20 m from the facility s property boundary line. This is the closest sensitive receptor. Receptors were placed at ground level and in 3 m height increments to measure impacts at operable windows at all levels on the retirement home. Three churches were identified near the facility, located 80 to 400 m from the property boundary line. The vacant land to the east of the facility, 2150 McNicoll Avenue, is zoned under Scarborough General Zoning Bylaw as Heavy, General and Special Industrial (M, MG and MS). Regardless of the industrial nature of this zoning, permissions allow for educational facilities, daycares and places of worship. There are currently no publically-made plans for development, and no building permits on the property. In the absence of specific direction on how to assess vacant lots in the air quality guidelines; and to be consistent with the approach taken in the noise assessment, a vacant lot surrogate receptor has been placed on the property consistent with the requirements of MOE Publication NPC-300 noise guideline. The point of reception has been located at the centre of a 1 Ha. building envelope, located on the lot consistent with the setback restrictions of the zoning by-law with the typical building pattern in the area. The receptor is located approximately 100 m from the proposed McNicoll facility property boundary line, and was considered when predicting worst-case impacts. Figure 6 shows the receptor locations in yellow, the proposed facility in blue and the property boundary line in red. It should be noted that since sensitive-receptors (the senior citizen s residence and potential for day care or educational facility) were identified nearby the proposed site, the relaxed standard for assessing emergency generators was not applied. Total NOx emissions from the site, including the emergency generator, were assessed against the ambient air quality guideline of 400 µ/m 3 for a 1-hour and averaging period for NOx at the identified sensitive receptors. Emissions from the emergency generator were not considered when assessing impacts on a 24-hour averaging period against the 200 µg/m 3 guideline, in accordance with MOE guidelines for assessing emergency generators. Novus Environmental 20

27 McNicoll Garage Air Quality Assessment December 3, 2014 Figure 6: Receptor Locations 4.3 Facility Operations and Exhaust Parameters Bus Operations The main emissions from buses will occur due to idling buses inside the facility prior to going into service. Hourly bus counts entering and leaving the existing Mount Dennis facility (which is similar to the proposed facility) were provided by URS, as well as a maximum vehicle count of 220 buses at the proposed McNicoll Bus Garage. The hourly vehicle distribution at the Mount Dennis facility was applied to the maximum number of buses proposed at the McNicoll Bus Garage to determine the number of buses that would be leaving/entering the facility during any given hour for this assessment. The same hourly vehicle distribution was assumed for every day of the week. As stated by the design team, a maximum idling time of 10 minutes for any vehicle within the facility was assumed. To be conservative, it was assumed that each vehicle moving in that hour (in or out of the facility) would idle for 10 minutes. Vehicle movements used in the assessment are provided in Table 15. Novus Environmental 21

28 McNicoll Garage Air Quality Assessment December 3, 2014 Table 15: Predicted Hourly Bus Movements at the McNicoll Facility Hour Buses Leaving Facility Buses Entering Facility Total Bus Movements Idling emissions from inside the storage bay will be emitted through the 20 air handling unit exhaust fans, with an exhaust diameter of 1.5 m and an average flow rate of 7 m 3 /s, as specified in the mechanical schedule for the facility provided by Stantec. It was assumed that the emissions from buses in the storage bay would be evenly mixed and emitted through the air handling units serving this area. Vehicles may also idle in the maintenance bay while being worked on, and are connected up to a bus fume exhaust hose system which exhaust on the rooftop. In the assessment it was assumed that one bus would be idling at all times through each of the six bus fume exhaust hose systems. This is conservative as it is not likely that buses will be idling while being worked on at all times. These fans exhaust 4 m above the rooftop, were modelled with a diameter of 0.4 m, which is conservative as it yields a low exit velocity and reduces dispersion of the exhaust. An average flow rate of 1.5 m 3 /s was provided in the mechanical schedule. Novus Environmental 22

29 McNicoll Garage Air Quality Assessment December 3, 2014 Buses were modelled leaving and entering the facility from the north entrance on Redlea Avenue. Buses entering drive south along the west side of the building, and around to enter on the east side of the building. Buses leaving exit at the east side of the facility, drive north around the facility and exit onto Redlea Avenue. The path for buses entering and leaving the facility is shown in Figure 7. Buses were modelled entering and leaving as per the schedule shown in Table 15. Figure 7: Path for Buses Entering and Leaving the Facility Comfort Heating Equipment and Standby Diesel Generator The total heat input for natural-gas-fired air handling units and unit heaters for comfort heating is 55,465,000 kj/hr. The heating input for each individual air handling unit and unit heater assessed is provided in Appendix A. Some air handling units are equipped with heat recovery units, for which a reduction in total heating input was applied. The reduction in heating capacity due to heat recovery units is also detailed in Appendix A. Air handling units were modelled with a flue diameter of 0.25 m and the unit heaters with a flue diameter of 0.1 m. Flow rates for each unit were calculated from the stoichiometric balance for the combustion of natural gas, and are also listed in Appendix A. Novus Environmental 23

30 McNicoll Garage Air Quality Assessment December 3, 2014 The facility will also have a diesel-fired 800 kw standby generator, located at grade at the northeast corner of the property. The generator was modelled conservatively with an exit diameter of 0.4 m and flow rate of 2.5 m 3 /s. Paint Booth and Shop Areas Several products will be used in the paint booth and shop areas as part of maintenance operations and work on the buses. A full list of the products with chemical composition is provided in Appendix B. Most of the products will be applied with High Volume Low Pressure (HVLP) spray gun, however, some products will be applied by hand. The spray gun used to apply products will have a maximum flow rate of 0.42 L/min. These products will be used in either the paint booth, millwrights shop, paint prep area, CIS control area or body shop, each of which has a dedicated exhaust stack. The paint booth stack is 8 m above rooftop, and has a flow rate of 19.8 m 3 /s, as per the provided mechanical schedule. A large stack diameter of 1 m was modelled with a low exit velocity, to provide conservative predictions. The other stacks from the shop areas were modelled with an average diameter of 0.2 m and flow rate of 0.2 m 3 /s, which is listed in the mechanical schedule. Modelling of both the paint booth and shop area stacks showed that lower dispersion levels (high offsite concentrations) occurred for emissions from the smaller shop area stacks. It was therefore conservatively assumed that all contaminants could be emitted from these stacks, to predict worst-case impacts. Typical usage of the paint gun for any product will be no more than for minutes at a time, 4-5 times per shift during each of the 3 shifts, which equates to 7.5 hours per day. This is a conservative maximum usage, in reality many products will be used less than 7.5 hours in a day, and some products only a few times a week. For contaminants with a 24-hour averaging period, the mass flow rate was determined for use of the spray gun 7.5 hours in a day, for a normalized flow rate of 0.13 L/min throughout the day. For contaminants with a 1-hour averaging period, it was assumed the gun could be used for the entire hour, at the full flow rate of 0.42 L/min. These flow rates were used to determine mass contaminant emission rates in Section Liquid Storage Tanks There are 12 liquid storage tanks onsite, containing diesel fuel and various vehicle oils and fluids. The tanks are located on the east side of the facility. Details for each of the tanks is provided in Table 16. Emissions from the vapour head space in the tanks were modelled seeping slowly out of the provided 0.05 m vent with an exit velocity of m/s. Novus Environmental 24

31 McNicoll Garage Air Quality Assessment December 3, 2014 Table 16: Liquid Storage Tank Specifications Tank ID Product Height (m) Diameter (m) Filling Frequency T-1 Diesel Once/day T-2 Diesel Once/day T-3 Diesel Once/day T-4 Engine Oil Once /3 months T-5 Engine Oil Once /3 months T-6 Transmission Fluid Once /6 months T-7 Transmission Fluid Once /6 months T-8 Engine Coolant Once /week T-9 Windshield Fluid Once/week T-10 Gear Oil Once/3 months T-11 Waste Oil As Required T-12 Waste Glycol As Required Employee Parking Lot Emissions from vehicles in the employee parking lot were also considered in the assessment. The shift schedule for bus operators and employees working on maintenance operations was provided by the TTC, and is shown in Table 17. It was assumed that every employee would drive to and from work. Vehicles were modelled driving at a slow speed (20 km/hr) in the parking lot, as per the schedule in Table 17. Novus Environmental 25

32 McNicoll Garage Air Quality Assessment December 3, 2014 Table 17: Schedule for Employees Arriving and Leaving the Parking Lot Hour Vehicles Arriving Vehicles Leaving Total 0: :00 0 2: : : : : : :00 0 9: : : : : : : : : : : : : : : Meteorological Data hourly meteorological data was obtained from Toronto Pearson International Airport. The full year of 2011 meteorological data is not available from the U.S. National Center for Atmospheric Research (NCAR), therefore was the most up to date and complete meteorological data available. Upper air data was obtained from the Buffalo Niagara International Airport, as per MOECC guidance. The combined data was processed to reflect conditions at the study area using Lakes Environmental s AERMET software program which prepares meteorological data for use with the AERMOD model. A wind frequency diagram (wind rose) is shown in Figure 8. As can be seen in this figure, predominant winds are from the southwesterly through northerly directions. Novus Environmental 26

33 McNicoll Garage Air Quality Assessment December 3, 2014 Figure 8: Wind Frequency Diagram for Pearson International Airport 4.5 Emission Rates Vehicle Emission Rates (Buses and Employee Parking Lot) MOVES is a computer program that provides estimates of current and future emission rates from motor vehicles based on a variety of factors such as local meteorology and vehicle fleet composition. MOVES 2014, updated in October 2014, is the U.S. EPA s latest tool for estimating vehicle emissions due to the combustion of fuel, and brake and tire wear. The model is based on an analysis of millions of emission test results and considerable advances in the Agency's understanding of vehicle emissions and accounts for changes in emissions due to proposed standards and regulations. For this project, MOVES was used to estimate emissions from diesel buses and passenger vehicles in the employee parking lot. Emission rates were estimated for a base year of 2011, as the fleet at the bus garage may be composed of buses as old as This is conservative as MOVES predicts vehicle emission rates to decrease in the future due to improved technologies and stricter regulations. Table 18 specifies the major inputs into MOVES. Novus Environmental 27

34 McNicoll Garage Air Quality Assessment December 3, 2014 Table 18: MOVES Input Parameters Parameter Input Scale Custom County Domain Meteorology Temperature and Relative Humidity were obtained from meteorological data from the Toronto Airport. Years 2011 Geographical Bounds Custom County Domain Fuels Diesel Fuel, Natural Gas Source Use Types Transit Bus, Passenger Car, Passenger Truck, Motorcycle Road Type Urban Unrestricted Access Pollutants and Processes NO2 / CO / PM2.5 / PM10 / Acetaldehyde / Acrolein / Benzene / 1,3- Butadiene / Formaldehyde Vehicle Age Distribution MOVES defaults based on years selected. Upon processing of the MOVES outputs, the highest monthly value was selected, which represents a worst-case emission rate. The emission rates used in the assessment for idling and moving buses are shown in Table 19. Table 19: MOVES Output Emission Factors for Diesel Transit Buses for 2011 Contaminant Diesel Buses Passenger Vehicles Idle (g/v-hr) 20 km/hr (g/vmt) Idle (g/v-hr) 20 km/hr (g/vmt) NO CO PM 2.5 Total PM 10 Total Acetaldehyde Acrolein Benzene ,3-Butadiene Formaldehyde In addition to tailpipe emissions, re-suspension of particulate matter from buses driving on site as well as from vehicles driving in the parking lot was considered. These emissions are estimated using empirically derived values presented by the U.S. EPA in their AP-42 report. The emissions factors for re-suspended PM were estimated by using the following equation from U.S. EPA s Document AP-42 report, Chapter and are summarized in Table 20. A silt loading factor of was used for the buses driving on-site, as per MOECC guidance, since the facility has limited access and the buses will be moving very slowly onsite, and are therefore not likely to re-suspend a large amount of particulate matter. A silt loading factor of 0.2 was used for vehicles in the parking lot, which is the recommended silt loading factor for roadways with unrestricted access and an annual average daily traffic (AADT) count of Novus Environmental 28

35 McNicoll Garage Air Quality Assessment December 3, 2014 E = k(sl) 0.91 (W) 1.02 Where: E = the particulate emission factor k = the particulate size multiplier sl = silt loading W = average vehicle weight (Assumed 3 Tons based on Toyota fleet data and U.S. EPA vehicle weight and distribution) Table 20: Re-Suspended Particulate Matter Emission Factors Vehicle Type AADT K (PM 2.5/PM 10) sl (g/m 2 ) W (Tons) E (g/vmt) PM2.5 PM10 Buses < / Cars (Parking Lot) 500-5, / Heating Equipment and Standby Generator Emission Rates All of the heating equipment will be equipped with low-nox burners. For NO2 emissions from the boilers, it was conservatively assumed that 100% of NOx would convert to NO2. Emission rates for each piece of heating equipment were calculated based on the individual heating input, and emission rates for small boilers provided in the U.S. EPA AP-42 Ch. 1.4 Combustion Natural Gas Combustion for low-nox burners. The 800 kw emergency generator was assumed to be a low-nox generator with a maximum emission rate of 2 g/bhp-hr. We understand that the design team will select a unit with this emission rate, or lower. Paint Booth and Shop Areas The majority of the products being used at the facility will be applied using a HVLP spray gun. Of the products being used by hand, all but one are solid at room temperature. For these contaminants, it was therefore assumed that there would be no emissions. The one contaminant applied by hand that is not solid at room temperature (styrene) was assessed as sprayed, which is conservative, because a much higher volume would be used when sprayed as opposed to applied by hand. For products applied with the spray gun, an average applied transfer efficiency rate of liquid being sprayed was determined from the U.S. EPA Environmental Technology Verification Program. In this program, several HVLP spray guns were tested for transfer of sprayed liquids onto a product. Of all the studies performed, the average transfer efficiency rate was 58%. It was therefore assumed that 42% of the product would not be applied to the buses, and would therefore be emitted out the stack. Novus Environmental 29

36 McNicoll Garage Air Quality Assessment December 3, 2014 Total emission rates were determined by summing the weight percentage of each contaminant in every product, and then multiplying the weight percentage by the flow rate of the spray gun (volume of product being used) and by the density of the contaminant, to determine a mass flow emission rate. The total emissions were then multiplied by 0.42 to represent the percent of product emitted through the stack. The emission rates for each contaminant and sample calculations are shown in Appendix B. It was assumed that only one product would be used at a time. Note that a density of 1 g/cm 3 was assumed for contaminants for which a density was not available. These are all contaminants for which there is no recommended guideline. It was noted that the weight percentage of Naphtha (petroleum) was 100% in the grease remover product, which resulted in a high emission rate. The conservative usage rate for this product was therefore further refined to reflect usage at the facility. TTC noted that one 6.36 US Gallon drum would last for one year, and would be used daily. This equates to a usage rate of L/min, for daily application. This usage amount was used to further refine the emission rate for the Naphtha (petroleum) contaminant only. The maximum predicted usage volumes using the HVLP spray gun were used to determine emission rates for all other contaminants. Liquid Storage Tanks Total vapour emissions from each of the tanks was determined using the U.S. EPA TANKS model, which is based on AP-42 Ch. 7.1 Organic Liquid Storage Tanks. Chemical properties of the tank products, fill rates and local meteorological data were all considered in the TANKS calculations. Both standing losses (emissions due to evaporation of product in the tank) and working losses (evaporation during filling) were considered to determine total emissions. It was assumed that working losses occurred throughout the entire day. This is conservative since working losses would typically only occur for a few overnight hours. The maximum predicted monthly emission rate was assumed to occur for the entire year, to be conservative. Table 21 shows the maximum total emissions for each tank. Novus Environmental 30

37 McNicoll Garage Air Quality Assessment December 3, 2014 Table 21: TANKS Model Emission Rates Tank ID Product Maximum Monthly Vapour Loss (g/s) T-1 Diesel 8.79E-03 T-2 Diesel 8.79E-03 T-3 Diesel 8.79E-03 T-4 Engine Oil 1.14E-04 T-5 Engine Oil 1.14E-04 T-6 Trans. Fluid 7.59E-05 T-7 Trans. Fluid 7.59E-05 T-8 Engine Coolant 5.62E-04 T-9 Windshield Fluid 1.26E-01 T-10 Gear Oil 3.39E-08 T-11 Waste Oil 1.14E-04 T-12 Waste Glycol 2.81E-04 As discussed in Section 2.3, benzene is the most volatile contaminant present in the tanks, and was assessed as a worst-case emitted contaminant for the tanks. Benzene vapour percentage in diesel headspace was determined from the U.S. EPA SPECIATE database, which is the EPA s repository of volatile organic gas and particulate matter (PM) speciation profiles of air pollution sources. Of the available measurements of diesel headspace in the SPECIATE database, a maximum benzene content of 0.9% was identified for Super America Diesel, and used for this assessment to be conservative. This benzene vapour content was applied to the TANKS output of total vapour loss to determine the benzene emissions from each tank. It was assumed that the vapour headspace in the tanks containing coolant and windshield fluids would be comprised of 100% propylene glycol and isopropyl alcohol, the identified criteria contaminants from these products, to be conservative. Total vapour emissions predicted from the TANKS model were small, and were assessed using the screening-out assessment of contaminants that are emitted in negligible amounts, in accordance with MOECC Guideline A- 10 Procedure for Preparing an Emission Summary and Dispersion Modelling Report. Total facility-wide emissions for each of the contaminants were considered in the assessment of negligibility Benzene is emitted from the buses and vehicles in addition to the tanks, while propylene glycol and isopropyl alcohol are emitted only from the tanks. Propylene glycol and isopropyl alcohol were found to be negligible and were not assessed further. Total benzene emissions did not meet the negligibility criteria, and were modelled in detail to predict impacts. Results of the assessment of negligibility for the liquid storage tanks are shown in Table 22. Further details regarding the assessment of negligibility calculations discussed in Section and sample calculations are provided in Appendix B, for the paint booth assessment of negligibility. Novus Environmental 31

38 McNicoll Garage Air Quality Assessment December 3, 2014 Table 22: Assessment of Negligibility for Liquid Storage Tanks Compound O.Reg 419 Limit O.Reg Guideline Averaging Time (hours) Emission Rate (g/s) Emission Threshold (g/s) Negligible? Benzene NO Propylene Glycol YES Isopropyl Alcohol YES 4.6 Modelling Methods Air Dispersion Modelling Using AERMOD The U.S. EPA s AERMOD dispersion model, based on the Gaussian plume equation, was used to predict air quality impacts from emissions at the McNicoll Bus Garage. The model inputs include local building information, topography, sensitive receptor locations, meteorology, emission rates and stack parameters. AERMOD uses this information to calculate hourly, 8- hour or 24-hour averages for the contaminants of interest at the identified sensitive receptor locations. Combined impacts were assessed for all emissions from the buses, employee vehicles, heating equipment and liquid storage tanks. Impacts from the contaminants from the paint booth and shop areas were assessed separately, as contaminants did not overlap with the remaining activities. Assessment of Negligibility for Contaminants in the Paint Booth and Shop Areas Many of the contaminants are small fractions of the products being used, and will therefore be emitted in small amounts. As such, a screening-out assessment of contaminants that are emitted in negligible amounts was conducted in accordance with MOECC Guideline A-10 Procedure for Preparing an Emission Summary and Dispersion Modelling Report. Emission rates for each contaminant were assessed against the emission threshold, using the urban dispersion factor at 20 m, the smallest separation distance provided in Guideline A-10. If the emission rate was less than the emission threshold, the contaminant was determined negligible and not assessed further. Contaminants that were not found to be negligible were modelled in AERMOD and assessed against their applicable guidelines for the applicable averaging periods. Contaminants that do not have a guideline were modelled in AERMOD and results have been presented. Sample calculations for the assessment of negligibility are shown in Appendix B. Novus Environmental 32

39 McNicoll Garage Air Quality Assessment December 3, Results 5.1 Combined Results for All Emission Sources, Not Including the Paint Booth and Shop Areas The maximum impacts were predicted to occur at the nearby senior s residence, at ground level, for all contaminants with the exception of benzene. The maximum benzene impacts were predicted to occur at the vacant lot to the east of the facility, at which current zoning would allow for a day care or educational facility. Note that NO2 impacts are due to emissions from buses, heating equipment, generators and vehicles in the parking lot. The benzene impacts are due to emissions from buses, vehicles in the parking lot and fugitive emissions from the tanks. The remaining pollutants are emitted only from buses and vehicles in the parking lot. The maximum facility induced concentrations were added to the maximum, 90 th and average 5-year background concentrations to show worst-case and reasonable worst-case impacts. Note that this methodology results in conservative worst-case concentrations as the maximum facility induced concentration likely does not occur at the same time as the maximum background concentration. The worst-case concentrations are shown in Table 23. Contour plots showing the concentrations surrounding the facility are shown in Appendix C. Note that since this assessment was completed as part of an environmental assessment, impacts were only presented at the identified sensitive receptors. For the Environmental Compliance Approval, impacts at property boundary line will need to be assessed. Impacts at the property line from the facility alone are shown in the contour plots in Appendix C, and are predicted to meet the guidelines for all contaminants and averaging periods. Contaminant Table 23: Worst-Case Predicted Concentrations as a Percentage of the Guideline Averaging Period Maximum Concentration Due to Facility Alone (µg/m 3 ) Maximum Concentration Due to Facility Alone (as % of Standard) Combined Concentration as % of Standard (Ambient + Project) Maximum 90 th Percentile Average Additional # of Guideline Exceedances due to Project Over 5 Years NO 2 1-hour % 79% 53% 47% 24-hour 32 16% 61% 40% 30% CO 1-hour % 6% 1% 1% 8-hour % 12% 3% 2% PM hour 1.6 6% 139% 53% 30% 3 PM hour 2 4% 137% 51% 29% 1 Acetaldehyde 24-hour % 1% <1% <1% Acrolein 24-hour % 38% 25% 20% Benzene 24-hour % 107% 50% 35% 6 1,3-Butadiene 24-hour % 3% 1% 1% Formaldehyde 24-hour % 13% 8% 5% 1 CWS guideline for PM2.5 is based on an average annual 98 th percentile concentration, averaged over 3 consecutive years. The maximum combined 3-year rolling average of the annual 98 th percentile concentration was 22.14, which is 82% of the guideline. Novus Environmental 33

40 McNicoll Garage Air Quality Assessment December 3, 2014 Overall, the maximum concentrations due to the facility alone are 8% or less of the applicable standard, except for NO2 concentrations, for which the worst-case concentration is 40% of the NO2 1-hour guideline. Combined with the maximum measured background concentration, two pollutants are above the guideline: PM10 and benzene. Note that though maximum concentration of PM2.5 exceeded the CWS, the guideline for PM2.5 is based on an average annual 98 th percentile concentration, averaged over three consecutive years. Combining the maximum facility induced concentration with the background 98 th percentile concentration of PM2.5 for each of the five years modelled, the maximum rolling 98 th percentile average was µg/m 3, which is below the guideline. Background PM10 concentrations already exceed the guideline 12 times in five years. Combining the maximum facility induced concentration with background concentrations, one additional exceedance of the guideline is predicted to occur for a total of 13 times, which is less than 1% of the time. The maximum background benzene concentration is 99% of the standard. Combining the maximum facility induced concentrations with background concentrations causes a slight exceedance of the standard. As mentioned earlier, ambient measured benzene concentrations are monitored infrequently, typically every 6 days. To complete the dataset, the measured concentration was applied for all days between measurements when there were 6 days or less between measurements. The maximum benzene concentration, which was 99% of the standard, was based on one measured value and then applied to 6 days of the five-year dataset. Therefore, combined concentrations add slightly to the background for a combined concentration of 107% of the guideline, conservatively predicted to occur for 6 days due to the methods described. It is important to note that these exceedances are primarily due to background concentrations and the contribution from the facility is small. All other contaminants met the guidelines with no exceedances. It should be noted that this approach, combining the maximum values to the maximum ambient measurements is extremely conservative. It is not likely that the maximum facility concentration will occur at the same time as the maximum ambient concentration. Furthermore, it is likely that the combined maximum concentration will only occur for one hour of one day, and it is not representative of what can be expected on a typical day. 5.2 Results for the Paint Booth and Shop Areas From the paint booth and shop areas, 29 of the 66 contaminants were found to have negligible emissions. The remaining contaminants were modelled in AERMOD. Results of the AERMOD modelling showed that all contaminants met their respective guidelines at the nearest sensitive receptor. Results of the modelling in comparison the guidelines are shown in Appendix B. Note that some contaminants do not have a recommended guideline, however, the predicted worst-case concentrations at the nearest sensitive receptor have been presented to show their impacts. Novus Environmental 34

41 McNicoll Garage Air Quality Assessment December 3, Conclusions The potential effects of the proposed facility on local air quality have been assessed. The following conclusions and recommendations are a result of this assessment. The maximum combined concentrations were all below their respective MOECC guidelines or CWS, with the exception of PM10 and benzene. Frequency analysis determined that the project exceeded the PM10 and benzene guidelines one and six additional days, respectively, over the 5-year period. This equates to <1% of the time. It is recommended that low-nox burners be installed on all heating equipment, in accordance with this assessment. It is recommended that the design team select a generator unit with a maximum NOx emission rate of 2 g/bhp-hr. Upon final selection of equipment and exhaust fans for the facility, an Environmental Compliance Assessment will need to be completed and submitted to the MOECC. Novus Environmental 35

42 McNicoll Garage Air Quality Assessment December 3, References CCME, Canadian Council of Ministers of the Environment. Canada-Wide Standards of Particulate Matter and Ozone. Endorsed by CCME Council of Ministers, Quebec City. [Online] Environment Canada Priority Substances List Assessment Report: Respirable Particulate Matter Less Than or Equal to 10 Microns. Canadian Environmental Protection Act, Environment Canada, Health Canada. [Online] Health Canada National Ambient Air Quality Objectives for Particulate Matter Part 1: Science Assessment Document. Health Canada. A report by the CEPA/FPAC Working Group on Air Quality Objectives and Guidelines. Lall, R., Kendall, M., Ito, K., Thurston, G., Estimation of historical annual PM2.5 exposures for health effects assessment. Atmospheric Environment 38(2004) Ontario Publication 6570e, Ontario's Ambient Air Quality Criteria. Standards Development Branch, Ontario Ministry of the Environment. Ontario Ministry of the Environment, Transboundary Air Pollution in Ontario. Queens Printer for Ontario. Randerson, D., Atmospheric Science and Power Production. United States Department of Energy. Seinfeld, J.H. and Pandis, S.P.,2006. Atmospheric Chemistry and Physics From Air Pollution to Climate Change. New Jersey: John Wiley & Sons. United States Environmental Protection Agency, AERSURFACE User s Guide. USEPA. United States Environmental Protection Agency, Document AP 42, Volume I, Fifth Edition, Chapter USEPA. United States Environmental Protection Agency, MOVES 2010 Highway Vehicles: Population and Activity Data. USEPA. United States Environmental Protection Agency, AP 42, Fifth Edition, Volume I Chapter 1: External Combustion Sources, Chapter 1.4: Natural Gas Combustion. USEPA United States Environmental Protection Agency, AP 42, Fifth Edition, Volume I Chapter 7: Liquid Storage Tanks. USEPA Novus Environmental 36

43 McNicoll Garage Air Quality Assessment December 3, 2014 United States Environmental Protection Agency, AP 42, Fifth Edition, Volume I Chapter 7: Liquid Storage Tanks. USEPA United States Environmental Protection Agency, Environmental Technology Verification Program Pollution Prevention Coatings and Coating Equipment. USEPA [ United States Environmental Protection Agency, SPECIATE DATABASE: Diesel Headspace Vapor Super America Diesel. USEPA WHO WHO air quality guidelines global update Report on a Working Group meeting, Boon, Germany, October 18-20, Novus Environmental 37

44 Appendix A Heating Equipment Specifications

45 This page intentionally left blank for 2-sided printing purposes

46 McNicoll Garage Air Quality Assessment Appendix A Table A1 Heating Equipment Parameters Source Heating Input (kw) Heat Recovery (kw) Modelled Heat Input (kw) Stack Height Above Grade (m) Diameter (m) Exit V (m/s) Flow m3/s NO x Emission Rate (g/s) generator boiler boiler AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU Novus Environmental i

47 McNicoll Garage Air Quality Assessment Appendix A Source Heating Input (kw) Heat Recovery (kw) Modelled Heat Input (kw) Stack Height Above Grade (m) Diameter (m) Exit V (m/s) Flow m3/s NO x Emission Rate (g/s) AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU AHU UH UH UH UH UH UH UH UH UH UH UH Novus Environmental ii

48 McNicoll Garage Air Quality Assessment Appendix A Source Heating Input (kw) Heat Recovery (kw) Modelled Heat Input (kw) Stack Height Above Grade (m) Diameter (m) Exit V (m/s) Flow m3/s NO x Emission Rate (g/s) UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH Novus Environmental iii

49 McNicoll Garage Air Quality Assessment Appendix A Source Heating Input (kw) Heat Recovery (kw) Modelled Heat Input (kw) Stack Height Above Grade (m) Diameter (m) Exit V (m/s) Flow m3/s NO x Emission Rate (g/s) UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH Novus Environmental iv

50 McNicoll Garage Air Quality Assessment Appendix A Source Heating Input (kw) Heat Recovery (kw) Modelled Heat Input (kw) Stack Height Above Grade (m) Diameter (m) Exit V (m/s) Flow m3/s NO x Emission Rate (g/s) UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH UH Novus Environmental v

51 This page intentionally left blank for 2-sided printing purposes

52 Appendix B Paint Booth and Shop Area Assessment

53 This page intentionally left blank for 2-sided printing purposes

54 McNicoll Garage Air Quality Assessment Appendix B 1.0 Products to be Used at the McNicoll Facility Table B1 lists of the products which will be used at the McNicoll facility and the contaminants which they contain. Contaminants and the weight percentage of each product was determined from the MSDS sheets for each product, provided by TTC. The application rate and usage frequency were also provided by TTC. Table B1 Products Used at the McNicoll Facility Chemical Product Hi-Strength Spray Aerosol Adhesive Fastbond Contact Adhesive 2000-NF, Blue Wax and Grease Remover Contaminant % by Weight Max % Application Method Usage Frequency (if known) Dimethyl Ether Sprayed Daily Methyl Acetate Sprayed Daily Non-volatile Components Sprayed Daily Cyclohexane Sprayed Daily 1,1-Difluoroethane Sprayed Daily Pentane Sprayed Daily Water Sprayed - Polychloroprene Sprayed - Glycerol Esters of Rosin Acids Sprayed - Phenolic Rosin Sprayed - Toluene Sprayed - Methyl Alcohol Sprayed - Zinc Oxide Sprayed - 2,2'-Methylenebis (6-Tert-Butyl-P- Cresol) Sprayed - Rosin Sprayed - Naphtha (petroleum), hydrotreated heavy Sprayed Daily Solvent naphtha (petroleum), light aliph Sprayed Daily Heptane Sprayed Daily Novus Environmental i

55 McNicoll Garage Air Quality Assessment Appendix B Chemical Product Contaminant % by Weight Max % Application Method Usage Frequency (if known) Methylcyclohexane Sprayed Daily Toluene Sprayed Daily Silanated Modified Polyether Flooring Sealant N/A N/A 0.00 Hand Daily Talc Hand Twice/week Polyester Resin Hand Twice/week Styrene Hand Twice/week Light Weight Bodyfiller Magnesite Hand Twice/week Calcium Carbonate Hand Twice/week Inert Filler Hand Twice/week Titanium Dioxide Hand Twice/week Talc Hand Twice/week Polyester Resin Hand Twice/week Fiberglass Reinforced Filler Styrene Hand Twice/week Magnesite Hand Twice/week Dolomite Hand Twice/week Inert Filler Hand Twice/week Hexane Sprayed Daily Kleen Slip Silicone Lubricant Petroleum Distillates Sprayed Daily Propane (Propellant) Sprayed Daily Isobutane (Propellant) Sprayed Daily Water > Sprayed - Lens and Mirror Cleaner Sodium lauryl sulfate < Sprayed - Titanium Dioxide Pigment < Sprayed - Three Omni-Pak MasterBlend EZ Touch DV Cans Propane Sprayed times/week Three Acetone Sprayed times/week Methyl Ethyl Ketone Sprayed Three Novus Environmental ii

56 McNicoll Garage Air Quality Assessment Appendix B Chemical Product Omni-Pak for Enamel Self Etching Primer Black Urethane Based Adhesive/Sealant (Sikaflex-252) Urethane Based Adhesive/Sealant (Sikaflex-221) Contaminant % by Weight Max % Application Method Ethyl 3-Ethoxypropionate Sprayed Usage Frequency (if known) times/week Three times/week Propane Sprayed - Butane Sprayed - Ethylbenzene Sprayed - Xylene Sprayed - Acetone Sprayed - Petroleum gases, liquefied, sweetened Sprayed Daily Acetone Sprayed Daily Ethyl Acetate Sprayed Daily Isobutyl Acetate Sprayed Daily Toluene Sprayed Daily Butanone Sprayed Daily Cellulose Nitrate Sprayed Daily Quartz Sprayed Daily n-butyl Acetate Sprayed Daily Propan-2-ol Sprayed Daily Xylene Sprayed Daily Tris (methylphenyl) Phosphate Sprayed Daily Xylene Hand Daily Polyol and Isocyanate Prepolymer Hand Daily Amorphous Silica Hand Daily Methylene Bis Phenyl Isocyanate Hand Daily Calcium Oxide Hand Daily Xylene Hand Daily Polyol and Isocyanate Prepolymer Hand Daily Solopol Hand Cleanser N/A N/A 0.00 Hand Daily Novus Environmental iii

57 McNicoll Garage Air Quality Assessment Appendix B Chemical Product Lacquer Thinner WD-40 Aerosol Contaminant % by Weight Max % Application Method Usage Frequency (if known) Lt. Aliphatic Hydrocarbon Solvent Sprayed Daily V. M. & P. Naphtha Sprayed Daily Toluene Sprayed Daily Ethylbenzene Sprayed Daily Xylene Sprayed Daily Methanol Sprayed Daily 2-Propanol Sprayed Daily 2-Methyl-1-propanol Sprayed Daily 2-Butoxyethanol Sprayed Daily Acetone Sprayed Daily Methyl n-amyl Ketone Sprayed Daily Isobutyl Acetate Sprayed Daily Aliphatic Petroleum Distillates Sprayed Daily Petroleum Base Oil Sprayed Daily Non-Hazardous Ingredients < Sprayed Daily Carbon Dioxide Sprayed - Novus Environmental iv

58 McNicoll Garage Air Quality Assessment Appendix B 2.0 Assessment of Negligibility The assessment of negligibility was conducted in accordance with MOECC Guideline A-10 Procedure for Preparing an Emission Summary and Dispersion Modelling Report. Emission rates for each contaminant were assessed against the emission threshold, using the urban dispersion factor at 20 m, the smallest separation distance provided in Guideline A-10. If the emission rate was less than the emission threshold, the contaminant was determined negligible and not assessed further. Sample calculations for determine the emission rate and emission threshold are shown below for butane. Table B-2 shows the results of the assessment of negligibility for each product. It was assumed one product would be used at a time. Note that for contaminants with a 1-hour standard, a nozzle flow rate for the spray gun of 0.42 L/min was modelled, as this is the maximum amount of product that could be used in an hour. A flow rate of 0.13 L/min was modelled for contaminates with a 24-hour standard, as this is the average amount of product that could be used in one day. One pollutant, naphtha (petroleum) had a high weight percentage (100%), therefore a conservatively high emission rate was predicted. An actual product usage of 6.36 gallons per year was provided by TTC, which equates to L/min, for daily usage. This usage rate was used only for the assessment of naphtha (petroleum). Sample Calculation Butane 0.5 X MOE POI Limit (µg/m Emission Threshold (g/s) = 3 ) Dispersion Factor (µg/m 3 per g/s emission) 0.5 X Emission Threshold Butane (g/s) = = g/s 8700 Spray Gun Flow Rate (L/min) Emission Rate (g/s) = X density (g/cm cm3 ) X X Wt % X Transfer Efficiency Rate 60 s/min L Emission Rate Butane (g/s) = 0.13 (L/min) 60 s/min X (g/cm3 ) 1000 cm3 L X 0.21 X 0.42 = g/s Emission Rate for Butane ( g/s) < Emission Threshold for Butane (1.31 g/s), therefore Butane emissions are considered negligible. Novus Environmental v

59 McNicoll Garage Air Quality Assessment Appendix B Table B-2: Assessment of Negligibility Compound CAS # Density (g/cm 3 ) O.Reg 419 Limit O.Reg Guideline JSL Limit Averaging Time (hours) Sum of Percent Weights Emission Rate (g/s) Emission Threshold (g/s) Negligible? Butane YES Methylcyclohexane YES Pentane YES Carbon Dioxide YES Ethyl Acetate YES Propane (Propellant) YES Propane YES Propane YES n-butyl Acetate YES Ethylbenzene YES Butane YES Methylcyclohexane YES Pentane YES Methyl n-amyl Ketone YES Hexane YES Cyclohexane YES 2-Butoxyethanol YES Carbon Dioxide YES Heptane YES Methyl Alcohol YES Methanol YES Propan-2-ol YES Novus Environmental vi

60 McNicoll Garage Air Quality Assessment Appendix B Compound CAS # Density (g/cm 3 ) O.Reg 419 Limit O.Reg Guideline JSL Limit Averaging Time (hours) Sum of Percent Weights Emission Rate (g/s) Emission Threshold (g/s) Negligible? Isopropyl Alcohol YES 1,1-Difluoroethane YES 2-Methyl-1- propanol YES Butanone YES Petroleum Distillates YES Propane (Propellant) YES Propylene Glycol YES Naphtha (petroleum), hydrotreated heavy YES Styrene NO n-butyl Acetate NO Isobutyl Acetate NO Dimethyl Ether NO Sodium Xylenesulfonate NO Glycerol Esters of Rosin Acids NO Isobutane (Propellant) NO Acetone NO Ethyl 3- Ethoxypropionate NO Methyl Acetate NO Toluene NO Dimethyl Ether NO Novus Environmental vii

61 McNicoll Garage Air Quality Assessment Appendix B Compound CAS # Density (g/cm 3 ) O.Reg 419 Limit O.Reg Guideline JSL Limit Averaging Time (hours) Sum of Percent Weights Emission Rate (g/s) Emission Threshold (g/s) Negligible? Xylene NO Titanium Dioxide Pigment NO Quartz NO Glycerol Esters of Rosin Acids NO Acetone NO Isobutane (Propellant) NO Methyl Ethyl Ketone NO Methyl Acetate NO Polychloroprene NO Solvent naphtha (petroleum), light aliph NO Petroleum gases, liquefied, sweetened NO V. M. & P. Naphtha NO Zinc Oxide NO Sodium lauryl sulfate NO 2-Propanol NO Phenolic Rosin NO 2,2'-Methylenebis (6-Tert-Butyl-P- Cresol) NO Rosin NO Novus Environmental viii

62 McNicoll Garage Air Quality Assessment Appendix B Compound CAS # Density (g/cm 3 ) O.Reg 419 Limit O.Reg Guideline JSL Limit Averaging Time (hours) Sum of Percent Weights Emission Rate (g/s) Emission Threshold (g/s) Negligible? Cellulose Nitrate NO Tris (methylphenyl) Phosphate NO Lt. Aliphatic Hydrocarbon Solvent NO Aliphatic Petroleum Distillates NO Petroleum Base Oil NO Novus Environmental ix

63 McNicoll Garage Air Quality Assessment Appendix B 3.0 AERMOD Modelling Results Contaminants that were not found to be negligible were modelled in AERMOD. Both the paint stack and body shop stacks were modelled to determine which stack would provide worst case results. The paint booth stack is 8 m above rooftop, and has a flow rate of 19.8 m 3 /s, as per the provided mechanical schedule. A large stack diameter of 1 m was modelled with a low exit velocity, to provide conservative predictions. The other stacks from the shop areas were modelled with an average diameter of 0.2 m and flow rate of 0.2 m 3 /s. The modelling showed lower dispersion levels for the shop area stacks (resulting in higher concentrations), therefore it was assumed that all contaminants could be emitted from the shop area stacks, in order to predict worst case results. Table B-3 shows the AERMOD results for each contaminant, and whether or not the guideline was met. The guideline was met for all contaminants. Table B-3 AERMOD Results Compound CAS # Density (g/cm 3 ) O.Reg 419 Limit O.Reg Guideline JSL Limit Averaging Time (hours) Sum of Percent Weights Emission Rate (g/s) AERMOD Result (µg/m 3 ) Meets Guideline? Styrene PASS n-butyl Acetate PASS Isobutyl Acetate PASS Dimethyl Ether PASS Glycerol Esters of Rosin Acids PASS Isobutane (Propellant) PASS Acetone PASS Ethyl 3- Ethoxypropionate PASS Methyl Acetate PASS Toluene PASS Dimethyl Ether PASS Quaternary PASS Novus Environmental x

64 McNicoll Garage Air Quality Assessment Appendix B Compound CAS # Density (g/cm 3 ) O.Reg 419 Limit O.Reg Guideline JSL Limit Averaging Time (hours) Sum of Percent Weights Emission Rate (g/s) AERMOD Result (µg/m 3 ) Meets Guideline? ammonium 9 chloride Xylene PASS Titanium Dioxide Pigment PASS Quartz PASS Glycerol Esters of Rosin Acids PASS Acetone PASS Isobutane (Propellant) PASS Methyl Ethyl Ketone PASS Methyl Acetate PASS Polychloroprene PASS Solvent naphtha (petroleum), light aliph PASS Petroleum gases, liquefied, sweetened PASS V. M. & P. Naphtha PASS Zinc Oxide No Guideline Sodium lauryl sulfate No Guideline 2-Propanol No Guideline Phenolic Rosin No Guideline 2,2'-Methylenebis (6-Tert-Butyl-P No Guideline Novus Environmental xi

65 McNicoll Garage Air Quality Assessment Appendix B Compound CAS # Cresol) Density (g/cm 3 ) O.Reg 419 Limit O.Reg Guideline JSL Limit Averaging Time (hours) Sum of Percent Weights Emission Rate (g/s) AERMOD Result (µg/m 3 ) Meets Guideline? Rosin No Guideline Cellulose Nitrate No Guideline Tris (methylphenyl) Phosphate No Guideline Lt. Aliphatic Hydrocarbon Solvent No Guideline Aliphatic Petroleum Distillates No Guideline Petroleum Base Oil No Guideline Novus Environmental xii

66 Appendix C Contour Plots for Each Contaminant

67 This page intentionally left blank for 2-sided printing purposes

68 McNicoll Garage Air Quality Assessment Appendix C Provided below are the contour plots from AERMOD for each of the pollutants and averaging periods assessed. Sensitive receptors are shown as yellow dots. Receptors just west of the facility alogn McNicoll Avenue represent the Mon Sheong residence/long-term care facility, and the other three individual receptors represent the identified churches. Figure C1: Contour Plot of Maximum 1-Hour NO2 Concentration Novus Environmental i

69 McNicoll Garage Air Quality Assessment Appendix C Figure C2: Contour Plot of Maximum 24-Hour NO2 Concentration Novus Environmental ii

70 McNicoll Garage Air Quality Assessment Appendix C Figure C3: Contour Plot of Maximum 1-Hour CO Concentration Novus Environmental iii

71 McNicoll Garage Air Quality Assessment Appendix C Figure C4: Contour Plot of Maximum 8-Hour CO Concentration Novus Environmental iv

72 McNicoll Garage Air Quality Assessment Appendix C Figure C5: Contour Plot of Maximum 24-Hour PM2.5 Concentration Novus Environmental v

73 McNicoll Garage Air Quality Assessment Appendix C Figure C6: Contour Plot of Maximum 24-Hour PM10 Concentration Novus Environmental vi

74 McNicoll Garage Air Quality Assessment Appendix C Figure C7: Contour Plot of Maximum 24-Hour Acetaldehyde Concentration Novus Environmental vii

75 McNicoll Garage Air Quality Assessment Appendix C Figure C8: Contour Plot of Maximum 24-Hour Acrolein Concentration Novus Environmental viii

76 McNicoll Garage Air Quality Assessment Appendix C Figure C9: Contour Plot of Maximum 24-Hour Benzene Concentration Novus Environmental ix

77 McNicoll Garage Air Quality Assessment Appendix C Figure C10: Contour Plot of Maximum 24-Hour 1,3-Butadiene Concentration Novus Environmental x

78 McNicoll Garage Air Quality Assessment Appendix C Figure C11: Contour Plot of Maximum 24-Hour Acetaldehyde Concentration Novus Environmental xi

79 Wang, Joanne To: Subject: Occhiogrosso, Leonard RE: TPH Comment McNicoll Bus Garage EPR From: Howard Shapiro Sent: Monday, December 22, :21 AM To: Barbara Lachapelle; MacDonald, Jason; Dimovski, John Cc: David McKeown; Jann Houston; Reg Ayre; Nagler, David; Romano, Lito; Favaro, Marcello Subject: TPH Comment McNicoll Bus Garage EPR Hi John, Toronto Public Health has completed our assessment of the air quality portion of the McNicoll Bus Garage EPR and have three comments that are listed below. We should have our assessment of noise to you in the next day or two. I have also attached the a copy of the Local Air Quality Study for Wards 5 & 6 which is relevant to the second point. If you have any questions please feel free to contact myself or Barbara Lachapelle. General AQ Comments: 1. Modelling of emissions the modelling of impacts from the facility should include estimating and modelling emissions from the additional traffic (due to the TTC facility) on the nearby roads. This cumulative impact assessment approach has previously been used by TPH on other projects, such as the BBTCA expansion project, and it allows for a more robust assessment of potential health impacts associated with the proposal. 2. The use of health protective comparison benchmarks Novus currently compares the estimated AQ concentrations against the Ontario AAQCs. Although that is appropriate for evaluating regulatory compliance, it is a departure from the TPH approach of evaluating health impacts. We recommend you evaluate the projected AQ impacts against health based carcinogenic and non carcinogenic benchmarks, and the AQBAT risk coefficients to evaluate health risk from criteria air pollutants in addition to the Ontario AAQCs. In the past, this approach has been used by TPH for Local Air Quality (LAQ) studies, the BBTCA project, and is currently being utilized for other proposals. For your convenience we have attached a copy of the Local Air Quality Study for Wards 5 & 6, the approach is outlined in the body of the reports and the benchmarks are listed in Appendix A. 3. Air monitoring Novus made a number of assumptions in order to justify the use of existing AQ data from monitoring stations located at a significant distance from the site. Although we recognize the methodology utilized by Novous was fairly conservative, the data may not represent the existing AQ conditions in the area. We recommend conducting air monitoring to validate the baseline data assumptions (prior to the commencement of any site/construction activities), during the construction phase (to ensure construction impacts are minimized similar approach as at the Leslie Barns project), and once the facility is built to ensure impacts have been properly assessed and mitigated. 1

80 Howard Shapiro MD MSc FRCPC Associate Medical Officer of Health & Acting Director Healthy Environments Toronto Public Health 277 Victoria Street, 5th Floor Toronto, Ontario M5B 1W2 Tel: The information transmitted is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review retransmission dissemination or other use of or taking any action in reliance upon this information by persons or entities other than the intended recipient or delegate is strictly prohibited. If you received this in error please contact the sender and delete the material from any computer. The integrity and security of this message cannot be guaranteed on the Internet. The sender accepts no liability for the content of this or for the consequences of any actions taken on the basis of information provided. The recipient should check this and any attachments for the presence of viruses. The sender accepts no liability for any damage caused by any virus transmitted by this . This disclaimer is property of the TTC and must not be altered or circumvented in any manner. 2

81 Health Assessment for the Cumulative Air Quality Modelling Study Wards 5 and 6 including the South Etobicoke and Lakeshore Neighbourhoods March 2014

82 Reference: Toronto Public Health. Health Assessment for the Cumulative Air Quality Modelling Study: Wards 5 and 6. Toronto: March Authors: Stephanie Gower and Ronald Macfarlane Peer Review: We are grateful for the comments of two peer reviewers who provided valuable feedback on the methodology used to create this report: Dr. Stephen McColl, University of Waterloo Dr. Dave Stieb, Health Canada Acknowledgements: Report at: We also thank the following people for their advice and insightful comments regarding this report: Monica Campbell and Kate Bassil (Toronto Public Health), and Christopher Morgan (Toronto Environment Office). For Further Information: Healthy Public Policy Toronto Public Health 277 Victoria Street, 7th Floor Toronto, Ontario Canada M5B 1W2 Tel:

83 Air Quality Health Assessment Wards 5 and 6 i Table of Contents Executive Summary... ii 1) Cumulative risk for non-cancer effects... iii 2) Cumulative risk of cancer... iii 3) Cumulative risk from common air contaminants (CACs)... iv How does this compare with the previous local air quality study?... v Background... 1 Assessing Air Quality... 2 Modelling Approach... 4 Selection of Air Contaminants to be modelled... 5 Health Assessment... 7 Overview... 7 Methods for the Health Assessment... 8 Calculating Cumulative Cancer Risk for Individuals... 8 Specific Substances and Weights... 9 Calculating cumulative non-cancer hazard Specific Substances and Weights Calculating Cumulative Risk from Common Air Contaminants (CACs) Specific Substances and Weights Results: Air Quality Model Findings Carcinogens Carcinogens Cumulative Risk Non-Carcinogens Non-carcinogens Cumulative risk Criteria Air Contaminants Discussion How does this compare with the previous local air quality study? Limitations of the Health Assessment Methodology References Appendix A... A-1

84 Air Quality Health Assessment Wards 5 and 6 ii Executive Summary This is the second local air quality study and health assessment conducted for Toronto. The first study was completed for South Riverdale, Leslieville, and the Beach. Toronto Public Health's full assessments of both studies are available at This health assessment complements the local air modelling study conducted by Toronto's Environment and Energy Division (EED) for Etobicoke-Lakeshore (Wards 5 and 6). Based on EED's findings, Toronto Public Health (TPH) prepared a health assessment estimating the cumulative health impacts of air pollution in the area. The health assessment considered the thirty pollutants included in EED's air quality model, which includes the 25 priority substances in Toronto's Environmental Reporting and Disclosure Bylaw (ChemTRAC program) and five other common air pollutants. This local air quality health assessment report covers: The background of the study The modelling approach used and substances chosen for the air quality assessment The health impacts, including the methods and findings, and A discussion of the results of the study. Air modelling studies typically compare the estimated levels of pollutants against air standards or health benchmarks to identify if releases could result in levels of concern in a specific area. The predicted ambient concentrations of most the individual pollutants considered in this study were below Ontario s Ambient Air Quality Criteria (AAQCs). However, the model predicted that levels of nitrogen dioxide (NO 2 ), coarse particulate matter (PM 10 ), fine particulate matter (PM 2.5 ), benzene, and benzo[a]pyrene * (used as an indicator of polycyclic aromatic hydrocarbons) might exceed air quality objectives in some areas, some of the time. The modelling study showed that transportation is the largest local source of these pollutants. As people are exposed to a mixture of pollutants, it is useful to also consider the combined impacts of these pollutants, even when most are individually below levels of concern. The science for assessing the health impacts of mixtures of chemicals continues to evolve and there is no common approach to combined exposures from the complete range of substances considered in this study. This study grouped pollutants according to similar mechanisms of action. This resulted in three categories of health effects, with the cumulative impact estimated for each group of pollutants separately. These categories were: 1) Substances associated with non-cancer effects, for which there is a health threshold; 2) Substances associated with cancer; and * The findings for benzo[a]pyrene require further review and validation.

85 Air Quality Health Assessment Wards 5 and 6 iii 3) Common air contaminants (CACs), which are mainly associated with cardiovascular and respiratory diseases, and which are assumed to have no health threshold. The results of these separate cumulative assessments are described below. 1) Cumulative risk for non-cancer effects In this study, non-cancer health effects include include neurological, immunological, and developmental health impacts. In general, for non-cancer effects it is assumed that there is a threshold of effect a level below which exposure to the substance will have no adverse health impacts. By comparing an exposure level with the threshold, it is possible to assess whether a health impact is expected. Each pollutant considered in the health assessment has a different threshold. To be able to compare them all on the same scale, a measure called the hazard ratio is obtained for each pollutant by dividing the exposure level for that pollutant by its health threshold. If the hazard ratio is less than one, then a person or community is being exposed at a level which current knowledge suggests is not a concern. As well, hazard ratios for multiple substances can be added to estimate a cumulative hazard. Of the 30 substances included in EED's modelling study, 22 are potentially associated with non-cancer health effects. The hazard ratio values for each of the individual noncarcinogenic substances are all much less than one; nickel had the largest hazard ratio at 0.1. This confirms that there is little or no risk of adverse health effects from exposures to these substances individually. When the hazard ratios for the 22 pollutants were added together, the cumulative hazard index is 0.42; this is still well below one. This suggests that the combined exposure to these air pollutants do not pose a health risk for non-cancer effects. 2) Cumulative risk of cancer Carcinogens are substances that are associated with an increased risk of developing cancer over a person's lifetime. For carcinogens, it is assumed that every amount of exposure has a risk of causing cancer. Low levels of exposure are associated with low levels of risk, and the risk rises as exposure increases. The risk of cancer for a single carcinogen is calculated by multiplying the level of carcinogen in the air by a risk factor that represents the likelihood of developing cancer over a lifetime. A risk of one in a million means that one out of every million people exposed would be expected to develop cancer during their lifetime. Toronto Public Health encourages actions to reduce exposures when the risk is above one in one million. There are limitations to this approach. It assumes that the effect of the individual pollutants is in direct proportion to the level of exposure and the effect of each pollutant is additive. In some circumstances, this could overestimate the risk since it does not take into account that different pollutants affect different parts of the body and ignores the natural mechanism of the body to eliminate or detoxify these substances. At the same time, the approach could underestimate the risk since it does not take into account potential interactions between these pollutants that could increase the health impacts.

86 Air Quality Health Assessment Wards 5 and 6 iv In this study, eight of the 19 carcinogens had modeled annual concentrations above the one in one million excess lifetime cancer risk benchmark in parts or the whole of the two wards. These carcinogens were: benzene, chromium (VI), 1,3-butadiene, tetrachloroethylene (or perchloroethylene), formaldehyde, 1,4-dichlorobenzene, acetaldehyde, and benzo[a]pyrene. Toronto Public Health has low confidence in the results for one of the carcinogens, benzo[a]pyrene. The City will gather additional information to verify the estimate of risk for this substance. Except for two of the substances (chromium and tetrachloroethylene), vehicles (on and off-road) are the largest local source of these carcinogens. The estimated risk for the carcinogens, excluding benzo[a]pyrene, was added to give a total estimate of the risk. When the average annual risk is summed across these 18 carcinogenic substances, the average cumulative cancer risk in these two wards is 44 in one million. While 44 in a million is greater than the benchmark that TPH uses for individual cancer risk, the total risk is still quite small. This total risk is less than two percent of the overall cancer incidence rate in Toronto which was around 400 per 100,000 in However, because benzo[a]pyrene was excluded from the risk estimates until further validation is possible, this value is expected to be an underestimate of the average cancer risk in the area. While the cumulative risk was above the one in a million cancer risk benchmark in all parts of the two wards, the areas with more elevated risks tend to be close the Gardiner Expressway and Highway 427. The elevated risk next to these highways is mostly from benzene and 1,3-butadiene. Although the contribution of benzo[a]pyrene to cancer risk was removed from the model pending further validation, this pollutant is also expected to contribute to cancer risk near major transportation corridors. Overall, the largest part of the total estimated cancer risk in these two wards comes from benzene and chromium (VI). Benzene emissions arise mainly from transportation sources. Most of the chromium comes from sources outside Toronto and thus is a health risk that is likely common to all parts of the city. 3) Cumulative risk from common air contaminants (CACs) Common air contaminants (CACs) are a group of air pollutants that are mainly associated with respiratory and cardiovascular health outcomes. There is no established threshold for the health effects from these five CACs, which are carbon monoxide, ozone, nitrogen dioxide, particulate matter (PM) and sulphur dioxide. Therefore, TPH used an approach similar to the one used for carcinogens to estimate the cumulative risk from this group of pollutants. Instead of excess cancer risk, TPH used the estimate of excess risk of premature death to calculate the cumulative impact. Premature death was selected because it is the most severe health outcome associated with exposure to CACs. Findings of elevated risk of premature death from cardiovascular and respiratory disease suggest that risk of other outcomes related to heart and lung conditions will also be elevated.

87 Air Quality Health Assessment Wards 5 and 6 v The cumulative excess risk of premature mortality attributable to the CACs is 10.2 per cent. That is, together they increase the existing mortality rate for respiratory and cardiovascular diseases by this amount. Fine particulate matter (PM 2.5 ) and nitrogen dioxide (NO 2 ) are the pollutants that contribute most to this risk. Similar to the analysis for carcinogens, there is higher risk from exposure to CACs in the area close to the local highways including Highway 427. There are also several industrial sources in the area that appear to emit CACs. Changes have taken place in the industrial facilities in the study area since the emissions data was compiled for this study. The most concentrated area of elevated risk located North of the Queensway and East of Kipling is attributable to O-I Canada Corporation, a facility that has now ceased operations. The emissions arising from the other two areas of concentrated risk in the Southwest and South-centre parts of the map are attributable to particulate matter and nitrogen oxides from Chrysler Canada Inc. and National Silicates. As described in the main body of the report, these facilities have reduced their emissions by between 75% and 80% since the data was compiled for this study. How does this compare with the previous local air quality study? The findings from the two local air quality studies that have been completed to date (South Riverdale, Leslieville and the Beach, and Etobicoke-Lakeshore) show similar patterns. In both, the same five substances exceed ambient air quality criteria or standards. Table 1 shows that the calculated cumulative health risks are of similar magnitude in both areas. Given that some uncertainty is inherent in any modelling exercise, it is reasonable to conclude that there may not be much difference in average risk for these two areas. In the case of carcinogens, benzene, chromium, and 1,3- butadiene are all among the top contributors to health risk in both areas. Among the CACs, PM 2.5 and NO 2 are the primary contributors to excess risk in both neighbourhoods. As well, transportation is an important source of local pollution and related health risk in both areas. Table 1: comparing health risks from air pollution in two Toronto neighbourhoods Type of Health outcome South Riverdale and Beach Etobicoke-Lakeshore Noncancer (eg.,immunological, No risk No risk neurological, developmental) Cancer 83 in one million 44 in one million Respiratory and Cardiovascular 8.9% increase 10.2% increase ǂ This value is an increase over what was previously reported in the staff report on this issue. Since that report was completed, a review of the information resulted in a slight increase in the estimate of impact from CACs in Wards 5 and 6. The cancer risk results are not directly comparable between neighbourhoods, as the findings from South Riverdale and the Beach include the contribution of Benzo[a]pyrene, while those from Etobicoke- Lakeshore do not. TPH is reviewing the B[a]P findings from both studies.

88 Air Quality Health Assessment Wards 5 and 6 vi Conclusions This health assessment suggests that many of the thirty air contaminants selected for this study, mainly the non-carcinogenic ones, occur below levels of concern to health in Wards 5 and 6 even when the combined exposure is taken into account. However, there is an indication that some carcinogens are present at levels above the one in one million excess cancer risk benchmark. Other pollutants such as ozone, nitrogen dioxides, and particulate matter are also found at levels that are known to have an adverse impact on health. For many substances of greatest concern, both among those related to cancer, and those related to cardiovascular and lung disease, the locally generated emissions are mainly from transportation sources. Therefore, it is important to continue efforts to reduce air pollution from both on and off-road transportation sources. The study and modelling used to estimate cumulative health effects have several limitations. It is difficult to compare the multiple health impacts into a single measure of health risk for the community. As the modelling is based on one year, 2006, the lifetime risk of diseases such as cancer are being estimated based on the air quality situation from one year. This assessment cannot account for past exposures from sources in the community that may contribute to current and future health problems. This study is a way of assessing cumulative health risks from multiple pollutants for a specific neighbourhood within a large urban area. Newly available data available about emissions from the small commercial and industrial sources of air pollutants in the area of study will improve the air quality model. This data is being collected through the Environmental Reporting and Disclosure Bylaw (ChemTRAC program) and will help improve future estimate of the cumulative exposure in these and other Toronto neighbourhoods. The results aid in setting priorities and determining effective strategies for pollution prevention to reduce exposures and improve the health of Toronto residents.

89 Air Quality Health Assessment Wards 5 and 6 1 Background The Etobicoke-Lakeshore local air quality study is the second in a series of studies that are expected to be carried out across the City of Toronto. The idea of characterizing health risk from air pollution at a neighbourhood scale was initially championed by residents of the Toronto neighbourhoods of South Riverdale and the Beaches. These communities have long been concerned about the health and environmental impacts of pollutants in their community. Like the Etobicoke-Lakeshore area, the community was historically home to a heavy industrial presence. While many of the facilities in the area closed down by the end of the 1990s, the potential for cumulative impacts from current and past exposures to pollutants remains a question for area residents. In response to these concerns, the Environment and Energy Division (EED) (formerly the Toronto Environment Office) completed a sophisticated air quality modelling study to estimate and map concentrations of thirty substances for Wards 30 and 32 which include the neighbourhoods of South Riverdale and The Beaches. While the South Riverdale and beaches neighbourhoods were the focus of the first local air quality study within Toronto, the Environment and Energy Division and Toronto Public Health recognized that air quality is a pressing concern for many communities. As a result, both the modelling and health assessments were designed to enable and simplify replication in other neighbourhoods across the City. In 2012, Wards 5 and 6 were selected for the second local air quality study within Toronto. The air quality modelling was designed to characterize how much of Toronto's air pollution comes from different sources such as industrial, residential/commercial, transportation, and biogenic sources. The modelling also estimates the proportion of pollution that comes from the United States, other parts of Ontario, and from within Toronto. By combining the contributions from each source type to local levels of air pollution, ambient concentrations can be calculated and mapped at high resolution to show how levels of each substance vary across the neighbourhood. Toronto Public Health (TPH) used the estimates from the modelling to conduct a health assessment of the findings to respond to the community's interest in understanding cumulative impacts from multiple pollutants. A cumulative health assessment approach can help to answer important questions related to the potential health effects of neighbourhood-level air pollution, including: Is air pollution present in the neighbourhood at levels that are a health concern? What are the patterns of exposure to air pollution? Which sources contribute most to potential health impacts? Which air pollutants should be priorities for reduction? Ultimately, the health assessment should facilitate identification of problem chemicals and support development, implementation and measurement of risk reduction strategies.

90 Air Quality Health Assessment Wards 5 and 6 2 This document outlines the selection of substances included in the project, and describes the approach taken to assess the health implications of the modelled concentrations. While it provides some background information about air quality modelling, the main focus is on the methods and results from the health assessment. More detailed information about the air quality modelling methods and findings is available from the Environment and Energy Division ( M d60f89RCRD). Assessing Air Quality Urban air pollution is made up of a complex mix of hundreds of substances. The potential for health risk from any airborne pollutant depends on (i) people s exposure to the substance, and (ii) its toxicity. Exposure refers to the amount of contact people have with a pollutant, and for air pollution it is often approximated using ambient pollutant concentrations. Toxicity refers to the type and degree of potential harm posed by the pollutant. Exposures to two different substances at the same concentration can result in different types of health effects (for example, impact on the respiratory system or impact on the nervous system). Similarly, exposure to two different substances at the same concentrations could affect the same body system, but at different severity (for example, nasal irritation or extended inflammation of the respiratory system). Pollutants which are of most concern are typically those that are linked to severe health outcomes and those that are linked to health outcomes which affect a large number of people. The air pollution mix in cities includes common air contaminants such as particulate matter, nitrogen oxides, and ozone that are linked to cardiovascular and respiratory effects (Brunekreef and Holgate, 2002; Dockery et al., 1993; Pope et al., 2002; Samet 2000; Schwartz, 1993). The air pollution mixture also includes low levels of air toxics substances for which long-term chronic exposure is associated with development of cancer or other serious health effects, such as reproductive effects or birth defects (Caldwell, 1998; Cohen, 2000; Environment Canada, 2009a; US EPA, 2009). Examples of air toxics include benzene, tetrachloroethylene, and lead. Using 1999 data, on Toronto, the common air contaminants were estimated to contribute to approximately 1,700 premature deaths and 6,000 hospitalisations each year (Toronto Public Health, 2004). In children, these pollutants are also linked to illness such as acute bronchitis and asthma. They also contribute to lost work days and diminished quality of life, especially for vulnerable populations including people already suffering from chronic illness. While the burden of illness from air toxics is unknown, reviews of local data suggest that some air toxics are likely to be present in Toronto s air at levels that pose a risk to health (Toronto Public Health 2002, 2008). At-risk populations include children, seniors, and individuals with pre-existing disease (Schwartz 2004; Samet and Krewski 2007; Krewski and Rainham 2007).

91 Air Quality Health Assessment Wards 5 and 6 3 The concentration of air pollutants at any urban location varies from day-to-day, and from one place to the other. Factors that affect neighbourhood pollution levels include the types of local sources such as cars, homes, and businesses, as well as pollution coming from other locations. The weather can affect whether pollution stays in the air for a short or long time, and the time of day can also affect pollution levels, since some pollution is more likely to be emitted at certain times (for example, emissions from cars are greater during the rush hour than in the middle of the night). There are two tools that are commonly used to evaluate air quality: monitoring and modelling. In air quality monitoring, specialized equipment is used to measure actual concentrations of pollutants. Depending on the equipment, measurement may be continuous, or may occur at specified time intervals. In Toronto, measurements for some pollutants are taken all the time. For others, the levels are only measured once every six days. Some benefits of air monitoring are that it allows investigation of trends in air quality over time and it provides information about actual concentrations at a specific location. However, air monitoring equipment is expensive to purchase, and must be maintained regularly. As well, most air quality monitors measure air quality at only one location. In Toronto, there are four monitoring stations that measure the most common air pollutants. They do not provide information about air pollution concentrations at other locations. Finally, most of the time, air monitoring data does not allow to identify the place of origin of the air pollutants. Air quality modelling uses information about known sources, typical emissions rates, the weather, and geography to predict the concentrations of pollutants for a place of interest. The predictions of air quality models are estimated based on complex calculations done by computers. Air quality modelling results can create a continuous picture showing expected air quality everywhere in a community, and can provide estimates about what sources contribute to the levels found in a community: for example, how much is from cars, and how much is from industry. As well, models can be used to see what might happen to air quality if a new source is added to the community, or if an existing source is eliminated. Finally, modelling is less expensive than monitoring. However, modelling requires extensive amounts of detailed data about air pollution sources and weather patterns, and modelling predictions are only as good as the data that is used as input. If there are problems with the data, or data are not available, this can affect the model s accuracy. Air quality monitoring and modelling are very different from each other, but taken together, the information from these tools can be very powerful. Air quality monitors in an area being modelled can be used to check air quality model results. If the models are good at predicting what happened in the past and what is happening now at the monitor s location, they should be good at predicting what is happening at other places in the

92 Air Quality Health Assessment Wards 5 and 6 4 neighbourhood. Air quality models are becoming increasingly reliable and sophisticated, and they facilitate analyses that monitoring data cannot do. For example they can help to identify which sources are responsible for most of the pollution in a community. They also can predict what would happen to local air quality if air pollution emissions changed. Modelling Approach Air pollution in any neighbourhood comes from a variety of sources, some of which are local. Others may be very far away: emissions from parts of the United States are known to affect air quality in Southern Ontario. In order to ensure that the impact of transboundary sources was fully captured in the model, information about emissions from large areas of the U.S. and Ontario was included in the model. Each substance was modelled for 1048 points in the studied area (Figure 1), which created a tight grid of estimated concentrations. Concentrations were estimated for points every metres. With this tight grid, it is possible to see variations in pollution concentrations across the neighbourhood. The model was used to estimate the average annual concentration and the maximum 24-hour concentration of each substance at each of the points. Figure 1: Illustration of 1048 points evenly distributed across Wards 5 and 6. The concentration of each substance was modelled at each point. Because of resource limitations that prevent modelling of hundreds of substances, this project modelled concentrations for 30 substances. It was important to narrow down the selection to those of greatest health concern. When deciding on which pollutants to model, the population health needs of the City as a whole was considered.

93 Air Quality Health Assessment Wards 5 and 6 5 Selection of Air Contaminants to be modelled The selection of pollutants to be modelled considered the City as a whole and was completed before the first local air quality study was carried out, with a view to creating a list of priority pollutants that would be relevant for each neighbourhood study. This selection process included several considerations: First, in 2008, Toronto Public Health established a list of 25 substances that are a priority concern for Toronto which were included in the Environmental Reporting and Disclosure bylaw (Toronto Public Health, 2008). This list includes air toxics as well as several criteria air contaminants (CACs). Toronto Public Health had two approaches to identify substances of potential concern in Toronto air: Prioritize estimated emissions data using a health-based ranking scheme. Using estimates of emissions, TPH applied a ranking scheme known as Toxicity Equivalence Potential (TEP) (Hertwich et al., 2001). The method uses estimates of the amount of a substance released, the potential for human exposure, and the toxicity of the substance to calculate a relative risk score for each substance. Identifying chemicals in Toronto s air that exceed health-based benchmarks. Data on air quality were obtained from Environment Canada and the Ontario Ministry of the Environment. The air quality data were used to identify substances in Toronto s air at levels that may cause adverse health effects. This was determined by comparing levels measured in our air with reference levels from the California Environmental Protection Agency, the Ontario Ministry of the Environment, and the New Jersey Department of Environmental Protection. The resulting list of 25 substances underlies the Priority Air Contaminants included in the cumulative air modelling study. An added advantage of selecting these substances is that the Environmental Reporting and Disclosure bylaw now collects information on smaller sources of these substances in Toronto. This will allow more accurate modelling of emissions of these priority air pollutants in the Toronto in future. Second, Toronto Public Health has previously estimated the burden of illness arising from exposure to the common air contaminants (CACs) (Toronto Public Health, 2004, 2007). While PM 2.5 and NO x were already among the substances identified as a priority concern for Toronto, not all of the CACs that were include in previous burden of illness studies were on the list. Given the weight of evidence that CACs can cause adverse health effects, and the research indicating that current levels of these substances affect the health of Toronto residents, all CACs are included in the modelling study. This added the following substances to the list of priority air contaminants" in this study: 1. PM 10 (includes PM 2.5 ) 2. O 3 3. SO x 4. CO

94 Air Quality Health Assessment Wards 5 and 6 6 Third, the substances included in the 2005 Ashbridges Treatment Plant (ABTP) air emissions study were also considered for inclusion. Some of those substances were already included on the list of Priority Air Contaminants. Those that remained were not associated with a health risk in the original study, and were excluded from the list of Priority Air Contaminants as there was no evidence that the substances would be emitted either locally or from transboundary emissions sources. Finally, some minor adjustments were made to the list of "Priority Air Contaminants" before it was finalized: Due to uncertainties in the ability to accurately capture modelling results for total volatile organic compounds (VOCs), the substance toluene was added to the list. Toluene is frequently used as a marker for VOCs, and is the most commonly released VOC in Ontario; NO x and SO x were replaced with NO 2 and SO 2, as these substances are associated with defined health endpoints; Benzo[a]pyrene (B[a]P) was used as a marker for the mixture of substances known as polycyclic aromatic hydrocarbons (PAHs). B[a]P is believed to be among the most potent of the PAHs and has been used by the Ontario Ministry of Environment (MOE) to set the air quality standards for PAHs in Ontario. Similarly, the World Health Organization and the UK Expert Panel on Air Quality Standards (EPAQS) have considered B[a]P as a marker of the carcinogenic potency of the PAHs mixture, when recommending their respective guidelines for PAHs in outdoor air. This led to the following final list of Priority Air Contaminants (PACs) for this study: 1. Acetaldehyde 11. 1,2-Dichloroethane 21. PM Acrolein 12. Dichloromethane 22. Tetrachloroethylene 3. Benzene 13. Ethylene dibromide 23. Toluene 4. 1,3-Butadiene 14. Formaldehyde 24. Trichloroethylene 5. Cadmium 15. Lead 25. Vinyl Chloride 6. Carbon tetrachloride 16. Manganese 26. Carbon Monoxide (CO) 7. Chloroform 17. Mercury 27. PM Chloromethane Nickel compounds 28. Sulphur Dioxide 9. Chromium 19. Nitrogen Dioxide 29. VOC ,4-Dichlorobenzene 20. B[a]P (as marker for PAHs) Ozone 1 This compound was later removed from the Environmental Reporting and Disclosure Bylaw. 2 B[a]P is modelled as a marker for the mixture of PAHs (not as PAH-equivalents). 3 Both Natural and anthropogenic sources of VOCs were included in the modelling.

95 Air Quality Health Assessment Wards 5 and 6 7 Health Assessment Overview This health assessment is intended to characterize the health impacts of the 30 priority air contaminants, when present at the modelled concentrations in Wards 5 and 6. The modelled concentrations provide a spatial representation of ambient airborne levels of each substance, and provide spatial proxies for exposure. In reality, any individual resident's exposure to air pollution depends on their individual characteristics and behaviours. This may include how much time they typically spend outside, and where they go during the day for work, errands, and play. Despite these limitations, using ambient concentrations as a proxy for exposure is a well-established methodology in the risk assessment and epidemiological literature, and is viewed as providing a reasonable estimate of the magnitude of exposure likely to be experienced by a local resident. In order to determine whether a given exposure is associated with a health risk, it can be compared with benchmark concentrations which have been established by reputable health agencies. Such benchmarks identify exposures that are associated with particular levels of risk, or are deemed to represent a safe level of exposure. Some health benchmarks are developed for chronic exposure where exposures to low levels of a substance occur over long periods of time, perhaps a whole lifetime. Benchmarks can also be established for acute exposures, where exposure occurs briefly, but to relatively high concentrations. Such health benchmarks are substance-specific and are often developed based on information from animal research, occupational studies, or epidemiological studies that used ambient concentrations as a measure of exposure. In this study, Toronto Public Health assessed the cumulative, or combined, impacts of exposure to multiple substances at the same time. Even when most individual pollutants are below levels of concern, people are typically exposed to more than one pollutant at a time. A cumulative assessment combines information about all of the pollutants modelled into summary estimates of health risk posed by air pollution as a whole. As described above, the model estimated the average annual concentrations and maximum 24-hour average concentrations for each substance at each of 1048 points within Wards 5 and 6. The annual average concentrations predicted at each point best reflect chronic exposure levels. The maximum 24-hour average concentrations are more representative of potential acute exposures. In this study, the maximum 24-hour average concentrations also represent a type of worst-case scenario, because the value at each point (e.g., the most polluted day of the entire year at each location) was combined into a single dataset to represent the neighbourhood. Using maximum concentrations in this way is a health-protective approach, since it is unlikely that the maximum concentration for any substance would occur at all locations at the same time. It is also unlikely that the maximum concentration for all substances would occur simultaneously at any of the locations. Instead, the spatial

96 Air Quality Health Assessment Wards 5 and 6 8 profile of individual and total concentrations would be in constant flux. Therefore, when characterizing the risk associated with 24-hour maximum concentrations for individual substances, the risk should be viewed as the worst-case scenario for the community. While it is tempting to explore cumulative impacts by directly comparing or summing the modelled concentrations for each substance, this is not an appropriate way to estimate cumulative health impacts. Each substance may induce health effects at a different concentration, such that some could be harmful at relatively low exposures, while others may not pose a significant threat unless concentrations are much higher. These differences in toxicity must be accounted for. The following section outlines the approach TPH used to conduct a cumulative health assessment. Methods for the Health Assessment Methods for characterizing the health risk from the carcinogens and non-carcinogens were developed based on work conducted in Portland Oregon, and in California (Department of Environmental Quality, 2006; Morello-Frosch et al., 2000). Calculating Cumulative Cancer Risk for Individuals Cancer risks can be assessed using inhalation unit risk (IUR) values for each carcinogenic compound. The inhalation unit risk is the upper-bound excess lifetime cancer risk estimated to result from continuous exposure to an agent at a concentration of 1 µg/m 3 in air. Estimated cancer risks for each carcinogenic substance at each location can therefore be calculated using the formula R ij = C ij *IUR j Where R ij is the estimate of individual lifetime cancer risk from pollutant j at location i, C ij is the concentration of pollutant j at location i in µg/m 3, and IUR j is the inhalation unit risk for a 70-year lifetime, for pollutant j in (µg/m 3 ) -1. The cancer risks of different air toxics are assumed to be additive, and can be summed together at each location to estimate a total individual lifetime cancer risk for that location: Cumulative cancer risk i = Σ i R ij The calculated cumulative risk can then be compared to a benchmark to characterize the level of concern that may be associated with the cumulative risk. The definition of tolerable risk may vary by jurisdiction. Many jurisdictions, including the USEPA, use one in one million (10-6 ) as the maximum lifetime risk benchmark for carcinogen. Health Canada often uses as benchmark from one in one hundred thousand to one in one million. Typically, Toronto Public Health uses one in one million.

97 Air Quality Health Assessment Wards 5 and 6 9 A common health-protective approach is to assume that most cancer types develop according to a similar multi-stage biological mechanism. Under this assumption, it makes sense to add the potential risk from different substances (which may be linked to different types of cancer) to estimate a cumulative cancer risk arising from multiple substances. Specific Substances and Weights Since over two years have passed since the first local air quality study was completed, all health benchmark values (including those for carcinogens, non-carcinogens, and Criteria Air Contaminants) were reviewed in 2013 and updated if warranted. Table 2 lists the substances treated as carcinogens and provides inhalation unit risk values used for weighting. The inhalation unit risk values for most substances are drawn from the California Office of Environmental Health Hazard Assessment (Cal OEHHA) database (OEHHA 2009). This database includes values for almost every pollutant listed in Table 2, and is regularly updated. The Cal OEHHA method is respected and viewed as being health-protective. The OEHHA database was not used for four of the substances considered in the study. Alternate inhalation risk values were adopted for benzo[a]pyrene, 1,3-butadiene, chloromethane, and chromium VI for the following reasons.: OMOE recently adopted new ambient air quality criteria (AAQC) for benzo[a]pyrene, based on carcinogenicity. These AAQC are based on an inhalation unit risk from the World Health Organization. This value was used since it is more conservative and specifically intended as a as a surrogate for the total carcinogenicity of PAHs, whereas the OEHHA value is specific to B[a]P. OMOE recently released an updated annual AAQC for benzene of 0.45, based on carcinogenic effects. This AAQC was based on an inhalation unit risk value from Texas, which is more recent than the OEHHA value, and based on occupational data rather than animal data. The texas value was therefore used in this study. The OEHHA database does not include an inhalation unit risk value for chloromethane, so for this substance, an inhalation unit risk derived by the state of New Jersey was used (New Jersey Department of Environmental Protection 2008). OMOE recently released a new annual average AAQC for chromium VI in the PM10 size fraction. TPH adopted the inhalation unit risk used by OMOE, which is based on combining information from two occupational cohorts that are more recent and incorporate fewer uncertainties than the one used by OEHHA. Links to the sources for all of the values in Table 2 are provided in Appendix A. Air quality modelling was done for total chromium. However, the health risks arising from exposure to chromium VI and other forms of chromium are different. For example, while chromium VI is primarily associated with lung cancer, chromium III is associated

98 Air Quality Health Assessment Wards 5 and 6 10 with impaired lung function and irritation. Therefore, when estimating the risk from the predicted chromium concentrations, a weight should be applied to account for the proportion that is likely to be chromium VI. While the proportion of chromium VI and chromium III is likely to vary by emission source, measurements of chromium present in ambient air can be used to estimate the typical proportions that might be expected to reach the population. Previously, TPH selected 15% to be a health protective, reasonable estimate 5 for the proportion of chromium that is likely to be chromium VI. Table 2: Carcinogens Air Pollutant Cancer Inhalation Unit Risk (IUR) (μg/m 3 ) -1 Acetaldehyde 2.70 x 10-6 Benzene 2.90 x 10-5 Benzo[a]pyrene 8.70 x ,3-Butadiene 5.00 x 10-7 Cadmium 4.20 x 10-3 Carbon tetrachloride 4.20 x 10-5 Chloroform 5.30 x 10-6 Chloromethane 1.80 x 10-6 Chromium VI 2.43 x ,4-Dichlorobenzene 1.10 x ,2-Dichloroethane (note: also called ethylene dichloride) 2.10 x 10-5 Dichloromethane (note: also known as methylene chloride) 1.00 x 10-6 Ethylene dibromide (note: also known as EDB) 7.10 x 10-5 Formaldehyde 6.00 x 10-6 Lead 1.20 x 10-5 Nickel compounds 2.60 x 10-4 Tetrachloroethlyene (also known as perchloroethylene) 5.90 x 10-6 Trichloroethylene 2.00 x 10-6 Vinyl chloride 7.80 x 10-5 Calculating cumulative non-cancer hazard The hazard posed by air pollutants that exhibit non-cancer effects can be assessed using a reference concentration (RfC). The RfC is an estimate of a continuous inhalation exposure to the human population (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime. To assess non-cancer 5 CEPA 1994 states that 3-8 % of air concentrations of total chromium could be chromium VI (CEPA, 1994). Estimates from Marshall, Macklin and Monaghan show that 13% of the total chromium air emissions are chromium VI (Toronto Public Health, 2007). Studies conducted in concluded that 20% of the routinely monitored chromium in Southwestern Ontario was in the hexavalent form (Bell and Hipfner, 1997).

99 Air Quality Health Assessment Wards 5 and 6 11 risks, the hazard ratio (HR) for each pollutant is calculated at each location by dividing the modelled concentration by its reference concentration: HR ij = C ij /RfC j Where HR ij is the hazard ratio for pollutant j at location i, C ij is the concentration of pollutant j at location i in µg/m 3, and RfC j is the reference concentration for pollutant j in µg/m 3. An indicator of total non-cancer hazard can be calculated by summing together the hazard ratios for each non-carcinogen pollutant to derive a total hazard index: HI i = Σ i HR ij There are no universal values for tolerable hazard ratios. The value of a tolerable hazard ratio depends upon the jurisdiction using it. Many agencies, including Health Canada and the USEPA, assume that a hazard ratio of less than one means that the concentration is less than the benchmark and so is not expected to be a concern for health. Health Canada also considers hazard ratios of 0.2 or less as not of concern for health for a single exposure pathway or when exposure is compared to the total acceptable daily intake. This reflects the possibility that hazard may accumulate from exposure through multiple exposure pathways. As well, air standards and benchmarks typically already consider multiple pathways of exposure when they are set. Specific Substances and Weights Table 3 lists the substances treated as non-carcinogens in this study. RfCs can be developed for various averaging time periods. The values in Table 3 represent chronic values wherever possible. There are five substances where RfCs are based on 24-hour averaging times because RfCs were unavailable for longer averaging times (denoted with ** ). All others are based on annual averaging periods. The chronic reference exposure levels used were drawn mainly from Cal OEHHA s database and existing or proposed ambient air quality criteria set by the Ontario Ministry of the Environment (MOE) (OEHHA 2008; Ontario Ministry of the Environment 2008). Both databases include values for almost every pollutant on our list, and are regularly updated (i.e., new values were adopted for acrolein, manganese, and mercury by Cal OEHHA in 2008, and the MOE adopted new standards for chromium on June 2011). Where an MOE annual ambient air quality criterion value for a non-carcinogen endpoint was lower than a California reference concentration, the MOE value was adopted. Otherwise, California's values were used.

100 Air Quality Health Assessment Wards 5 and 6 12 Table 3: Non-carcinogens (based on annual averaging periods unless otherwise indicated.) Air Pollutants Non-cancer RfC 1 μg/m 3 Acetaldehyde 140 Acrolein 0.35 Benzene 60 1,3-Butadiene 20 Cadmium Carbon tetrachloride** 2.4 Chloroform 300 Chloromethane** 320 Chromium III ** 0.5 Chromium VI 0.2 1,4-Dichlorobenzene** 95 1,2-Dichloroethane (note: also called ethylene dichloride) 400 Dichloromethane (note: also known as methylene chloride) 400 Ethylene dibromide (note: also known as EDB) 0.8 Formaldehyde 9 Lead *** 0.09 Manganese 0.09 Mercury compounds 0.03 Nickel compounds Tetrachloroethlyene (note: also known as perchloroethylene) 35 Trichloroethylene 600 Toluene 300 ** indicates that the RfC is based on a 24-hour averaging time. *** indicates that the RfC is based on a 30-day averaging time. Links to the sources for all of the values in Table 3 are provided in Appendix A. Some substances are classified as both carcinogens and non-carcinogens. These substances were included in the estimate of cumulative cancer risk as well as the hazard index calculation. Calculating Cumulative Risk from Common Air Contaminants (CACs) Common air contaminants (CACs) are associated with multiple respiratory and cardiovascular outcomes. The risk from CACs was evaluated for an endpoint which is common to all CACs and for which rigorous risk coefficients exist: premature mortality. Using acute premature mortality may be akin to selecting a single most significant endpoint: it is the most severe outcome, and enables the risks associated with each individual CAC to be compared to the others. However, it should be recognized that CACs are associated with a significant burden of illness from respiratory and cardiovascular health conditions in Toronto.

101 Air Quality Health Assessment Wards 5 and 6 13 The outcomes associated with CAC exposure are common, and would occur in the population even in the absence of CAC exposure. Thus, to characterize the risk posed by CACs, it is best to assess the additional or excess risk posed above baseline levels. The excess risk of premature mortality due to CAC exposure can be estimated based on the set of concentration response function (CRF) coefficients endorsed by Health Canada for use in its Air Quality Benefits Assessment Tool (AQBAT). These CRF coefficients represent statistically derived estimates of the percent (%) excess health endpoint associated with a unit increase in the pollutant concentration (Health Canada, 2006). Estimated percent excess per capita risk for each CAC at each location can be calculated using the formula 6 : R ijk ( CijCRF ) = ( e 1) * 100 Where R ijk is the estimate of percent excess per capita risk for a one unit increase in pollutant j at location i for outcome k, C ij is the concentration of pollutant j at location i in µg/m 3, and CRF ijk is the coefficient representing percent excess per capita risk for outcome k associated with a unit increase in pollutant j (in applicable units). Overall, the approach is analogous to the approach used for calculating cumulative risk from carcinogens. The percent excess per capita risks from four of the individual CACs (NO 2, O 3, CO, and SO 2 ) are assumed to be additive, and can be summed together at each location to estimate a total percent excess individual lifetime risk for that location: Cumulative CAC risk ik = Σ i R ijk As Table 4 suggests, the estimates for premature mortality for PM 2.5 are based on chronic exposure, whereas those for the remaining CACs (NO 2, O 3, CO, and SO 2 ) are for acute exposure. They are added together to derive a cumulative percent excess per capita risk under the assumption that over the long-term, the acute risk posed by PM 2.5 each day reaches a steady-state, and can be adequately represented as an annual risk. The approach described above is consistent with methods previously used by TPH to calculate the burden of illness from CACs (Toronto Public Health 2004, 2007). The above calculations generate percent excess per capita risk values, while the burden of illness reports applied percent excess per capita risks to current population incidence to estimate the number of people affected. For a small neighbourhood such as South Riverdale, the health outcome data is not reliable enough to enable a full burden of illness calculation. 6 The formula requires specifying a health outcome. Here, we specify acute premature mortality for NO 2, O 3, CO, and SO 2, and chronic premature mortality for PM 2.5. Premature mortality was assessed in epidemiological research that related air pollution levels with the number of people dying from nontraumatic causes. For premature mortality to be linked to air pollution exposure, the number of deaths must be higher than expected after exposure to elevated levels of air pollution, either in the long-term (for chronic exposures) or short-term (acute exposures).

102 Air Quality Health Assessment Wards 5 and 6 14 Specific Substances and Weights The concentration response function coefficients for the CACs are regression coefficients drawn from Health Canada s Air Quality Benefits Assessment Tool (AQBAT) (Health Canada, 2006). The substances treated as CACs are shown in Table 4. The CRF values were obtained directly from Health Canada. Table 4: Concentration Response Function coefficients for the CACs studied. Air Pollutants (CRF units) CRF coefficient (concentration) -1 (acute premature mortality) 7 CRF coefficient (concentration) -1 (chronic premature mortality) NO 2 (ppb -1 ) 7.48 x 10-4 PM 2.5 (μg/m 3 ) x 10-3 Ozone (ppb -1 ) 8.39 x 10-4 CO (ppm -1 ) 1.90 x 10-3 SO 2 (ppb -1 ) 4.59 x 10-4 PM 10 and total VOCs were not included in this analysis. This is to prevent doublecounting when estimating cumulative risk. PM 10 includes PM 2.5, and there is general consensus now that of the two measures for PM, PM 2.5 is the best indicator of risk and the best target for policy interventions (COMEAP, 2009). Several of the individual substances modelled including benzene, 1,3-butadiene, and formaldehyde qualify as VOCs, so including total VOCs would double-count these substances. Additionally, there is no health benchmark available for total VOCs. Such a benchmark would be difficult to identify because the toxicity of any VOC mixture depends on the specific combination of VOCs under consideration. Results: Air Quality Model Findings Some key findings from the air quality modelling are presented below to aid in interpreting the health assessment findings. A comprehensive presentation of the methods and findings of the air quality modelling component of this research is available in the technical materials made available by EED at d60f89RCRD. For most substances, the predicted ambient concentrations met Ontario s ambient air quality criteria (AAQCs). The AAQCs are provincial standards that are developed for a 7 While the CRFs for NO 2, CO, PM 2.5, and SO 2 are based on 24-hour averaging times, the CRF for O 3 is based on a 1-hour averaging time

103 Air Quality Health Assessment Wards 5 and 6 15 large number of substances to protect human health and the environment. The province has established 24-hour standards for all of the priority air contaminants except PM 2.5 and VOCs. For PM2.5, the ambient concentrations were compared to a Canada-Wide Standard, which serves a similar function to the AAQCs but is developed at the Federal level. The model predicted that these standards would be exceeded for NOx (24-hour average and 24-hour max), PM10, and benzene at some points and times. Concentrations of benzo[a]pyrene may also be above the provincial AAQC. These findings require further review as the predictions for this substance did not verify well with some limited available monitoring data. The air quality modelling enabled identification of the proportion of ambient concentrations of each substance attributable to sources outside of Toronto. Of the emissions that originate in Toronto, it is possible to identify the proportion originating from each of five sectors. For each of the substances, Table 5 shows the proportion originating from the U.S., from Ontario, and from within Toronto. The Toronto sources are further broken down by sector, where "industrial" refers to large industrial polluters listed in the national pollutant release inventory (NPRI), "residential/commercial" refers to homes, autobody shops, solvent users, and drycleaners, "mobile" refers to all transportation sources, "mobile non-road" refers mainly to airport, marine, rail, lawn and garden equipment, and "biogenic" refers to emissions arising from living organisms or biological processes. The model estimates that on average, 39% of the air pollution (by weight) comes from the United States and 25% comes from other parts of Ontario. Of the 36% of air pollution that comes from Toronto sources, residential and commercial sources appear to be the most important, contributing about 18% to the total ambient concentrations while mobile sources contribute about 13%. Off-road mobile sources contribute a further 4%, and large industrial sources in Toronto also contribute about 4% to the total concentrations in Toronto. The proportion of each individual pollutant that comes from inside or outside Toronto, or from a specific sector varies strongly by substance. For example, while the model estimates that 100% of carbon tetrachloride originates in the United States, 83% of 1,4- dichlorobenzene is generated within Toronto. For locally generated pollutants, residential and commercial sources in Toronto appear to be the primary contributors to concentrations of aromatic and halogenated substances, mainly due to emissions from dry cleaners, solvent users and auto-body shops. Residential and commercial sources also release trace metals which come from natural gas consumption and other combustion sources.

104 Air Quality Health Assessment Wards 5 and 6 16 Table 5: Contribution of Source Categories to Annual Average Predicted Air Concentrations (2006) in Toronto (Toronto Environment Office 2011). Substance Contribution from U.S Contribution from rest of Ontario Industrial Contribution from within Toronto Residential/ Commercial Mobile Nonroad Mobile Biogenic / Agri Nitrogen Oxides (1) 22% 21% 5.2% 11.3% 32.6% 7.9% 0% Carbon Monoxide 22% 19% 1.4% 2.8% 44.3% 10.5% 0% Sulphur Dioxide 71% 17% 0.64% 0.83% 2.4% 8.1% 0% PM2.5 32% 20% 10.9% 16.0% 16.0% 5.1% 0% PM10 30% 20% 5.1% 6.4% 36.3% 2.2% 0% VOCs 19% 25% 24.4% 14.8% 12.6% 4.0% 0.24% Formaldehyde 34% 21% 1.2% 2.1% 22.3% 19.3% 0% Acetaldehyde 40% 16% 4.8% 0% 21.4% 16.8% 0% Acrolein 14% 36% 1.2% 5.3% 26.0% 17.5% 0% 1,3-Butadiene 22% 24% 0% 0% 49.8% 4.2% 0% Benzene 26% 19% 8.7% 0% 39.0% 8.3% 0% Toluene 23% 15% 6.2% 35.9% 17.3% 2.7% 0% 1,4-Dichlorobenzene 2% 15% 0% 84.0% 0% 0% 0% PAHs (as B[a]Ps) 68% 8% 0.07% 0% 23.9% 0% 0% Chloromethane 4% 96% Dichloromethane 14% 15% 4.4% 67.6% 0% 0% 0% Chloroform 44% 8% 0.0% 48.0% 0% 0% 0% Carbon Tetrachloride 100% 0% 0% 0% 0% 0% 0% Vinyl Chloride 92% 8% ,2-Dichloroethane 42% 50% 0% 8.0% 0% 0% 0% Trichloroethylene 12% 85% 0.12% 2.9% 0% 0% 0% Tetrachloroethylene 7% 18% 0% 75.0% 0% 0% 0% Ethylene Dibromide 100% 0% Lead 57% 32% 7.2% 2.2% 0% 1.6% 0.0% Cadmium 26% 16% 11.4% 46.3% 0% 1.4% 0.0% Chromium 49% 27% 10.8% 7.8% 1.0% 3.4% 0.0% Nickel compounds 38% 49% 2.5% 9.8% 0.63% 0% 0% Mercury 61% 22% 3.4% 13.6% 0% 0% 0% Manganese 71% 26% 1.6% 1.0% 0.14% 0.35% 0.0% Average 39% 25% 4% 18% 13% 4% 0%

105 Air Quality Health Assessment Wards 5 and 6 17 When the predicted concentrations were mapped, substances exhibited significant spatial variation. This typically occurred if the substance is emitted mostly by a large industrial facility, where the concentrations are highest close to the facility. Similarly, transportation-related pollutants appear to be most concentrated along major highways and roadways. Within Wards 5 and 6, concentrations of transportation-related pollutants are highest close to Highway 427 and the Gardiner Expressway (see for example, Figure 2). The results of the model were validated where possible by comparing predicted concentrations with monitored data. The comparison was made with data collected at the closest air quality monitor. Monitoring data was available for 11 substances. For most, the match between monitored and modelled concentrations was very close, giving an average monitoring to modelling ratio of However, the modelled concentrations for nitrogen dioxide and sulphur dioxide were higher than the monitored concentrations by more than a factor of 2. This likely occurs because the modelling methodology used is unable to account for the chemical transformations that degrade these substances as they travel long distances from the U.S or parts of Ontario. However, it is important to note that high concentrations of nitrogen dioxide have been recognized as problem for the air quality of Toronto. For dichloromethane, the modelled concentrations were lower than the monitored by more than a factor of 2. This needs to be taken into account when the results of the health assessment are interpreted. Further validation is also required for the predicted concentrations of benzo[a]prene. Because of limited availability of monitoring data for this substance, it was not possible to do a full validation against monitored data. However, comparison with existing monitoring data suggests that additional review of the predicted concentrations for benzo[a]pyrene is warranted.

106 Air Quality Health Assessment Wards 5 and 6 18 Figure 2: Spatial distribution of predicted benzene concentrations in Wards 5 and 5. Transportation sources are a key contributor to ambient concentrations of this substance.

107 Air Quality Health Assessment Wards 5 and 6 19 Carcinogens The predicted risk associated with the nineteen carcinogens is summarized in Table 6. Predicted annual average concentrations were used to estimate risk as they are the best representation of chronic exposure, and the cancer risks are estimated based on lifetime exposure. TPH typically uses a risk level of 10-6 or "one in a million" to represent a situation without appreciable risk. In Table 6, values that exceed a risk level of 10-6 are in bold. The range from minimum to maximum represents the variation in risk that occurs within the neighbourhood. The minimum values represent the risk at the location in the neighbourhood where the lowest yearly average concentration is predicted, and the maximum values indicate the risk at the location of the highest predicted concentration in the neighbourhood. Table 6: Summary of Cancer risks for exposures at average annual levels (Values greater than 10-6 are shown in bold) SUBSTANCE Cancer risk (average annual) minimum mean maximum Acetaldehyde 5.2x x x10-06 Benzene 1.6x x x ,3-Butadiene 3.2x x x10-06 Cadmium 4.4x x x10-09 Carbon tetrachloride 5.0x x x10-07 Chloroform 3.0x x x10-08 Chloromethane 2.8x x x10-09 Chromium VI 5.2x x x ,4-Dichlorobenzene 8.7x x x ,2-Dichloroethane 6.5x x x10-09 Dichloromethane 1.5x x x10-07 Ethylene dibromide 1.2x x x10-09 Formaldehyde 2.5x x x10-06 Lead 2.4x x x10-08 Nickel compounds 4.1x x x10-07 Polyaromatic hydrocarbons* findings require further validation Tetrachloroethylene 1.7x x x10-05 Trichloroethylene 1.5x x x10-07 Vinyl Chloride 6.8x x x10-08 * As benzo[a]pyrene Figure 3 is a graphical representation of the data from Table 6. This figure plots the data shown in and shows that there are six substances where the average neighbourhood risks arising from the average annual concentrations were associated with greater than one in a million risk (benzene, chromium (VI), 1,3-butadiene, tetrachloroethylene, formaldehyde,

108 Air Quality Health Assessment Wards 5 and 6 20 and 1,4-dichlorobenzene). Also, acetaldehyde appears to be associated with a greater than a one in a million risk at locations within the neighbourhood where the concentrations are highest. Benzo[a]pyrene is excluded from this figure as the predictions require further validation. Figure 3 also shows that individual substances are associated with dramatically different risk (note that the y-axis is a logarithmic scale). Figure 3: Average, maximum, and minimum cancer risk values estimated for each substance based on average annual concentrations from the 1048 receptor sites. Carcinogens Cumulative Risk If the average annual risk is summed across eighteen carcinogenic substances (excluding benzo[a]pyrene), the mean cumulative risk is 4.4 x This estimated risk is above the one in a million benchmark. This total risk is less than two percent of the overall cancer incidence rate in Toronto which was about 400 per 100,000 in 2007 (Vital Statistics, 2007). As Table 7 below shows, these risks are driven mainly by a few substances: benzene, chromium VI, 1,3-butadiene, tetrachloroethylene, and formaldehyde. Benzo[a]pyrene was excluded from this analysis but would be expected to contribute to the overall cancer risk. As a result, the cumulative risk presented here may be an underestimate of the actual cumulative cancer risk in the neighbourhood.

109 Air Quality Health Assessment Wards 5 and 6 21 Table 7: Percentage Contribution of Individual Air Pollutants to the Cumulative Cancer Risk Estimates risks for exposures at average annual levels ( "---" indicates a contribution of < 0.1%). SUBSTANCE Cancer risk (mean for neighbourhood) Acetaldehyde 1.8% Benzene 52.0% 1,3-Butadiene 10.4% Cadmium --- Carbon tetrachloride 1.5% Chloroform 0.1% Chloromethane --- Chromium (VI) 13.6% 1,4-Dichlorobenzene 2.9% 1,2-Dichloroethane --- Dichloromethane 0.6% Ethylene dibromide --- Formaldehyde 7.1% Lead 0.1% Nickel compounds 1.0% Polyaromatic hydrocarbons findings require further validation Tetrachloroethylene 8.5% Trichloroethylene 0.3% Vinyl Chloride 0.2% * As benzo[a]pyrene A map of cumulative cancer risk is not presented here, as it is considered to be incomplete until the benzo[a]pyrene findings are validated. Substance-specific data indicate that two of the main contributors to cumulative risk, 1,3-butadiene and benzene, are primarily emitted near the two main highways in the area. Benzo[a]pyrene is also emitted primarily from transportation sources. Therefore, the areas of elevated cumulative cancer risk in Wards 5 and 6 are expected to be around the highways: Highway 427 and the Gardiner Expressway. Non-Carcinogens The predicted hazard associated with the 22 non-carcinogen substances is summarized in Table 8. For most of the non-carcinogens, it is most appropriate to use the average annual concentrations to predict hazard. This is because the health benchmarks selected represent chronic exposure periods or lifetime risk. However, there were four substances for which no chronic exposure health benchmark was available: chloromethane, carbon tetrachloride, chromium (III), and 1,4-dichlorobenzene. For lead, a reference value with a 30-day averaging time was used. These substances are identified in the table with the symbol "**". For these substances, the available evidence suggests that non-cancer health effects are more likely to arise as a result of acute exposures than

110 Air Quality Health Assessment Wards 5 and 6 22 chronic exposures. For these substances it is more appropriate to examine the hazards using the maximum 24-hour concentrations. Table 8: Summary of Noncancer HRs for exposures at average annual levels, and maximum daily levels. SUBSTANCE Noncancer HR (average annual) Noncancer HR (max 24-hr) min mean max min mean max Acetaldehyde 1.4 x x x Acrolein 6.6 x x x Benzene 9.4 x x x ,3-Butadiene 3.2 x x x Cadmium 2.4 x x x Carbon tetrachloride ** x x x Chloroform 1.9 x x x Chloromethane ** x x x Chromium (III) ** x x x Chromium (VI) 1.1 x x x ,4-Dichlorobenzene ** x x x ,2-Dichloroethane 7.7 x x x Dichloromethane 3.8 x x x Ethylene dibromide 2.0 x x x Formaldehyde 4.7 x x x Lead ** x x x Manganese 4.1 x x x Mercury 1.1 x x x Nickel compounds 1.4 x x x Tetrachloroethylene 8.2 x x x Toluene 1.1 x x x Trichloroethylene 1.2 x x x ** indicates that the hazard ratio was calculated based on an acute health outcome rather than a chronic health outcome. The range from minimum to maximum represents the variation in risk that occurs within the neighbourhood. The minimum values represent the risk at the location in the neighbourhood where the lowest concentration is predicted, and the maximum values indicate the risk at the location of the highest predicted concentration in the neighbourhood. Health Canada uses a hazard ratio of 0.2 as an indicator for potentially high exposure for a single exposure pathway and individual substance. All values for the hazard ratios, including at the points of maximum exposures, are below 0.2. The substance with the largest hazard ratio is nickel which a maximum hazard ratio of This indicates that exposures to these substances are well below their corresponding health benchmark (see Figure 4 and Table 8. Figure 4 is a graphical representation of the data in Table 8)

111 Air Quality Health Assessment Wards 5 and 6 23 Figure 4: Average, maximum and minimum Hazard Ratio (HR) values estimated for non-carcinogens substances based on average annual concentrations at the 1048 sites. Non-carcinogens Cumulative risk The sum of the average annual hazard indexes of all twenty-two non carcinogenic substances sums to As this cumulative hazard ratio is below 1, this suggests that exposure to the combination of substances would not result in adverse health impacts 8 Table 9 shows the contribution of each substance to the cumulative HR. Nickel is the main contributor (28.3%) followed by acrolein (19.9%). 8 There are limitations to this approach. It assumes that the impacts of exposure are additive. This can over estimate the hazard that substances act through different mechanisms, but at the same time, it does not takes into account potential synergistic effects.

112 Air Quality Health Assessment Wards 5 and 6 24 Table 9: Summary of Percentage Contribution of Individual Air Pollutants to the Cumulative Non-cancer Hazard Index estimates for exposures at average annual levels, and maximum daily levels. SUBSTANCE Non-cancer Hazard (mean for neighbourhood) Acetaldehyde 0.5% Acrolein 19.9% Benzene 3.2% 1,3-Butadiene 1.1% Cadmium 7.8% Carbon tetrachloride 0.1% Chloroform --- Chloromethane ---- Chromium (III) 2.9% Chromium (VI) 0.3% 1,4-Dichlorobenzene 0.1% 1,2-Dichloroethane --- Dichloromethane 0.1% Ethylene dibromide --- Formaldehyde 14.0% Lead 1.1% Manganese 10.2% Mercury 1.1% Nickel compounds 28.3% Tetrachloroethylene 4.4% Toluene 3.6% Trichloroethylene indicates a contribution of < 0.1% Figure 5 shows the distribution of the cumulative non-cancer risk across Wards 5 and 6. Although there is one easily identifiable source of non-carcinogens in the central southwest quadrant of the map, the levels of contamination in the middle of this area are still well below levels of concern for health.

113 Air Quality Health Assessment Wards 5 and 6 25 Figure 5: Spatial distribution of cumulative hazard from non-carcinogens for Wards 5 and 6.

114 Air Quality Health Assessment Wards 5 and 6 26 Criteria Air Contaminants The predicted percent excess per capita risk of premature death associated with the five criteria air contaminants is summarized in Table 10. Annual average values were used for estimating percent excess per capita risk, as they are most representative of chronic, longterm exposures. This is the approach previously used by Toronto Public Health when estimating burden of illness using these coefficients (Toronto Public Health, 2007). The range from minimum to maximum represents the variation in risk that occurs within the study area. The minimum values represent the risk at the location in the area where the lowest concentration is predicted, and the maximum values indicate the risk at the location of the highest predicted concentration in the area. Modelling for ozone was conducted separately from all the other substances. Ozone arises from chemical reactions of other pollutants and it is not possible to predict neighbourhood concentrations at a high spatial resolution. The model was only able to predict average concentrations. Table 10: Summary of CAC percent excess per capita risks for exposures at average annual levels 1. SUBSTANCE CAC risk (average annual) minimum mean maximum CO SO Ozone NO X (as NO 2 ) PM All are at levels of concern and therefore in bold. 2 Because ozone is produced primarily from chemical transformation rather than from emission and dispersion, it was modelled based on annual average NO x values. Only an annual average value could be produced to represent the neighbourhood concentration (no maximum or minimum values could be predicted). The distribution of predicted CAC percent excess per capita risk for average exposures across the neighbourhood for each substance is shown in Figure 6. The figure is a graphical representation of the data in Table 10, and shows that the excess risk is predominantly associated with PM 2.5 and NO 2. The cumulative excess per capita risk is estimated at 0.102, or 10.2%. Fine particulate matter (PM 2.5 ) and nitrogen dioxide are the pollutants that contribute most to this risk. This level of excess risk is similar to what has previously been calculated in the Burden of Illness in Toronto (Toronto Public Health, 2004).

115 Air Quality Health Assessment Wards 5 and 6 27 Figure 6: Average, maximum and minimum percent excess per capita risks of premature death for five criteria air contaminants based on average annual concentrations at the 1048 sites. Figure 7 shows the spatial distribution of cumulative percent excess per capita risk from CACs. There is higher risk from exposure to criteria air contaminants in the area close to the local highways including Highway 427. There are also several industrial sources in the area that appear to emit these common air pollutants. Changes have taken place in the industrial facilities in the study area since the study was conducted suggesting that local facilities have reduced their emissions of particulate matter and nitrogen oxides.

116 Air Quality Health Assessment Wards 5 and 6 28 Figure 7: Spatial distribution of cumulative percent excess per capita risk from CACs for Wards 5 and 6.

117 Air Quality Health Assessment Wards 5 and 6 29 Discussion Some of the thirty priority air contaminants selected for this study may be present in the neighbourhood at levels above recognized health benchmarks in some places in some occasions. The health assessment suggests that many substances that are emitted into Toronto's air are not likely to have an adverse impact on health in the study area. This is valuable information, enabling pollution prevention resources to be focussed on those substances and sources which are priorities from a health perspective. How does this compare with the previous local air quality study? The findings from the two local air quality studies that have been completed to date (South Riverdale, Leslieville and the Beach, and Etobicoke-Lakeshore) show similar patterns. In both, the same five substances exceed ambient air quality criteria or standards. Table 11 shows that the calculated cumulative health risks are of similar magnitude in both areas. Given that some uncertainty is inherent in any modelling exercise, it is reasonable to conclude that there may not be much difference in average risk for these two areas. In the case of carcinogens, benzene, chromium, and 1,3- butadiene are all among the top contributors to health risk in both areas. Among the CACs, PM 2.5 and NO 2 are the primary contributors to excess risk in both neighbourhoods. As well, transportation is an important source of local pollution and related health risk in both areas. Table 11: comparing health risks from air pollution in two Toronto neighbourhoods Type of Health outcome South Riverdale and Beach Etobicoke-Lakeshore Noncancer (eg.,immunological, No risk No risk neurological, developmental) Cancer 83 in one million 9 44 in one million* Respiratory and Cardiovascular 8.9% increase 10.2% increase As described earlier in the section on air quality modelling results, the air quality modelling enables identification of the proportion of ambient pollution attributable to specific source types. Table 12 identifies those substances that are at levels above health benchmarks and the proportion originating from the U.S., from Ontario, and from within Toronto. Toronto sources are further broken down by sector. These results are helpful in setting priorities and strategies for pollution prevention that could improve air quality in Wards 5 and 6. 9 The cancer risk results are not directly comparable between neighbourhoods, as the findings from South Riverdale and the Beach include the contribution of Benzo[a]pyrene, while those from Etobicoke- Lakeshore do not. TPH is reviewing the B[a]P findings from both studies.

118 Air Quality Health Assessment Wards 5 and 6 30 Table 12: Origins of substances identified as exceeding health benchmarks in Wards 5 and 6. Substance Contribution Contribution from Within Toronto Contribution from rest of Residential/ Mobile from U.S Industrial Mobile Ontario Commercial Non-road NO2 22% 21% 5.2% 11.3% 32.6% 7.9% CO 22% 19% 1.4% 2.8% 44.3% 10.5% SO2 71% 17% 0.64% 0.83% 2.4% 8.1% PM2.5 32% 20% 10.9% 16.0% 16.0% 5.1% Acetaldehyde 40% 16% 4.8% 0% 21.4% 16.8% 1,3-Butadiene 22% 24% 0% 0% 49.8% 4.2% Benzene 26% 19% 8.7% 0% 39.0% 8.3% Chromium 49% 27% 10.8% 7.8% 1.0% 3.4% 1,4-Dichlorobenzene 2% 15% 0% 84.0% 0% 0% Formaldehyde 34% 21% 1.2% 2.1% 22.3% 19.3% Tetrachloroethylene 7% 18% 0% 75.0% 0% 0% The results suggest that for some substances that affect these neighbourhoods, most emissions originate outside the City of Toronto. This is the case with SO 2, 88% of which comes from either the U.S. or Ontario, chromium, of which 76% originates outside Toronto, and PAHs, 76% of which originate outside Toronto. The findings for SO 2 in particular highlight the success of Canadian initiatives to reduce the presence of sulphur in fuels: very little of the substance is emitted in Canada now. The modelling highlights a need for regional strategies and multi-jurisdictional co-operation in addition to local strategies to effectively reduce the presence of these substances in Toronto. The results also indicate that for many substances, the transportation sector is the main source of the locally-generated emissions. This appears to be especially true for NO 2, CO, 1,3-butadiene, and benzene. These findings suggest that options to reduce exposure to transportation-related emissions are needed. Strategies for doing this could include new technologies for fuels and vehicle emissions reduction or programs that encourage people to choose transit or active transportation over driving when feasible. For some substances such as 1,4-dichlorobenzene and tetrachloroethylene, most local emissions fall under the residential/commercial category. These substances are not typically emitted from homes, and so are attributed to commercial sources such as solvent users and drycleaners respectively. A limitation of the modelling effort was that only three types of commercial emitters were included in the model: drycleaners, solvent users, and autobody shops. As described earlier in this report, in the future, information collected from the Environmental Reporting and Disclosure Bylaw will support more accurate modelling of local concentrations for priority air pollutants in Toronto ( In turn, the air modelling and health assessment results can inform the ChemTRAC program and

119 Air Quality Health Assessment Wards 5 and 6 31 help to identify which substances should remain priorities for reporting in the long term. As well, a component of the program is to support smaller businesses in finding ways to reduce emissions and adopt safer alternatives where possible. On average, large industrial sources are small contributors to overall air pollution and the cumulative risk arising from air pollution in those two Wards. While the study identified some areas of elevated pollution concentrations, these were reviewed because the industrial emissions data used in the model was from Examination of more recent data (2011) from NPRI and discussions with the Ontario Ministry of Environment and members of the Industrial Liaison Committee identified that the industries in the areas of elevated pollution concentrations that are identifiable in Figure 7 have either closed or reduced their emissions by between 75% and 80% since The recent air quality modelling study suggests that the main reasons for the reductions in industrial emissions are that industries have closed and/or moved away, industries have reduced their operations, and industries have improved their processes and reduced air emissions. This project assessed the air quality in Wards 5 and 6, and the conclusions about patterns of exposure and which substances are priorities for reduction apply specifically to this area of Toronto. Other neighbourhoods may be characterized by different source types which could influence the local air quality mix and create different priorities for emissions reduction. Limitations of the Health Assessment Methodology There are some challenges associated with conducting a cumulative assessment of health risk. For example, substances are thought to act through various mechanisms of action. Substances with similar mechanisms of action and affecting similar body systems may exert an additive effect on the health of a population. However, it is more difficult to sum up effect on different body systems, particularly as some effects may be deemed more severe than others. For example, the presence of common air contaminants may worsen symptoms among asthmatics in the community, while certain air toxics may increase the lifetime risk of cancer. It is difficult to compare or add these two health impacts into a single measure of health risk for the community. To tackle this difficulty TPH separated substances into three broad categories: those associated with cancer outcomes, those associated with non-cancer outcomes, and CACs, which are associated mainly with cardiovascular and respiratory outcomes, and premature death. Because cancers are thought to arise according to a given biological process, it is logical to accumulate the potential risk of cancer from individual substances. Similarly, because the risk estimates for CACs are focussed on a single health outcome, accumulating the risks across substances is a logical approach.

120 Air Quality Health Assessment Wards 5 and 6 32 Non-cancer health effects attributed to different substances affect different systems/target organs and each may cause harm as a result of different biological mechanism. Some would argue that it may therefore not be appropriate to sum the hazard quotients for the substance-specific non-cancer health endpoints into a single index. However, assessing the hazard separately for each target system/organ ignores the potential for lowered overall resilience from the cumulative impact of multiple environment assaults on all organs/systems at low concentrations. Finally, there is no guarantee that a person may not be subject to more than one health outcome at a time. If non-cancer hazard indices are calculated separately for each target system/organ, each cumulative (organ-specific) hazard index will be lower than if the hazard index is calculated for all substances as a group. Accumulating all hazard ratios is a way of characterizing the maximum possible non-cancer impact. If a hazard ratio of one is not exceeded from such an accumulation, then the organ-specific and substance-specific noncancer hazards can also be assumed to be less than one. In this study, the hazard ratio was 0.42 well below the threshold of It appears that there is no need to examine the hazards by body system or target organ, since the hazard ratios for individual target system/organs will all be less than the cumulative hazard, and will therefore all be less than Some pollutants may interact with each other synergistically. This means that their combined effect is greater than what would be expected based on assessing the substances in a simple additive way. Another challenge is that some people and not others may be especially vulnerable to certain pollutants. With no data on these possible interactions or range of sensitivities, these aspects could not be accounted for in the analysis. The approach to characterizing risk for CACs examines only one health endpoint: premature mortality. There are other health endpoints associated with exposure to CACs such as cardiovascular and respiratory morbidity and the risk coefficients (and associated incidence of outcomes) for these health outcomes are higher than those for acute mortality. Thus, the estimated impact of CACs must be viewed as an indicator of potential impact for the neighbourhood and not the full burden of illness. This approach also represents an aggregate level estimate which may not accurately represent the percent excess risk arising for specific age groups. The predicted concentrations of each substance are based on the year Thus, lifetime risk of diseases such as cancer are being estimated based on the air quality situation from one year. This assessment cannot account for past exposures from sources which no longer exist in the community. However, past exposures may contribute to current and future health problems.

121 Air Quality Health Assessment Wards 5 and 6 33 References Bell, R.W., and J C Hipfner Airborne hexavalent chromium in southwestern Ontario. Air Waste Manag Assoc. 47 (8): Canadian Environmental Protection Agency Priority Substances List Assessment report Chromium and its Compounds. Minister of Supply and Services Canada. Available from Cohen, Aaron J, H Ross Anderson, Bart Ostro, Kiran Dev Pandey, Michal Krzyzanowski, Nino Künzli, Kersten Gutschmidt, et al The global burden of disease due to outdoor air pollution. Journal of Toxicology and Environmental Health. Part A 68, no. 13 (July 9): doi: / Committee on the Medical Effects of Air Pollutants (COMEAP) Long-Term Exposure to Air Pollution: Effect on Mortality. 9report.pdf Department of Environmental Quality (DEQ) Portland Air Toxics Assessment. Portland, OR. Health Canada Air Quality Benefits Assessment Tool. Ottawa, ON. Hertwich EG, Mateles SF, Pease WS, McKone TE Human toxicity potentials for life cycle assessment and Toxics Release Inventory risk screening. Environmental Toxicology and Chemistry, 20, no. 4: Lura Consulting Access to Environmental Information: Environmental Reporting in Toronto. Gaps and Opportunities. Prepared for Toronto Public Health. Available from Martuzzi, M., M. Krzyzanowski, and R. Bertollini Health impact assessment of air pollution: providing further evidence for public health action. European Respiratory Journal 21, no. 40 (May 1): 86s -91s. doi: / Morello-Frosch, Rachel A., Tracey J. Woodruff, Daniel A. Axelrad, and Jane C. Caldwell Air Toxics and Health Risks in California: The Public Health Implications of Outdoor Concentrations. Risk Analysis 20, no. 2: doi: / New Jersey Department of Environmental Protection NJDEP - Toxicity Factors for Methyl Chloride (chloromethane). OEHHA OEHHA Acute, 8-hour and Chronic Reference Exposure Level (REL) Summary. OEHHA OEHHA Cancer Potency Values. Ontario Ministry of the Environment Ontario's Ambient Air Quality Criteria. Toronto Environment Office, An All Sources Cumulative Air Quality Impact Study of South Riverdale - Leslieville Beaches. Available from

122 Air Quality Health Assessment Wards 5 and Toronto Public Health Air Pollution Burden of Illness in Toronto. Available from Toronto Public Health Mortality, Cancer Incidence and Hospitalization Among Residents of the Neighbourhoods of South Riverdale, The Beaches, and Comparison Communities. Available from Toronto Public Health Air Pollution Burden of Illness from Traffic in Toronto Problems and Solutions. Available from Toronto Public Health Environmental Reporting and Disclosure. A Proposed Program for the City of Toronto. United States Environmental Protection Agency. U.S. Risk Maps 1999 National-Scale Air Toxics Assessment Technology Transfer Network Air Toxics Web site. Available from Vital Statistics 2007, Ontario Ministry of Health and Long-Term Care, IntelliHEALTH ONTARIO. Respiratory Disease (ICD10 Codes J00-J99): deaths = 1382; agestandardized mortality rate = 36.4 deaths per 100,000 (95% CI = to 38.4); crude rate = 52.5 deaths per 100,000. Cardiovascular Disease (ICD10 Codes I00- I99): deaths = 4646; age-standardized mortality rate = deaths per 100,000 (95% CI = to 128.9); crude rate = deaths per 100,000. Date Extracted: May 2011

123 Air Quality Health Assessment Wards 5 and 5 A-1 Appendix A

124 Air Quality Health Assessment Wards 5 and 5 A-2 Toronto Air Pollutants: reference levels summary Method for choosing Reference Concentrations (RFCs): For Air Toxics: Draw from Chronic Reference Exposure Levels (CRELs) which are based on annual exposure. Where a MOE AAQC based on non-carcinogen enpoint is lower than a CREL, adopt AAQC Priority is to select annual values where possible Method for choosing inhalation unit risk (IUR) values: Adopt CalEPA values which enable use of a complete list from a single source. Where a MOE AAQC based on carcinogen enpoint uses a lower IUR that CalEPA value and based on more recent data, adopt IUR used to develop AAQC Method for choosing risk coefficients (representing percent excess risk) Adopt AQBAT CRFs (Concentration response Functions) for all CACs Air Pollutants RFC = reference exposure concentration (eg., CREL) Reference Cancer Inhalation Unit Risk (IUR) Reference unit risk coefficient (premature mortality) Reference µ g/m 3 ( µ g/m 3 ) -1 units as noted Acetaldehyde E-06 2 N/A N/A Acrolein N/A N/A N/A N/A Benzene E-05 2 N/A N/A Benzo[a]Pyrene N/A N/A 8.70E-02 5 N/A N/A 1,3-Butadiene E-07 3 N/A N/A Cadmium E-03 2 N/A N/A Carbon tetrachloride E-05 2 N/A N/A Chloroform E-06 2 N/A N/A Chloromethane E-06 4 N/A N/A Chromium III N/A N/A N/A N/A Chromium VI E-02 3 N/A N/A 1,4-Dichlorobenzene E-05 2 N/A N/A 1,2-Dichloroethane (note: also called ethylene dichloride) E-05 2 N/A N/A Dichloromethane (note: also known as methylene chloride) E-06 2 N/A N/A Ethylene dibromide (note: also known as EDB or 1,2- Dibromoethane) E-05 2 N/A N/A Formaldehyde E-06 2 N/A N/A Lead E-05 2 N/A N/A Manganese N/A N/A N/A N/A Mercury compounds N/A N/A N/A N/A Nickel compounds E-04 2 N/A N/A Perchloroethylene (note: also kno E-06 2 N/A N/A Trichloroethylene E-06 2 N/A N/A Toluene N/A N/A N/A N/A Vinyl chloride N/A N/A 7.80E-05 2 N/A N/A total VOC do not include (eliminated since it will cause double-counting for the analysis) NO2 N/A N/A N/A N/A 7.48E-04 6 PM2.5 N/A N/A N/A N/A 6.76E-03 6 PM10 do not include (eliminated since it will cause double-counting for the analysis) ozone N/A N/A N/A N/A 8.39E-04 6 CO N/A N/A N/A N/A 1.90E-03 6 SO2 N/A N/A N/A N/A 4.59E-04 6 References 1 CalEPA CRELs (chronic) 2 CalEPA Cancer Inhalation Unit Risk 3 MOE AAQC: (see individual OMOE substance decision documents for IURs; links on the carcinogen tab) 4 State of New Jersey 5 World Health Organization (as cited in 6 AQBAT CRFs (obtained direct from Health Canada)

125 General AQ Comments: 1. Modelling of emissions the modelling of impacts from the facility should include estimating and modelling emissions from the additional traffic (due to the TTC facility) on the nearby roads. This cumulative impact assessment approach has previously been used by TPH on other projects, such as the BBTCA expansion project, and it allows for a more robust assessment of potential health impacts associated with the proposal. Response Industry practice is to only model emissions from the facility and assume that emissions offsite are included in the background concentrations. For this reason, very conservative concentrations were derived and used in the assessment in order to predict worst case impacts. Roadway assessment was undertaken as per TPH s recommendation. The contribution of the roadway is marginal, with the highest concentration as a percent of the standard being PM2.5 at 5.2%. There are a predicted two additional exceedances of both the PM10 and PM2.5 standards as a result of the roadway traffic over the 5 year period and no additional exceedances of benzene. This equates to less than 1% of the time. It should be reiterated that despite the minimal impact of the roadway, these predicted concentrations are likely over conservative. The chances that the maximum impacts from ambient, the garage, and the roadway would occur simultaneously is very small. 2. The use of health protective comparison benchmarks Novus currently compares the estimated AQ concentrations against the Ontario AAQCs. Although that is appropriate for evaluating regulatory compliance, it is a departure from the TPH approach of evaluating health impacts. We recommend you evaluate the projected AQ impacts against health based carcinogenic and non carcinogenic benchmarks, and the AQBAT risk coefficients to evaluate health risk from criteria air pollutants in addition to the Ontario AAQCs. In the past, this approach has been used by TPH for Local Air Quality (LAQ) studies, the BBTCA project, and is currently being utilized for other proposals. For your convenience we have attached a copy of the Local Air Quality Study for Wards 5 & 6, the approach is outlined in the body of the reports and the benchmarks are listed in Appendix A. Response Concentration response function (CRF) relationships developed by Health Canada for their AQBAT model to correlate increases in concentrations of criteria air contaminants to excess risk of premature mortality is a perfectly reasonable approach for evaluating changes in air quality over a large airshed area with a sufficient population base. However, the epidemiological basis to this approach rapidly breaks down for small areas with limited populations. There is no scientific

126 basis for it to be used for something like the incremental very local increases in CAC concentrations in the immediate vicinity of the proposed facility Health Canada did not intend these CRFs to be used in evaluating a small point source emission with very localized impacts. This would be highly inappropriate use of these epidemiologically derived functions. 3. Air monitoring Novus made a number of assumptions in order to justify the use of existing AQ data from monitoring stations located at a significant distance from the site. Although we recognize the methodology utilized by Novus was fairly conservative, the data may not represent the existing AQ conditions in the area. We recommend conducting air monitoring to validate the baseline data assumptions (prior to the commencement of any site/construction activities), during the construction phase (to ensure construction impacts are minimized similar approach as at the Leslie Barns project), and once the facility is built to ensure impacts have been properly assessed and mitigated. Response Considering that conservative assumptions were made in the modeling in order to predict worst case impacts and in general predicted results showed compliance with regulatory criteria, it is our opinion that air monitoring is not warranted in this situation. Considering the above, a monitoring program costing hundreds of thousands of dollars can t be justified from a technical perspective. Construction impacts are temporary in nature and are best managed utilizing a construction code of practice such as Environment Canada s Best Practices for the Reduction of Air Emissions From Construction and Demolition Activities, March Monitoring of construction activities is not necessary considering that a construction code of practice will be implemented and monitored for adherence. The EPR will meet monitoring requirements as outlined in the TPAP.

127 January 26 th, 2015 AECOM 4 th Floor, 30 Leek Crescent Richmond Hill, ON L4B 4N4 Re: TTC McNicoll Bus Garage FINAL REPORT Redlea/McNicoll Roadway Assessment Novus File No AECOM retained Novus Environmental Inc. (Novus) as part of an environmental assessment for the proposed TTC McNicoll Bus Garage to be located at the corner of McNicoll Avenue and a future extension of Redlea Avenue. This assessment evaluates the increased bus traffic and employee vehicle traffic on the roadways surrounding the facility. A detailed air quality assessment of all on-site emission sources was previously conducted ( TTC McNicoll Bus Garage TPAP Air Quality Assessment, dated December 3, 2014). The results of this assessment take into account the combined effect of background concentrations, facility on-site emissions and offsite vehicle emissions from buses and employee vehicles. 1.0 Introduction The project includes the construction of a new bus storage and maintenance facility for the Toronto Transit Commission (TTC). The proposed facility is located on McNicoll Avenue, just east of Kennedy Road in the City of Toronto. The new facility will be used to house buses when they are not in use, and for general maintenance and repair on the buses. Figure 1 shows the facility s location as well as the proposed extension of Redlea Avenue and the Employee Parking Lot at the facility. This assessment specifically considers the impact of the additional bus traffic on Redlea Avenue as well as the additional employee vehicle traffic. Contaminant concentrations have been predicted at sensitive receptors surrounding the facility consistent with the previous air quality assessment. Five years of meteorological conditions and ambient contaminant concentrations have been considered in this assessment. Air Quality Sound & Vibration Sustainable Water Wind & Climate Novus Environmental Inc. 150 Research Lane, Suite 105, Guelph, Ontario, Canada N1G 4T2 Novus West Inc Avenue SW, Suite 600, Calgary, Alberta, Canada T2R 1K7

128 TTC McNicoll Bus Garage Roadway Assessment Final Report January, 2015 Figure 1: Project Site and Study Area 2.0 Assessment Approach The roadway assessment has been conducted following typical methodologies for roadway projects in Ontario. The methodology followed is summarized below: Roadway alignment, volume, and signal parameters were identified for the project modelling year (2011). Motor vehicle emission rates were developed using the US EPA s MOVES software, which is based on an analysis of millions of emission test results and considerable advances in the Agency s understanding of vehicle emissions and accounts for changes in emissions due to proposed standards and regulations. Nearby receptors were identified, including any: o Residence; o Place of worship; o School or Daycare; and o Health Care Facility. The roadway traffic due to the project (i.e. buses and employee vehicles) was modelled using the US EPA s CAL3QHCR model which predicts contaminant air dispersion from a roadway. The model was run using 5-year of local meteorology ( ) to ensure various meteorological conditions are considered. Worst-case predicted concentrations of the Contaminants of Concern (CAC s) at the nearest sensitive receptor were added to the worst-case predicted values from the Novus Environmental 2

129 TTC McNicoll Bus Garage Roadway Assessment Final Report January, 2015 facility assessment (completed by Novus dated December 3, 2014) as well as the worstcase background contaminant concentrations. The worst-case predicted concentrations of each contaminant were compared against the relevant provincial standards to determine compliance. It should be noted that this approach is extremely conservative as it combines the worst-case concentration from three different independent sources (i.e. the facility, the roadway, and ambient contaminant concentrations) over a 5-year period. It is very unlikely that these three maximum values would occur simultaneously. 3.0 Local Meteorology Local meteorological conditions are used in the CAL3QHCR model to predict the dispersion of contaminants from the roadway. Surface measurements were taken from Pearson International Airport while upper air data was taken from the closest measurement site, the Buffalo International Airport. Figure 2 presents a wind rose illustrating the meteorological conditions modelled for the 5-year period. Figure 2: Wind Frequency Diagram for Pearson International Airport Novus Environmental 3

130 TTC McNicoll Bus Garage Roadway Assessment Final Report January, Ambient Pollutant Concentrations The predicted contaminant concentrations from the project were added to ambient concentrations in order to evaluate a worst-case scenario. The CAC s which were evaluated are shown in Table 1 along with their respective standards. Table 1: Applicable Contaminant Guidelines Pollutant Averaging Period Guideline (µg/m 3 ) Source NO 2 1 hr 400 AAQC 24 hr 200 AAQC CO 1 hr 36,200 AAQC 8 hr 15,700 AAQC PM hr 27 * AAQC (CWS) PM hr 50 Interim AAQC Acetaldehyde 24 hr 500 AAQC Acrolein 1 hr hr 0.4 Environmental Registry Benzene 24 hr 2.3 Environmental Registry 1,3-Butadiene 24 hr 10 Environmental Registry Formaldehyde 24 hr 65 AAQC * The CWS is based on the annual 98 th percentile concentration, averaged over three consecutive years. The standard becomes 27 in year Each of the CAC s concentrations from the nearest four MOECC and six NAPS monitoring stations were compared to determine the worst-case ambient background concentrations. Determination of the worst-case monitoring station for each contaminant and averaging period was performed in the detailed Air Quality Assessment. Worst-case background concentrations used in this assessment are consistent with those considered in the Air Quality Assessment. Figure 3 shows the ambient concentrations of each contaminant from the worst-case monitoring stations as a percentage of the respective guideline. Novus Environmental 4

131 TTC McNicoll Bus Garage Roadway Assessment Final Report January, 2015 Figure 3: Summary of Background Conditions For the contaminants which exceed their respective guidelines (i.e. PM2.5 and PM10), the frequency of exceedance was considered. Both PM2.5 and PM10 exceeded their respective standards 12 times over the 5-year period. It should be noted that PM10 is not measured in Ontario and is calculated form measured PM2.5 concentrations. Additionally it should be noted that PM2.5 is regulated based on the 3-year 98 th percentile value, which was not exceeded. 5.0 Previous Assessment The operations within the bus garage facility and the on-site bus movements have previously been assessed by Novus. This assessment considered emissions from bus movements on-site, buses idling on site prior to going into operation and during maintenance activities, comfort heating equipment, the paint booth and shop areas, liquid storage tanks and the employee parking lot. Table 2 shows the predicted concentration of the CAC s, the combined contaminant concentrations and the predicted additional exceedances of the guideline due to emissions from the facility. Novus Environmental 5

132 TTC McNicoll Bus Garage Roadway Assessment Final Report January, 2015 Table 2: Worst-Case Predicted Concentrations as a Percentage of the Guideline Contaminant Averaging Period Maximum Concentration Due to Facility Alone (µg/m 3 ) Maximum Concentration Due to Facility Alone (as % of Standard) Combined Concentration as % of Standard (Ambient + Project) Maximum 90 th Percentile Average Additional # of Guideline Exceedances due to Project Over 5 Years NO 2 1-hour % 79% 53% 47% 24-hour 32 16% 61% 40% 30% CO 1-hour % 6% 1% 1% 8-hour % 12% 3% 2% PM hour 1.6 6% 139% 53% 30% 3 PM hour 2 4% 137% 51% 29% 1 Acetaldehyde 24-hour % 1% <1% <1% Acrolein 24-hour % 38% 25% 20% Benzene 24-hour % 107% 50% 35% 6 1,3-Butadiene 24-hour % 3% 1% 1% Formaldehyde 24-hour % 13% 8% 5% 1 CWS guideline for PM2.5 is based on an average annual 98 th percentile concentration, averaged over 3 consecutive years. The maximum combined 3-year rolling average of the annual 98 th percentile concentration was 22.14, which is 82% of the guideline. 6.0 Assessment Methodology This assessment followed the methodologies established in the air quality assessment for the garage. Emission rates, bus and car volumes and receptor locations were all consistent with the previous report with the exception of vehicle emission rates at 60 km/hr which were modelled for this assessment (in comparison to lower speeds modelled on-site). Traffic signal parameters at the proposed Redlea Avenue and McNicoll Avenue intersection were assumed based on the traffic signal parameters at the existing intersections to the east and west of the proposed intersection (Kennedy Road and Silver Star Boulevard). 6.1 Traffic Volumes Hourly bus counts entering and leaving the existing Mount Dennis facility (which is similar to the proposed facility) were provided by AECOM, as well as a maximum vehicle count of 220 buses at the proposed McNicoll Bus Garage, which represents the number of buses used for daily service. The hourly vehicle distribution at the Mount Dennis facility was applied to the maximum number of buses proposed at the McNicoll Bus Garage to determine the number of buses that would be leaving / entering the facility during any given hour for this assessment. The same hourly vehicle distribution was assumed for every day of the week. The projected bus movements in and out of the facility were used to determine traffic volumes north and south on Redlea Avenue for each hour. Table 3 gives the estimated daily bus and car volumes as a result of the facility, and Figure 4 shows the bus movements from the facility. It was assumed that 50% of employee vehicles would arrive / leave north on Redlea Avenue, and 50% would arrive / leave south. Novus Environmental 6

133 TTC McNicoll Bus Garage Roadway Assessment Final Report January, 2015 Note that traffic was not modelled on McNicoll Avenue because only the buses operating on the existing bus route along McNicoll Avenue will take this route from the facility. Therefore, there is no additional bus traffic on McNicoll Avenue due to the facility. Buses and vehicles were, however, modelled idling at the new traffic light to be installed at the intersection of McNicoll Avenue and the proposed Redlea Avenue in all four directions. Table 3: Predicted Vehicle Volumes at the Facility Segment Daily Bus Volume Daily Car Volume Redlea Ave North of the Facility Redlea Ave South of the Faciltiy Figure 4: Project Bus Movements in and out of the Facility Novus Environmental 7

134 TTC McNicoll Bus Garage Roadway Assessment Final Report January, Emission Rates Modelled emissions rates were determined from the US EPA s MOVES model. Table 4 presents the predicted emission rates from the contaminants of concern. In addition to the emission rates presented below, re-suspended silt loading was also considered as a source of particulate matter emissions. Table 4: Predicted Vehicle Emission Rates (g/s) Carbon Monoxide Oxides of Nitrogen Benzene 1,3- Butadiene Formaldehyde Acetaldehyde Acrolein Nitrogen Dioxide PM10 PM2.5 Bus - idle 3.5E E E E E E E E E E+00 Bus - 60 km/hr 3.1E E E E E E E E E E-01 Car - idle 3.6E E E E E E E E E E-01 Car - 60 km/hr 7.0 Results 4.4E E E E E E E E E E-02 Table 5 presents the worst-case concentrations from the roadway as well as from background and from the facility assessment. It can be seen that the contribution of the roadway is marginal, with the highest concentration as a percent of the standard being PM2.5 at 5.2%. There are a predicted two additional exceedances of both the PM10 and PM2.5 standards as a result of the roadway traffic over the 5-year period and no additional exceedances of benzene. This equates to less than 1% of the time. Table 5: Worst-Case Predicted Concentrations as a Percentage of the Guideline including the Roadway Assessment Contaminant Averaging Period Maximum Concentration Due to Roadway Alone (µg/m 3 ) Maximum Concentration Due to Roadway Alone (as % of Standard) Combined Concentration as % of Standard (Ambient + Facility + Roadway) Maximum 90 th Percentile Average Additional # of Guideline Exceedances due to Project Over 5 Years (Exceedances due to Roadway) NO 2 1-hour % 84% 58% 52% 24-hour % 61% 40% 30% CO 1-hour % 6% 1% 1% 8-hour % 12% 3% 2% PM hour % 144% 58% 35% 5 (2) PM hour % 140% 54% 32% 3 (2) Acetaldehyde 24-hour % 1% <1% <1% Acrolein 24-hour % 43% 30% 25% Benzene 24-hour % 109% 52% 37% 6 (0) 1,3-Butadiene 24-hour % 3% 1% 1% Formaldehyde 24-hour % 13% 8% 5% Novus Environmental 8

135 TTC McNicoll Bus Garage Roadway Assessment Final Report January, CWS guideline for PM2.5 is based on an average annual 98 th percentile concentration, averaged over 3 consecutive years. The maximum combined 3-year rolling average of the annual 98 th percentile concentration was 22.14, which is 82% of the guideline. It should be reiterated that despite the minimal impact of the roadway, these predicted concentrations are likely over-conservative. The chances that the maximum impacts from ambient, the garage, and the roadway would occur simultaneously is very small. 8.0 Conclusions This assessment shows that the impact of the additional bus and vehicle traffic on Redlea Avenue and McNicoll Avenue is very low. Given the conservative nature of the modelling and the predicted low number of exceedances of the guidelines, mitigation measures are not warranted. Sincerely, Novus Environmental Inc. Hamish Corbett-Hains, M.A.Sc., P.Eng. Air Quality Engineer Scott Shayko, Hon.B.Comm, B.Sc. Principal Novus Environmental 9

136 Ministry of the Environment and Climate Change Central Region Technical Support Section 5775 Yonge Street, 8 th Floor North York, Ontario M2M 4J1 Tel.: (416) Fax: (416) Ministère de l Environnment et de l Action en matière de changement climatique Région du Centre Section d'appui technique 5775, rue Yonge, 8 ième étage North York, Ontario M2M 4J1 Tél. : (416) Téléc. : (416) March 2, 2015 TO: FROM: Subject: Solange Desautels Marinha Antunes Technical Support Air Quality Comments TTC McNicoll Bus Garage Draft Report Redlea/ McNicoll Roadway Assessment Transit Project Environmental Assessment The following memorandum summarizes Central Region Technical Support Section (TSS) Air Unit comments on the Roadway Assessment Report in support to McNicoll Bus Garage Transit Assessment prepared by Novus Environmental and dated January 26, As discussed during the January 13, 2015 teleconference with TTC, TPH and ministry staff, the Traffic Assessment Report addressed Toronto Public Health (TPH) comments and public concerns with respect to the proposed increased bus traffic and employee vehicle traffic on the proposed Redlea Avenue and McNicoll Avenue. The methodology used to estimate vehicle emissions and re-suspension of particulate follows the typical transportation air quality assessments in Central Region. MOVES model was used to estimate vehicle emissions and CAL3QHCR was the dispersion model employed to assess air quality impacts at the closest sensitive receptors, which follows the ministry s guidelines. At this time, TSS has no comments on the Roadway Assessment; however, the recommendations below are offered for the proponent s consideration: Dust mitigation measures should be in place to minimize dust impacts at the most impacted receptors during construction activities. For a more comprehensive list of dust mitigation measures during construction activities can be found in a report prepared for Environment Canada entitled Best Practices for the Reduction of Air Emissions from Construction and Demolition Activities (Cheminfo Services Inc., March 2005). Ministry Comments - McNicoll Bus Garage Transit EA Page 1 of 2 March 2 nd, 2015