Background



Double Double Your Drive-thru Emissions

A Case Study of an Edmonton Tim Horton’s Facility and

a City Wide Analysis of This Service

Prepared by:

Miranda Baniulis, Curtis Boyd, Amanda Dacyk, Jody Gobert,

Jocelyn Howery, Kendra Issac, Todd Keesey, Jennifer Martin

Submitted to:

Dr. Peter Boxall

Methods and Applications in Environmental Economics (ENCS 410)

Edmonton, Alberta

April 27, 2006

1. INTRODUCTION

Air pollution is one of the most pressing issues facing communities around the world today. Emissions from the commercial, industrial and private sectors contribute to global problems including ozone layer depletion, acid rain and the degradation of air quality in urban centers. The growth of the human population and our increasing demands on the environment continue to challenge governments and NGOs trying to address air quality issues. One of the biggest problems with air pollution is that it cannot be contained within domestic borders. The fact that air pollution does not obey borders means international agreements are often necessary to prevent some countries from abating air pollution while others do not. This issue of free-riding can make international efforts to curb emissions difficult, as it is often costly to take on abatement technologies or increase regulations that could potentially deter foreign investment. This can be especially true for developing countries that make extensive use of manufacturing and production to increase their income. These nations may be willing to trade environmental degradation for a better standard of life.

It is estimated that the global temperature increase resulting from greenhouse gas (GHG) induced climate change will have severe and far-reaching implications. Although it can be difficult to attribute precise weather events to climate change, severe weather events, such as the devastating hurricanes seen in the southern United States in 2005, are expected to increase. Droughts and floods may also become regular occurrences or could take place in areas that had previously never experienced these phenomena. Other issues that may arise include severe water shortages, as glaciers that normally feed streams and river recede, the inability to grow food in drought stricken areas, and severe heat waves like the one that killed 150,000 Europeans in 2003 (Government of Canada 2005). The Government of Canada (2005) fears that these climate related incidents may lead to the displacement of millions of climate change refugees.

Canada became a signatory to the Kyoto Protocol in 1997, agreeing to reduce GHG emissions by 270 Megatons during the 2008 to 2012 commitment period (Government of Canada 2005). The 2005 federal plan to meet this target included capping emissions by Large Final Emitters, creating a domestic offset credit system and enticing source reductions and sink development. The plan also addressed the fact that individual Canadian citizens are collectively responsible for 28% of Canada’s total emissions (Project Green 2005).

According to the United State’s Environmental Protection Agency (EPA), emissions from personal automobiles make up the largest proportion of urban air pollution in numerous cities across the continent (EPA 1994). The Government of Canada (2005) estimates that 50% of the average Canadian’s GHG emissions come from their use of passenger vehicles. In order to fully understand this component of GHG emissions literature regarding the chemical makeup of vehicle emissions, how these emissions vary by vehicle type and the environmental and human health impacts associated with these compounds was reviewed. Once this information is considered, the true costs of actions creating GHG emissions may be recognized. It is then possible to debate the necessity of these activities and discuss the possibility of their reduction.

This study looks at the amount of emissions created by vehicles idling in drive-thru lines in Edmonton, Alberta, Canada. Average daily emissions produced by a single drive-thru, daily and weekly usage patterns and total daily emissions by all Edmonton drive-thrus were calculated. This process shows the relative significance of drive-thrus as an emission source. It is hoped that this study will initiate discussion on the necessity of drive-thrus and the policies surrounding them.

2.0 BACKGROUND

2.1 Chemical Composition of Vehicle Emissions

Though the by-products of perfect combustion are relatively innocuous (carbon dioxide and water), the typical process is less benign. This process includes the production of a variety of gaseous compounds including hydrocarbons (HC), nitrogen oxides (NOx), carbon monoxide (CO) and carbon dioxide (CO2) (EPA 1994). In a study of over one-hundred thousand vehicles, Beydoun and Guldmann (2006) found CO emissions varied between 0 and 899 grams/mile (g/mile), (median = 4.93 g/mile); HC emissions between 0 and 97.3 g/mile, (median = 0.51 g/mile); and nitrous oxides between 0 and 5.65 g/mile, (median = 1.17 g/mile)[1]. In addition to these main components, other chemicals such as sulphur, formaldehyde and acetaldehyde are often found in vehicle exhaust (Kirchstetter et al. 1996).

Concerns about these chemicals are related to their environmental and health impacts. The reaction of hydrocarbons and certain nitrogen oxides produces ground-level ozone, a major component of smog and currently one of the most prevalent urban air pollution problems (EPA 1994). The EPA (1994) also lists hydrocarbons as potential carcinogens. Nitrogen oxides contribute to acid rain. Carbon monoxide reduces a person’s ability to utilize oxygen (EPA 1994). Though carbon dioxide is not a direct health hazard, this gas plays a large role in global warming (EPA 1994). This impact is particularly important, as fourteen percent of vehicle exhaust by volume is carbon dioxide (Bishop and Stedman 1996). The identification of the negative implications of these chemicals indicates that emission-reduction policies are desirable, and thus, studies on potential sources of emissions, such as the one proposed, are socially desirable.

In relation to the proposed drive-thru study, it is important to note that the above figures may not be representative of the emissions of idling vehicles, and instead should serve as an exaggerated illustration of the potential amounts of chemicals produced and a basis for assessing negative affects. This caution is supported by the findings of Wenzel et al. (2000) who determine that NOx emissions are generally very low during idling. These investigators also find that HC and CO emissions, measured when a vehicle is under load, can differ greatly from those produced in resting situations (Wenzel et al. 2000).

2.2 Factors Influencing Emission Rates

Many factors influence the emissions rates of vehicles. These include: vehicle make, vehicle age and maintenance, vehicle weight and engine size, seasonality and others. Variation in emission rates may span several orders of magnitude between vehicle manufacturers (Wenzel et al. 2000). Differences in technology, engine design and mechanical components may result in certain automobile makes producing fewer emissions than others, that is, being environmentally “cleaner” (Beydoun and Guldmann 2006). In their analysis of data from the Enhanced Emissions Testing Program from three American states, Beydoun and Guldmann (2006) argue that emission levels vary significantly by make. They name manufacturers whose vehicles failed emission tests in less than 5% of observations (Kia, Lexus, Infiniti, Saturn, Honda, Suzuki and Toyota) and in more than 12% of cases (American Motors, Eagle, Mercedes-Benz, Cadillac and Audi) (Beydoun and Guldmann 2006) [2]. In a regression analysis, these researchers analyzed the influence of a vehicle’s manufacturer on its probability of failing an emissions test. It was found that BMWs were least likely to fail, while Hyundai, Mitsubishi, Chrysler and General Motors vehicles were most likely to fail. The extent of this variation is such that the probability of failure of the cleanest vehicles is ten times less likely than that for the dirtiest vehicles (Beydoun and Guldmann 2006). These results conform closely to the findings of Bin (2003) who determined the rate of emissions test failure is significantly lower for foreign vehicles than domestics.

The significance of these findings suggests that, to maximize the accuracy of the estimation of drive-thru emissions, researchers must record the makes of the vehicles they time in drive-thru lines. Ignoring this variable implies all types of vehicles produce equal emissions, which is inaccurate. The huge range of emission rates reduces the effectiveness of using an average value. Instead, a profile of the vehicles most commonly found in drive-thru lines should be developed and the appropriate adjustments made. It must be recognized that the current study does not make distinctions between makes of vehicles. Although, this would have given a better estimate of the emissions from fast food drive-thrus, it was beyond the scope of the study.

Overall, older vehicles produce significantly higher emissions than newer ones (Beydoun and Guldmann 2006). This finding is not surprising as increased mileage (generally corresponding well to increased age) tends to increase the deterioration of parts and reduce functioning of emission-control equipment (Beydoun and Guldmann 2006). The progression towards cleaner vehicles in recent years is also attributable to the tightening of emission laws (Wenzel et al. 2000) and technological development (Bishop and Stedman 1996).

Vehicle condition has been shown to overshadow the impact of age (Bishop and Stedman 1996). Bishop and Stedman (1996) document numerous cases of old non-catalyst vehicles with good maintenance records producing lower emissions than newer cars that had not received regular repairs and adjustments. In addition to poor maintenance, excessive maintenance, known as tampering/modifications, often disrupts the emission-control systems and increases pollution (Wenzel et al. 2000).

These findings suggest that the most accurate estimation of vehicle emissions would require data on vehicle condition to be collected. The mechanics of doing this could become quite subjective if non-disruptive sampling is the only method available. A possible means of incorporating partially reliable data on age/condition would involve adopting the previous recommendation, that is, recording vehicle make. Since the year of introduction of each make is known, the maximum age of each vehicle can be approximated and a model generated incorporating mechanical deterioration. It is likely impossible to determine maintenance history accurately without inspecting the car or interviewing the owners.

Vehicle size was found to significantly impact emission levels, with trucks generating higher emissions than cars (Beydoun and Guldmann 2006). The EPA provides idling emission rates for three major weight classes, with the emissions of heavy-duty trucks being approximately double that of light-duty cars (Appendix C) (EPA 1998). As it is these values that would be most useful to the proposed study, cars passing through the drive-thru were separated into three groups: car, small truck, van or SUV (less than three quarters of a ton) and large truck, van or SUV (greater than three quarters of a ton).

Beydoun and Guldmann (2006) found the probability of a vehicle failing an emissions test in winter is significantly higher than that for the same vehicle in summer. This is likely attributable to a reduction in the effectiveness of emission-control technology at low temperatures (Wenzel et al. 2000). This trend implies that different emission production rates should be used for spring and summer versus fall and winter. In this way, the accuracy of annual city-wide emission predictions will be maximized. It is important to note that, as average winter temperature in Edmonton is likely lower than -1.1 degrees Celsius, the rates of idling emission production reported by the EPA (1998) for this season (Figure 1.) will be conservative values when used in this study.

In addition to the previously mentioned determinants of emission rates, several other factors exist with the potential to influence these values, including: fuel type (EPA 1998), fuel economy (Harrington 1997), driving patterns (Beydoun and Guldmann 2006) and mechanical engine characteristics (Bin 2003). Due to the relatively small variation in these factors between vehicles in the studied city these factors will not be discussed.

When all the above factors interact in a negative way; that is, when a vehicle is a domestic make, old and poorly maintained, heavy and operates in a cold environment, there is a high probability that the vehicle is a “gross-polluter” (Bishop and Stedman 1996). Wang et al. (2005) note that, though the exact figures vary slightly, it is generally accepted that this minority of vehicles produces the majority of emissions. Beydoun and Guldmann (2006) elaborate this conclusion by illustrating that gross-polluters, described as including between 5 and 20 percent of urban vehicles, produce 50 to 80 percent of total automobile emissions. The presence of this group is relevant to the proposed study, as it will be important to make some allowance for the presence of gross-polluters in the sampling strategy.

2.3 Health Risks from Vehicle Emissions

Due to the fact that vehicle emissions are harmful to human and environmental health, it is important to include these impacts in our discussion. The relationship between children’s health and air pollution is the subject of many studies. To test the hypothesis that traffic-related air pollution causes childhood asthma, Gauderman et al. (2005) sampled NO2 concentrations at the residences of 208 Californian children. These researchers examined the influence of the proximity of the residence to the nearest freeway, the average number of vehicles traveling within 150 meters of the residence each day and the model-based estimates of traffic related air pollution at the residence (Gauderman et al., 2005). This study strengthened the emerging body of evidence stating that air pollution can cause asthma and that traffic related pollutants are partly responsible for this effect (Gauderman et al., 2005).

McCubbin and Delucchi (1999) attempted to approximate the costs associated with the ailments and deaths resulting from four criteria of vehicle pollutants and six toxic vehicle pollutants. They did so by estimating emissions related to motor vehicle use, and changes in exposure to air pollution. They then related changes in air pollution exposure to changes in physical health effects, and changes in physical health effects to changes in economic welfare (McCubbin and Delucchi, 1999). Four emissions sources were studied. Tailpipe and evaporative emissions from vehicles are of particular importance to this drive-thru study.

2.4 Mixed-Use Zoning and the Urban Environment

Urban sprawl is often characterized by vehicle dependence. Goldberg (2002) states that the combination of neighbourhoods not suited to walking, sedentary lifestyles and the drive-thru diet, causes one in four of today’s kids to suffer from diabetes as an adult. As well, since urban sprawl may be associated with an increase in driving rates, the gains made in reducing air pollution in other areas may be negated. This occurs at a time when asthma rates among children are soaring (Goldberg, 2002).

City design largely determines how one travels within a city. Because restaurants depend on a high volume of vehicular traffic and a high turnover rate of customers within the site (Bedford and Dill, 2002), they are most often situated close to major roadways. This can have significance for pedestrian safety and for the design of the urban landscape.

Angotti and Hanhardt (2001) discuss the issues city planners must address when allocating land for mixed uses (residential, commercial and industrial) using New York City to illustrate these issues. In this paper, two noteworthy observations were made. The first is that, although pollution may decrease from people not using their vehicles, increased pollution may result from the commercial and industrial sectors (Angotti and Hanhardt, 2001). The second observation is that mixed-use areas with industrial uses are disproportionately allocated to lower income neighbourhoods (Angotti and Hanhardt, 2001).

In the paper written by Handy et al. (2005), it was hypothesized that an environment, where residents are closer to destinations and that has viable alternatives to driving, is associated with less driving. As well, moves to environments where residents are closer to destinations and have viable alternatives to driving, are associated with a decrease in driving (Handy et al., 2005). The change in neighbourhood design was captured by questioning five hundred individuals who had recently moved to California neighbourhoods and five hundred individuals who were already situated in these neighbourhoods (Handy et al., 2005). Results showed significant associations between changes in travel behaviour and changes in neighbourhood design (Handy et al., 2005). This supports prior evidence which states that mixed land-use decreases vehicle travel, thereby reducing vehicle emissions.

In contrast to Handy et al. (2005), Lam and Niemeier (2005), constructed a simple model that found mixing residential and business uses in one city may result in increasing residential housing prices, which could displace incumbent residents to a neighbouring city. Their model, which was a simplified theoretical model used to look at policy options, found that an increase in housing price, due to mixing of residential and commercial uses, may increase inter-city traffic which would increase vehicle emissions (Lam and Niemeier, 2005). This could be an interesting factor relating to drive-thru usage in Edmonton, as Edmonton has several surrounding communities. This factor, however, was beyond the scope of this study.

2.5 Current and Future Trends in the Fast Food Industry

Trends in food consumption are continuously changing. Currently, there is a strong reliance on fast food restaurants. Research has shown a decline in in-home activities, such as cooking meals, while there has been an increase in activities requiring travel (e.g. picking up food) (McGuckin and Nakamoto 2004). Sales at fast food restaurants surpassed sales of full service restaurants for a short period in the mid 1990s, but there is constant competition among food service firms for the consumer’s away-from-home-dollar (Stewart et al. 2004). The focus on convenience and on low priced food has allowed fast food chains such as McDonald’s to capture a significant portion of this growing market. In an attempt to make their food more available to consumers, many fast food outlets now have two or three drive-thru windows (Jekanowski 1999). These trends have sparked growing concerns about the health impacts, environmental impacts and energy use resulting from the consumption of fast food.

Byrne et al. (1998) described the demographic and socioeconomic variables that lead to spending on food away from home at different types of food facilities. They developed a framework for the decision process and estimated how socioeconomic variables would affect each household’s choice of facility type. Their findings suggest that a large family size has a positive effect on expenditures at fast food restaurants, as does seasonality (summer) and increased hours worked by the home manager (Byrne et al 1998).

Stewart et al. (2004) examined the changing economic and demographic trends that they felt would affect the future demand for food away from home. They observed trends in key U.S. characteristics that influence the demand for food away from home and found that incomes are rising, the population is aging, and household sizes are getting smaller. They also identified the changing structure of households from “traditional families” to single person or multiple adult dwellings with no children. Using the Shonkwiler and Yen statistical model, they estimated the relationship between household characteristics and spending at fast food and full service restaurants. They then used projected changes of these characteristics (for the year 2020) in a simulation and found the effect was increased per capita spending of 18% at full-service restaurants and 6% in fast-food outlets. This simulation was based on the assumptions that there will be no change in the relative prices of fast and full-services foods, and that household characteristics will continue to influence consumer spending in the same manner. The model also holds the number and locations of restaurants and the food and service mix constant (Stewart et al. 2004).

There is a prevailing idea in North America that a strong economy is founded on cheap food (Cummings, 1999). Fast food companies have become vertically integrated in all aspects of food production, processing and retailing giving them significant market influence. Through their buying practices, these businesses have the ability to influence the agricultural sector, but usually not in favour of small-scale farms. Cummings (1999) also suggests that these corporations have used this influence to manipulate public perception by marrying food and entertainment, thus increasing their influence in consumers’ spending habits. She feels that their sway has been instrumental in inciting demand for convenience food, which has led to obesity and health problems for many North Americans (Cummings, 1999).

Jekanowski (1999) suggests that it is unlikely that time spent preparing meals in the home will increase. He says that, even if incomes decline, the growing time constraints faced by working people, as well as the general decline in cooking knowledge, will continue to increase the demand for convenient fast foods (Jekanowski, 1999). The advantage of franchise chains to respond to consumer preferences through economies of scale as suggested by Stewart (2005), may allow them to keep costs down by spreading them across many stores. These cost advantages and consumer trends, along with the growing demand for convenience in away from home food, support the perception that the fast food industry will continue to play a significant role in Canadian diets.

Jekanowski (1999), reports that 60% of burger sales at Burger King and 54% of burger sales at McDonald’s were made at the drive-thru window. Apprehensions about the impacts of drive-thrus have led some communities in the United States, including Carrboro in North Carolina, San Juan Capistrano and Sierra Madre in California, to ban construction of drive-thrus (Guilford 1998). This has also become a predominant issue in Canada, as seen in the push for a similar drive-thru ban in Toronto (Hume 2004). In this city, a ban was put into place in 2003 and upheld in 2004, despite strong legal opposition from McDonald’s, Tim Horton’s, Wendy’s, Burger King, five major banks and Canadian Tire Real Estate.

Much of the literature concerning the future state of the fast food industry suggests that growth will continue be positive, but that it may be smaller than in previous years. These predictions are based on general trends in attitudes and spending seen in the U.S.A. and Canada. The application of current literature to our study would seem to indicate that we will see small positive increases in future volumes of GHGs resulting from fast food drive-thrus, ceteris paribus. This information will be useful to individuals and government in addressing concerns about emissions from automobiles in the City of Edmonton.

3.0 OBJECTIVES

The objective of this study was to determine the number of cars, light duty trucks and heavy duty trucks that utilize the drive-thru window at the Tim Horton’s restaurant located at 11084 Street and 51 Ave NW in Edmonton, Alberta. Estimating the average idling time per vehicle was another primary goal. By gathering this information, the average daily emissions produced by this drive-thru could be determined, daily and weekly usage patterns could be uncovered, and a regression on the significance of time and day’s effect on emission rates could be created.

In addition to the illustration of consumer behaviour at one specific drive-thru location, a second objective of this study was to extrapolate these basic results into an estimation of the total daily emissions produced by all Edmonton drive-thrus. By using measures of traffic density and relative franchise popularity, the daily emissions of all the major drive-thru chains in this city could be approximated. This process would allow the relative significance of drive-thrus as an emission source to be assessed and initiate discussion on the necessity of policy changes regarding these establishments.

4. METHODS

4.1 Determining Consumer Behaviour for the “Base-Case” Tim Horton’s

Data was collected in 15-minute periods to allow for sampling flexibility, although, samples were often taken in 3 hour blocks. To calculate the quantity of emissions produced from this Tim Horton’s location, the number and type of vehicles using the drive-thru as well as the time each vehicle spends idling in the drive-thru was monitored. In order to complete this task, the following steps were taken. As each vehicle entered the drive-thru line the time was recorded and the vehicle type (classification) was recorded. Vehicles were classified into three categories; cars were grouped into one type with an assumed maximum gross vehicle weight (GVW) of 6000lbs. As per the U.S. EPA categorization, light-duty vehicles include those vehicles with a GVW of 6000lbs to 8500lbs, while heavy-duty vehicles include those with a GVW greater than 8500lbs. To simplify this classification, heavy-duty vehicles were those vehicles with an engine equal to, or greater than, 3500 (example an F-450). The time at which the vehicle came to a stop at the end of the drive-thru line was recorded as “time-in”. The time the vehicle began to pull away from the drive-thru window was recorded as “time-out”. Observations continued until the last car arriving during the period left the drive-thru window. All times were recorded to the nearest second.

216 samples, totalling 54 hours, were conducted at a single Tim Horton’s, located at 11084-51 Avenue, Edmonton, through five weeks in February and March. This location may be characterized as a very busy location due to its proximity to a traffic corridor on 111th St. Sample times were distributed to ensure that every 15-minuted time period was recorded at least once from 5am to 10pm. Samples were taken through varying days to ensure a distribution such that an ‘average day’ could be computed.

All observations were entered into an Excel spreadsheet. Idling time was calculated by taking the difference between the “time-out” and the “time-in” for each observation. For each 15-minute sample, sub-total idling times were calculated for each size of vehicle: cars, light duty trucks/vans/SUVs, and heavy-duty trucks/vans/SUVs. The total idling time for each 15-minute sample was calculated by adding these three values together. A full day (5am to 10pm) of idling time was compiled for this facility, using the average idling time calculated for each hour if this period had been sampled more than once. This full day was used in conjunction with U.S. EPA and Government of Canada data to compute average GHG emissions per day and per year from this restaurant’s drive-thru. The U.S. EPA data (Figure 1) outlines the amount of NOx (g/min), CO (g/min) and VOCs (g/min) emitted from passenger cars, light-duty vehicles and heavy-duty vehicles. Estimates of per day emissions of these various pollutants were then calculated.

In addition to compiling the data to comprise an average day, data was combined into its respective hourly blocks with dummy variables for time of observation, day of the week, and week observed. SPSS 11.0 was then used to perform a regression analysis to evaluate the existence of any significant difference in hourly idling times according to these dummy variables.

Realizing that this 5am to 10pm data is an underestimate daily pollutants, average hourly idle times were correlated with hourly traffic flow (City of Edmonton, Transportation and Streets Department, 2006) to attempt to extrapolate a full 24 hours of emissions.

4.2 Extrapolation of Observations for the City of Edmonton

In order to estimate the total quantity of GHG emissions produced by drive-thru lines across Edmonton, it was necessary to determine the number of these restaurants in the city and their rates of use. To accomplish the former, the address and the drive-thru hours of every location of eight selected restaurant chains were determined by contacting each locating. The restaurants were selected based on the groups’ informal observations of common drive-thru restaurants and were not assumed to be complete. The traffic density on the nearest roadway was determined for each location using a traffic density map. If the location was between two roads, an average of the densities was used.

Once an estimate of the number of drive-thrus in Edmonton was complete, the evaluation of relative use rates was performed. It was assumed that idling time at these other restaurants was a function of nearby traffic density and relative restaurant popularity. Although this assumption was undoubtedly simplifying, it allowed us to use a traffic density map and a small number of observations to predict the usage of each location’s drive-thru.

First, to establish the relationship between traffic density and idling time, idling-times were observed at five Edmonton Tim Horton’s restaurants in areas of varying traffic density (one low density and three medium density) (City of Edmonton, Transportation and Streets Department 2006), between the times of 5 and 6 pm on Friday, March 17th. Next, hour-long samples were conducted at seven drive-thru locations around Edmonton, each at a different major restaurant chain, including McDonald’s, Kentucky Fried Chicken (KFC), Burger King, Arby’s, Harvey’s, Dairy Queen, and Taco Bell. Sampling occurred at a time and day that had previously been sampled at the 11084-51 Avenue Tim Horton’s. Since the relationship previously observed between traffic density and idling time could only be assumed to hold between 5 and 6 pm, and the sampling at other restaurants occurred at different times throughout the day, these observations had to be adjusted to represent the idling time that would have been observed if the sample had been taken between 5 and 6 pm. To do this, the compiled average hourly idling times for the base case Tim Horton’s were presented as proportions of the total daily idling time (see Appendix A for sample calculations). These proportions were assumed to hold for all restaurant chains, however slight adjustments were made for restaurants whose drive-thru were not open the entire day (5 am to 10 pm).

Once the proportion of total daily idling time was known for each hour period, the samples taken at the other restaurants at varying times were adjusted to what they would have been predicted to show if they had occurred between 5 to 6 pm. The original traffic-density/idling time relationship was then used to predict the Tim Horton’s idling time for a restaurant on a roadway of that density. The observed hourly idling time (adjusted for time of day) for restaurant x was divided by the idling time predicted for a Tim Horton’s at that traffic density to give a “popularity factor”. Total daily idling time for each restaurant was calculated by multiplying the inverse of the proportion of daily idling time occurring between 5 and 6 pm by the coefficient for traffic density for that restaurant by the popularity factor for each restaurant chain.

The daily idling times of each restaurant were summed to give estimated total daily idling time for Edmonton drive-thrus. This value was used to calculate daily and annual emissions of carbon dioxide (CO2), carbon monoxide (CO), hydrocarbons (HC) and nitrous oxides (NOx). The coefficients used to calculate the non-CO2 emissions were available from the U.S. Environmental Protection Agency and given for each vehicle type. Winter production rates were used (Figure 1.). The rate of CO2 production varies greatly depending on vehicle mileage (less-related to vehicle weight); therefore, a constant rate of 64.9 g per minute of idling was used (Natural Resources Canada, 2006).

5. RESULTS

5.1 Results of the “Base Case” Tim Horton’s Analysis

Average emissions for one day at the Tim Horton’s location on 11084 Street and 51 Ave NW, are a product of the average idling time per day per vehicle type and the emissions emitted by each vehicle type.

Overall, through the 54 hours of observations, there were 3756 vehicles observed contributing 320 hours, 55 minutes and 48 seconds of idling emissions. From the entire data set, an average idling time of five minutes and eight seconds was recorded with the shortest average idling time being 59 seconds and the longest average idling time being twelve minutes and 37 seconds.

Depicting the average day, Figure 2 displays the average hourly wait times and observed vehicle classifications. 1085 vehicles used this drive-thru facility for a total idling time of 88 hours 23 minutes and 49 seconds on an average day. The distribution of vehicles categories using the drive-thru is seen in Figure 3.

Using the average day, it was possible to combine this data with the U.S. EPA data to find the average daily emissions from this site (Figure 4). The total emissions, including HC, CO, NOX, and CO2, total 385.5 kg per day. 82% of these total emissions are comprised of CO2 emissions. In noting the quantity of emissions produced it should be again noted that these numbers are assumed to be an underestimate, and because there is no correlation between daily traffic distribution and hourly drive-thru usages, it was not possible to extrapolate a full day with certainty.

The SPSS regression output did depict some significant variation from the base as defined as 7:00 Sunday in the first week observed. The regression reveals an R2 value of 84% which is relatively large. The regression output (figure 5) shows that Wednesday and Saturdays are significantly busier than Sunday. There were no significant differences between weeks, although those times which are significantly less busy than 7:00AM are 5:00AM, 5:00PM and 6:00PM. The fact that there were fewer observations recorded in the tails of the day may impact the reliability of these numbers as depictions of an average day.

5.2 Results of the City Wide Analysis

Using the observations made at the base case Tim Horton’s, the total levels of certain vehicle emissions (CO2, CO, HC and NOx) produced from drive-thru restaurants throughout the City of Edmonton were predicted. In total, 114 branches from nine restaurant chains were considered (Figure 6). Both traffic density and restaurant popularity were assumed to play a role in determining the relative amount of use of each drive-thru (compared with that of the base case Tim Horton’s). Thus, this results section includes findings on the influence of these factors as well as actual estimations of city-wide emissions.

Firstly, it was found that traffic density predicted approximately 64% of the variation in Tim Horton’s drive-thru usage between locations between 5 and 6 pm (Figure 7). As this relationship could only be assumed to hold for this particular hour, we were forced to assume that the daily trends in idling times observed for the base case Tim Horton’s remained constant across all restaurants in the city (Figure 8). It was determined that idling observed between 5 and 6 pm made up between three and six percent of a restaurant’s total daily idling time (depending on the hours of operation of the drive-thru window).

It was also found that drive-thru use at Tim Horton’s was substantially higher than the use of this service for all other restaurant chains. That is, the popularity of Tim Horton’s drive-thrus far exceeded that of any other restaurants included in this study (Table 1).

Daily idling times for each restaurant were estimated to range from 233 hours/day for a Tim Horton’s to 1.7 hours/day for a Burger King. The average daily idling time for each drive-thru location in Edmonton was predicted to be 47 hours/day. The estimated total daily drive-thru idling times for Edmonton was 5383 hours/day. This idling time was distributed unequally between the restaurant branches, with McDonald’s and Tim Horton’s together accounting for 81% of total daily drive-thru idling time in this city.

The total amount of GHGs produced by vehicles in drive-thru lines in Edmonton was estimated to be 23.5 tonnes per day [approximately 8600 tonnes per year]. Carbon dioxide constituted approximately 90% of these emissions at 21 tonnes per day, while carbon monoxide, nitrous oxides and hydrocarbons together made up the remaining 10% (2.5 tonnes per day). Of the non-carbon dioxide emissions, carbon monoxide was the most highly produced (Figure 9). Although cars were the most common category of vehicle observed in drive-thru lines, light trucks were the largest contributor to Edmonton’s daily drive-thru line emissions (Table 2).

6. DISCUSSION

6.1 Limitations of the Base-Case Analysis

The majority of data was collected by pairs of samplers to increase the accuracy of the results however, some observations were collected by individuals and may be less accurate if the sampler was busy and failed to notice the exact time a vehicle entered or left the drive-thru line-up. Lengthy drive-thru line-ups and multiple entrances to line-ups made it difficult to see when vehicles arrived. As a result, some observations may be an underestimate of the actual idling time. Time restrictions were a major hindrance in data collection, limiting the total number of samples that could be taken.

Another limiting factor of this study was the lack of detailed vehicle knowledge of some surveyors, potentially causing them to assign vehicles to the wrong class. It was also difficult to tell, without directly speaking with the driver (which we did not do), the make, age, and maintenance of a vehicle. As Wenzel et al. (2000) demonstrate that these factors are highly influential in determining a vehicle’s emission rates, ignoring these characteristics reduces the accuracy of our findings.

6.2 Extrapolation of Observations for the City of Edmonton

6.2.1 Limitations of this Study

This study was performed under conditions where sampling effort and research/analysis time were limited. In many cases, these limitations forced us to make large assumptions reducing the accuracy of our findings. The shortcomings of our methods must be presented in some detail, in order that possible errors in our results are understood.

Firstly we were forced to assume that there is no seasonality in drive-thru usage. As well, our methods imply that all drive-thrus follow the same daily usage patterns as Tim Horton’s. In addition, we were forced to assume that sampling was sufficient to get representative samples of the relative busyness of each drive-thru chain.

These assumptions are unlikely to be completely realistic. For example, the number of people visiting a Dairy Queen drive-thru in July is likely to be higher then it is in April. As well, variation in the busy times for each restaurant was not considered. This could be an important factor that may have led us to over/underestimate the emissions. Since we did not have a full day of sampling at each restaurant we were forced to assume that idling time observed between 5 and 6 pm made up the same proportion of total daily idling time for all restaurants sampled. This figure was based on data from the main Tim Horton’s. As Tim Horton’s is a coffee/breakfast restaurant and does a large portion of their business in the morning, using this franchise as our base-case potentially skews our results.

Adding to the problem of a lack of detailed data for the additional restaurants is the lack of sufficient data for the traffic density-idling time correlation. Although our R2 was quite high (0.64), additional samples can only decrease our confidence in the linear regression. Our ‘Traffic Density’ linear regression (which was used to predict idling minutes based on vehicles per week on the nearest roadway), was based on 5 samples, covering a range of 7500 to 28 700 vehicles per week on the nearest roadway. This regression was extrapolated out to 57 100 vehicles per week, it was assumed that the regression was linear and went through the origin. We are also assuming that each sampling time was a representative sample. These assumptions may not hold true and thus, decrease the reliability of our citywide emissions predictions.

To add reliability to our predictions, more data would need to be collected to add to the strength of our correlation, make our data more representative and overcome the problems of different restaurants having different daily time distributions. As well, data from different seasons would be necessary to account for seasonality in drive-thru usage.

The current lack of detailed data may considerably decrease the reliability of our results. Therefore, our results should not be considered as fact, but rather an estimation which is subject to many sources of error. The findings of this investigation should be presented only in context with the limitations of the methods.

6.2.2 Comparison to Existing Literature

Published literature on emissions produced by vehicles in drive-thru lines is somewhat limited, however, the results of this investigation do show some similarities to those obtained by a group of students in Massachusetts in 1999. These researchers estimated the daily amount of emissions produced by the drive-thru at one selected McDonald’s location in their county (Shusas, 2000). Their recorded average of 2237 idling minutes per day (Shusas, 2000) is approximately 20% lower than the overall daily average by store obtained in our study (2833). It is approximately 40% lower than the daily idling time average obtained for Edmonton McDonald’s restaurants. This finding is slightly curious, however, it may be explained partially by the fact that that Tim Horton’s restaurants were included in our investigation. Since daily idling times at Tim Horton’s are well above average rates (McDonald’s showed only approximately 50% of the daily idling time in our study), including this chain predictably raised the average daily idling time estimated for all Edmonton drive-thrus. As well, although some adjustments were made for restaurant popularity, since a Tim Horton’s was used for the base-case (given a value of 1 for popularity), daily emissions could potentially be slightly skewed upwards for the more popular restaurants like McDonald’s.

The difference in observed emission rates between the studies is relatively small and appears logical when the influence of certain factors is examined. Shusas (2000) estimated the daily production of about 11 kg of non-carbon dioxide emissions by their studied drive-thru. In our study, it was found that the average production of non-carbon dioxide emissions was approximately 22 kg per day per drive-thru. The variation in observed daily production of these emissions is likely attributable our higher average daily idling time and the higher proportion of our sampled vehicles in the light truck and heavy truck categories (our study found that 42% of observed vehicles were light trucks, while only 32% of their observed vehicles were in this class). Since these vehicle classes produce emissions at a greater rate than cars, observing more of light and heavy duty trucks is likely to yield inflated emissions rates. Another contributor to the observed variation could be the choice of emission production rates used. Although the U.S. Environmental Protection Agency’s figures were used by both research groups, Shusas (2000) used those rates applicable for summer temperatures while researchers in this study used those adjusted for winter conditions. Since the emission rates are higher for summer than for winter, it is likely that this factor also accounts partially for the higher average daily emission production for each store.

Since Shusas (2000) did not account for the production of carbon dioxide, it is impossible to compare methods. If however the rate of carbon dioxide production used in this study is applied to the average daily idling time observed by Shusas (2000), we may predict that this McDonald’s emits approximately 145 kg of carbon dioxide daily. This is relatively close to our average estimation of 184 kg/drive-thru/day. The variation is again explained by a combination of the above factors and their lower average daily idling time.

6.2.3 Significance of These Emissions

Despite the limitations of this study, it is undeniable that idling vehicles at Edmonton fast-food drive-thrus contribute to local, provincial, and national annual total vehicle emissions. Edmonton fast-food drive-thrus contribute approximately 8600 tonnes of emissions (mainly CO2, CO, HC, and NOx) per year to the atmosphere accounting for approximately 0.1865% of Alberta’s annual transportation-related GHG emissions and 0.0037% of Alberta’s annual total GHG emissions. Edmonton’s residential transportation emissions, including idling activities, account for approximately 15.84% of Edmonton’s total annual GHG emissions (CO2RE 2005).

Commonly used Air Quality Indices report Edmonton’s air quality history as primarily “good” with occasional occurrences of “fair” (Alberta Environment 2005). As the city of Edmonton has not previously had air quality problems and emissions from drive-thrus do not seem to contribute a considerable amount to Edmonton’s total, it may be difficult to argue that action is needed to reduce drive-thru use in this city.

6.2.4 Negative Impacts of These Emissions

The argument in favour of policy-induced adjustments to drive-thru use is based upon the recognition that vehicle emissions can cause a variety of environmental problems. Climate change, an issue currently at the forefront of environmental policy, results primarily from the burning of fossil fuels, creating GHGs. Actions are being taken both on a provincial level, through Alberta’s climate change action plan, as well as on a national/international level through the Kyoto Protocol. Alberta Environment’s (2002) climate change action plan is directed towards reducing GHG emissions: “The Alberta government will reduce the GHG emissions intensity of its economy (emissions relative to GDP) by 50 per cent below 1990 levels by the year 2020. 60 million tonnes by 2020 is a translation of what that level of intensity reduction would mean in tonnes of carbon dioxide equivalent.” Completely eliminating Edmonton drive-thrus until the year 2020 would result in a 114,128 tonne reduction in emissions – keeping in mind pollutants other than those emitted from vehicles contribute to climate change. Maintaining the current level of drive-thru use does not help either Alberta or Canada in meeting their stated reduction targets (Government of Canada 2001).

As well as posing risks to the environment, vehicle emissions are known to have negative implications towards human health, including eye irritation, headaches, acute and chronic respiratory illness, and death (McCubbin and Delucchi 1999). McCubbin and Delucchi (1999) apply a dollar value to health risks caused by vehicle emissions in the United States. They estimate the upper limit cost per kilogram of motor vehicle emissions in 1990 for CO and NOx as $0.09 and $17.29, respectively. If we apply this value to our estimated production of NOx (approximately 38 kilograms per day), the cost of drive-thru emissions of this gas is approximately $250,000 per year. People living in residential areas near drive-thrus and fast-food employees who typically work the drive-thrus are particularly at risk for the above noted health risks.

6.2.5 Taking Action

In spite of the current low contribution of Edmonton drive-thrus to total city emissions, two main factors suggest that taking action today is justifiable. Firstly, the fast-food industry is continuing to grow as evident from the 125 new Tim Horton’s locations opened in Canada between 2004 and 2005. This increase brings the number of these stores in Alberta to 205 (Wendy’s International Inc. 2006). Comparatively, McDonald’s Canada has opened an average of 35 locations per year throughout the country since its emergence here in 1966 (McDonalds Canada, 2006). As actual yearly figures are not available to the public, it is difficult to determine modern trends in fast-food growth. It is defensible however, to assume that the trend of an increase in new franchise stores in the United Kingdom are likely to be experienced in Canada (Biz/ed, 2006). In the United States in the 1990s, an average of 1000 new McDonald’s stores appeared each year (Reuters Newsroom, 2002). Though it is unlikely that this rate of growth will ever be seen in Canada, recognizing the enormous potential of the fast-food (drive-thru) industry and its current rate of growth supports the conclusion that the issues and risks associated with idling vehicle emissions will only intensify unless action is taken.

The second factor justifying taking action to reduce drive-thru use is the fact that this service is not essential for the vast majority of the population. Eliminating or reducing this feature would not prevent individuals from consuming a meal of their choice. This policy would only alter the mode of consumption. Though it has been shown that drive-thru revenues substantially contribute to the revenues of many franchises it is unlikely that all current drive-thru users would simply stop consuming fast-food if this service was removed. The losses that would accrue due to this policy (though potentially minimal) should be compared to the benefits that would arise from less air pollution and lower GHG emissions to determine the true value of such a policy.

If the benefits of action are shown to outweigh its costs, there are three primary policy approaches Edmonton could take: completely eliminate drive-thrus, reduce the use of drive-thrus or do nothing. Banning drive-thrus is an unrealistic approach to reducing idling vehicle emissions. Not only would it likely result in lost profits to the fast-food industry, but it would likely result in foregone capital investment and be extremely unpopular with society. Reducing drive-thru use is more reasonable. Individuals can be deterred from idling in drive-thru line-ups through a variety of economic mechanisms such as a discount on purchases made inside and taxes on purchases made at the drive-thru. Equity issues arise when imposing the above taxes or ban. Such mechanisms penalize individuals for whom getting out of their vehicle is difficult – individuals with small children or pets, or individuals with disabilities. Education campaigns could serve as alternatives to economic measures. Since these programs, designed to generate voluntary emission reductions, are common (eg. One-Tonne Challenge) the introduction of new campaigns, specifically targeting drive-thru use, could be a more acceptable method of curbing this practice (Climate Change, 2006).

6.2.6 Future Research

Any further research into vehicle emissions from drive-thrus should begin with some of the limitations discussed above. Taking more samples at more fast-food establishments could greatly reduce some of the inaccuracies present in our data. Sampling all hours of the day, taking into account drive-thrus that are open 24hrs, would also provide more representative samples. Our study can be expanded to include a variety of other variables including: seasonality, vehicle characteristics such as make, age, and maintenance and more restaurant franchises.

Expanding our research and improving its accuracy will play a key role in determining the true costs and benefits of policies designed to reduce drive-thru usage. Establishing the physical amount of emissions produced each day by drive-thrus is a fundamental part of determining the above benefits, and thus, is a key component of performing an accurate cost/benefit analysis. Eventually, other costs of drive-thru usage, such as a reduction in transportation route efficiency and social implications, should be considered along with benefits of this service to businesses and users.

Though seemingly small when compared to other emission sources in this province, emissions from drive-thru lines appear more substantial when this practice is recognized as almost completely avoidable. Since the use of this service creates negative health and environmental effects, a review of the expansion of drive-thrus is certainly warranted.

7.0 CONCLUSION

Currently, average Canadian lifestyles are vehicle dependent. Many individuals face long commutes to work and as a result are constrained for time. The fast-food industry has responded to the demand for quick, convenient service by creating and expanding drive-thru operations. Fast-food restaurants, banks, and dry cleaners are just some industries that utilize drive-thrus. Although convenient, this service contributes to environmental and public health degradation as a result of the emissions created by idling vehicles.

Vehicle emissions are a major source of GHGs (CO, NOx, HC, and CO2), which not only contribute to such health risks as eye irritation, asthma, and even death, but are also linked to the growing climate change issue. 163 countries have ratified the Kyoto Protocol to combat climate change. Urging the public to reduce the use of their vehicles is one mechanism through which the Canadian government is attempting to meet their Kyoto commitments.

This study was conducted to achieve two objectives. The first objective was to estimate the number of cars using the drive-thru at a local Edmonton Tim Horton’s restaurant as well as the average time each vehicle idled in line at the drive-thru. The second objective was to estimate the total daily vehicle emissions produced by all fast food drive-thrus in the city of Edmonton. Several sources of error plague this study, the majority stemming from a lack of sampling effort. This leads us to caution readers that the findings of this investigation should be treated as estimates only. However, as this study has achieved its objectives, both characterizing the use of a single drive-thru location and estimating the amounts of certain emissions produced from city-wide drive-thrus, these estimations should not be ignored. It is recommended that further research be conducted to produce more conclusive results regarding the societal and environmental impacts of emissions from idling vehicles at drive-thrus. In this way, debate on potential policies to curb this practice may occur in an educated and fact-based manner.

8.0 REFERENCES

American Transportation Research Intitute. 2005. Compendium of Idling Regulations.

Available on-line at 2005.ATRI.Idlingcompensation.

Angotti, T., and E. Hanhardt. 2001. Problems and prospects for healthy mixed-use

Communities in New York City. Planning Practice and Research 16: 145-154.

Bedford P.J., and P.M. Dill. 2002. Proposal to amend the zoning by-law for the former

City of Toronto regarding development standards to address drive through

restaurants and other drive-thru operations. Available online at

welivehere.ca/interim%20control%20bylaw.pdf

Beydoun, M. and J.M. Guldmann (2006), ‘Vehicle Characteristics and Emissions: Logit

and Regression Analyses of I/M Data from Massachusetts, Maryland and

Illinois’,Transportation Research Part D 11, 59-76.

Bin, O. (2003), ‘A Logit Analysis of Vehicle Emissions Using Inspection and

Maintenance Testing Data’, Transportation Research Part D 8, 215-227.

Bishop, G.A. and D.H. Stedman (1996), ‘Measuring the Emissions of Passing Cars’,

Accounts of Chemical Research 29, 489-495.

Biz/ed. 2006. Business Profiles. McDonald’s Company Facts. Accessed April 25, 2006



Byrne, Patrick J., Oral Capps, Jr. and Atanu Saha. “Analysis of Quick-serve, Mid-scale and Up-scale Food Away From Home Expenditures.” International Food and Agribusiness Management Review 1(1)(1998): 51-72.

City of Edmonton, Transportation and Streets Department. 2006. 2005 Traffic Flow Map Average Annual Weekday Traffic.

City of Edmonton. Anti-Idling Infromation. Accessed March 10, 2006

)2)688762_0_0_18/ Anti-Idling+Information.htm

Climate Change. 2005. Taking Action on Climate Change. Government of Canada. Accessed on April 25, 2006.

Cummings, Claire Hope. “Entertainment foods (the health of individuals threatened by the spread of fast food culture).” The Ecologist 29.1 (Jan-Feb 1999): 16(4). 

Enns, P. (2000), ‘A Framework for Modeling In-Use Deterioration of Light-Duty Vehicle Emissions Using MOBILE6’, Journal of Transportation and Statistics 3(2), 39-

48.

EPA (1998), Emission Facts: Idling Vehicle Emissions. United States Environmental

Protection Agency: Office of Mobile Sources, , February 16.

EPA (1994), Automobile Emissions: An Overview. United States Environmental

Protection Agency: Office of Mobile Sources, , February 16.

Gauderman, J.J. 2005. Childhood Asthma and Exposure to Traffic and Nitrogen Dioxide. Epidemiolgoy. 16:737-743

Goldberg, D. 2002 Children, Youth and Families and Smart Growth: Building Family Friendly Communities. Network for Smart Growth. 9

Guildford, Roxanne. “America's Car Culture Collides with Bans on Drive-Thrus”. Christian Science Monitor 8/24/98, Vol. 90 Issue 189, p3, 0p,1c

Handy, S., S. Cao, and P. Mokhatarian. 2005 Correlation or causality between the built environment and travel behaviour? Evidence from Northern California. Transportation Research Part D:427-444

Harrington, W. (1997), ‘Fuel Economy and Motor Vehicle Emissions’, Journal of

Environmental Economics and Management 33, 240-252.

Hume, Christopher. “For a McHappy Ending: No New Drive Throughs” The Globe and Mail February 10, 2004

IPSOS World Monitor. “Fast Food Faces an Uphill Battle.” IPSOS World Monitor Research Alert. December 07, 2004.

Jekanowski, Mark D.. “Causes and Consequences of Fast Food Sales Growth”. Food Review. 22.1 (Jan. 1999).

Kirchstetter, T., B. C. Singer, R. A. Harley, G. R. Kendall and W. Chan (1996), ‘Impact

of Oxygenated Gasoline Use on California Light-Duty Vehicle Emissions’, Environmental Science and Technology 30(2), 661 -670.

Lama T., and D. Niemeierb, 2005. An exploratory study of the impact of common land- use policies on air quality. Transport and Environment 10:5 265-383

McCubbin, D.R., and M.A. Delucchi. 1999. The health costs of motor-vehicle-related air pollution. The Journal of Transport Economics and Policy. 33:253-386

McDonald’s Canada. 2006. Accessed on April 25, 2006.

McGuckin, Nancy and Nakamato, Yukiko. “Differences in Trip Chaining by Men and Women”. Conference on Women’s Travel Issues. (November 18-20 2004).

Natural Resources Canada. 2006. Idling and Climate Change Go Hand in Hand. Accessed April 17, 2006.

Reuters Newsroom. 2002. McDonald’s growth binge behind its woes. News for McDonald’s Franchise Owners. Accessed on April 25, 2006



Shoup D., 1999. The Trouble with Minimum Parking Requirements. Transportation Research Part A 33: 549-574

Statistics Canada

Stewart, Hayden. "Food Market Structure: Food Service." Briefing Room. (17 June 2005). United States Department of Agriculture. 20 Feb. 2006

Stewart, Hayden and Steven T. Yen. “Changing household characteristics and the away-from-home food market: a censored equation system approach.” Food Policy. 29.6 (Dec. 2004): 643-658.

Wang, Q., K. He, T. Li and L. Fu (2005) ‘Strategies for Controlling Pollution from

Vehicular Emissions in Beijing’, International Journal of Environmental Technology and Management 5(1), 87-106.

Wendy’s International Inc. 2006. Corporate and Investor Website. Restaurant Data Tim Horton’s. Accessed on April 25, 2006.



Wenzel, T., B.C. Singer and R. Slott (2000), ‘Some Issues in the Statistical Analysis of

Vehicle Emissions’, Journal of Transportation and Statistics 3(2), 1-14.

Appendix A

Sample Extrapolation Calculation

i) Determining popularity factor

Observed 32 minutes of idling time (IT) at Harvey’s between 11 am and 12 pm. Traffic density at this location is 24000 vehicles per week.

Based on base-case Tim Hortons, 315 minutes of IT are observed on average between 11 and 12 pm. 164 minutes are observed on average between 5 and 6 pm

IT that would have been observed if measurement had occurred between 5 and 6 pm = 32*164/315 = 16.7

Now can use correlation observed for 5-6 pm between traffic density and IT [IT=0.009(traffic density)

If the restaurant had been a Tim Hortons would have seen [24000*0.009] 216 minutes of IT

Popularity factor= observed IT(adjusted for time of day)/predicted IT for a Tim Hortons

= 16.7/216

= 0.077

The IT observed at a Harveys is 7.7% that observed at a Tim Hortons on a road of comparable traffic density.

ii) Determining daily IT for an unsampled restaurant

eg. Harveys located on a street with weekly traffic density of 19100 vehicles per week

Total daily IT= (1/proportion of total daily IT occurring between 5 and 6 pm) (Popularity factor) (0.009*traffic density)

= (1/0.036) (0.077) (0.009*19100)

= 368 minutes of IT/day

*Note, the first term must be included to expand the hourly IT value given by the last term to a full day. In this example the 5 to 6 pm period accounted for 3.6% of the total daily IT (as given by base-case analysis), therefore multiply by 1/0.036 to get a full day’s IT.

Appendix B - List of Tables and Figures

Winter Conditions (-1.1ºC, 13.0 psi RVP gasoline)

|Pollutant |Units |Light-Duty Gasoline Cars |Light-Duty Gasoline |Heavy-Duty Gasoline |

| | |(8501 lbs) |

|HC |grams/minute |0.3525 |0.512 |0.734 |

|CO |grams/minute |6.19 |8.12 |11.4 |

|NOx |grams/minute |0.103 |0.125 |0.196 |

Figure 1. Average rates of production for non-carbon dioxide emissions for three vehicle categories (EPA, 1998).

|Average Day |

| |Car |Light Duty |Heavy Duty |Total |

|Vehicle Numbers |576 |461 |48 |1085 |

|Vehicle Idling Time (hh:mm:ss) |47:34:47 |36:56:16 |3:52:46 |88:23:49 |

| | | | | |

Figure 2. Distribution of daily idling time and total number of vehicles by vehicle category for base-case Tim Horton’s.

[pic]

Figure 3. Proportion of each vehicle category passing through the drive-thru line at the base-case Tim Horton’s.

|Average Daily Emissions |

| |Cars |Light Duty Trucks |Heavy Duty |Total |

| | | |Trucks | |

|HC (g) |1006.3 |1134.7 |170.9 |2311.9 |

|CO (g) |17671.1 |17996 |2653.5 |38320.6 |

|NOX (g) |294 |277 |45.6 |616.6 |

|CO2 (g) |185275.2 |143835.2 |15106.1 |344216.5 |

|Total emissions (g) |204246.7 |163243.1 |17976 |385465.8 |

| | | | | |

Figure 4. Average daily emissions (grams) produced by the base-case Tim Horton’s.

[pic]

Figure 5. Coefficients for the linear regression of daily idling time on dummy variables for days of the week and times of the day.

Figure 8. Daily trend in drive-thru usage at the base case Tim Horton’s. Observed hourly idling times are averages from all sampled times within a three week period (individual values are averages of two to seven observations—each one hour long).

[pic]

Figure 9. Proportion of non-carbon dioxide emissions produced by each vehicle type per day in Edmonton.

|Restaurant |Popularity |

|Harvey's |7.7% |

|Arby's |6.5% |

|McDonald's |46.8% |

|Kentucky Fried Chicken |12.3% |

|Burger King |3.2% |

|Taco Bell |7.8% |

|Dairy Queen |31.5% |

Table 1. Drive-thru popularity, as measured by relative daily idling times, of seven fast-food restaurant chains in Edmonton. Figures are presented as a percent of the popularity of Tim Horton’s.

|Vehicle Type |Percent of Total Vehicles Observed in |Percent of Total Daily Non-CO2 Emissions |

| |Drive-Thru Lines | |

|Car |54% |46% |

|Small Truck |42% |47% |

|Large Truck |5% |8% |

Table 2. Comparison of the frequency of use of drive-thru lines by each vehicle type and its contribution to total daily non-CO2 emissions. Note that data on CO2 emissions was not-available by vehicle category and thus, production rates were applied equally to all categories

-----------------------

[1] Medians are present instead of mean values because emission distributions are strongly skewed to the right.

[2] Note that the average failure rate of the three analyzed states was 8.6%.

-----------------------

Figure 6. Proportion of drive-thrus included in study belonging to each restaurant chain.

[pic]

Figure 7. Observed correlation between traffic density and hourly idling time for five different Tim Horton’s locations between 5 and 6 pm. Line represents a linear relationship forced through the origin.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download