Last month, the University of Toronto (UofT) hosted a webinar titled The Paradox of Net-Zero: Cities, Transportation and Health. The session was part of UofT Engineering’s monthly Skule Lunch and Learn series, which offers alumni the opportunity to keep up-to-date with broader world issues. December’s presentation was conducted by Marianne Hatzopoulou, Professor in Civil and Mineral Engineering at the University of Toronto and Canada Research Chair in Transportation and Air Quality.
Hatzopoulou leads an active research group that studies the interactions between transportation, built environments, air quality, climate and public health. She works closely with epidemiologists in the development of improved measures for air pollution exposure, and has received funding from federal and provincial health agencies to conduct integrative research in transportation engineering, air pollution and public health. Hatzopoulou also serves on the Transportation Research Board of the National Academies of Sciences, Engineering and Medicine, and is the associate editor of the journal Transportation Research Part D: Transport and Environment.
An Integrated Approach
Hatzopoulou kicked off December 2’s interactive Lunch and Learn session with the question, “What is the future of transportation?” Participants responded with a range of adjectives: affordable, shared, connected, accessible, electric, safe, efficient, health-promoting, decarbonized, automated, equitable. It was a bold vision—but were all transportation innovations guaranteed to mitigate climate change and improve public health?
As Hatzopoulou progressed with her presentation, she discussed the importance of considering transportation technologies using an integrated approach. One factor to examine would be carbon footprints, and how emissions would propagate in an urban environment. Another aspect would be air pollution and other environmental stressors to which pedestrians would be exposed. A portion of the public may be engaging in physical activity—so exposure models would be involved in addition to the environmental models. How could transportation policies be tested with respect to health?
Current Air Pollution Statistics in the GTHA
According to Hatzopoulou’s research, which was focused on the Greater Toronto and Hamilton Area (GTHA), transportation ranks second in terms of greenhouse gas (GHG) emissions when put in context of other sectors. Around 30 percent of GHG emissions are associated with transportation; this translates to 34,340 tons of CO2 equivalents generated per day for the GTHA alone.
Of all GHG emissions, 76 percent are produced by cars and SUVs. Only three percent of emissions come from public transit buses, and the remaining 21 percent are a result of commercial trucks.
Another pollutant to consider is black carbon, which is generated mostly by diesel engines. In the GTHA, 71 percent of black carbon emissions come from trucks—in comparison to 19 percent from cars and SUVs, and 10 percent from transit buses.
Yet another pollutant in the mix is nitrogen oxides (NOx), which are linked to numerous epidemiological issues when considering the impact of air pollution on health. According to Hatzopoulou’s findings, 52 percent of NOx emissions are generated by trucks, 36 percent by cars and SUVs, and the remaining 12 percent by public transit buses.
Rather than quantifying the health effects of these transport-related emissions through links to respiratory diseases, Hatzopoulou’s research group chose to look at premature mortality.
“In the world of epidemiology, there’s strong evidence linking long-term or chronic exposure to air pollution with dying sooner,” explained Hatzapulou. “On a personal level, you might die half a day sooner because of your air pollution exposure—so these are very small effects. But at the level of a population—and this is why it’s important from a public health point of view—it becomes really important to look at these things. So we chose premature death, keeping in mind that we’re very conservative because there are so many other health effects associated with transportation.”
When converting transport-related emissions to health effects, 332 premature deaths by cars and SUVs, 407 premature deaths by trucks, and 143 premature deaths by public transit buses were found per 100,000 people across the GTHA. The majority of the health burden was being incurred by people residing in hotspots close to the airport, and along a corridor or major arterial road.
Electrification and Renewal
Hatzopoulou’s research group went on to study the benefits of renewing these fleets of on-road vehicles.
According to their analysis, electrification of 100 percent of cars and SUVs would result in a 70 percent drop of GHG emissions. This would translate to a prevention of 313 premature deaths, which would generate USD$1.9 billion (CAD$2.4 billion) a year in social benefits. In fact, every electric car deployed would be associated with USD$7,720 (CAD$9,850) in social benefits based on the health burden that would be alleviated.
Hatzopoulou’s second scenario studied the impact of electrifying 100 percent of public transit buses. While the dent in terms of GHG emissions was not significant, air quality gains were still encouraging—especially in local hotspots like Toronto, which had the highest density of transit buses.
Finally, Hatzopoulou’s research group assessed the impact of having cleaner trucks. Rather than looking at the electrification of trucks, Hatzopoulou opted to go conservative and check what would happen if subsidies were incorporated for renewing trucks to 2008 models and newer—even if the trucks remained diesel. It was found that each cleaner truck would be associated with a reduction of 275 premature deaths, and gains of USD$241,300 (CAD$308,000) worth of social benefits over the lifetime of every truck.
Autonomous Vehicles and Air Pollution
In the next portion of her webinar, Hatzopoulou discussed what autonomous vehicles (AVs) would mean for emissions and sustainability. How would owning an AV change the behavior of individuals, and what would happen if families had AVs in their life?
Using data from GTHA’s Transportation Tomorrow Survey, Hatzopoulou first isolated households that had their own gasoline-fueled cars—i.e., households that did not need to take transit. Her transport research group designed an algorithm that would assess the impact of owning an AV on a family’s GHG emissions. The results were surprising.
“The first thing we noticed was a reduction in car ownership,” stated Hatzopoulou. “Instead of owning three cars, the family could get by with two cars or sometimes even one car. The AV would drive around members of the household in an optimal fashion. However, we found that daily greenhouse gas emissions overall increase significantly—by 51 percent. So yes, sure, you may be dropping a car, but you have an increase in the total amount of greenhouse gas emissions!”
The reason? Induced trips.
“While the AV is fulfilling all of these different trips, it still has to move from one person to another,” described Hatzopoulou. “And so, there is all of the empty mileage that is occurring within these households. What they end up with is this car running empty with 51 percent extra time to actually fulfill all of these trips.”
When considering households that own a car but also take transit, the addition of an AV was found to double GHG emissions due to transit trips becoming replaced by the AV’s empty kilometers.
The impact of connectivity and automation is another angle to explore when it comes to AVs. For their research, Hatzopoulou’s group partnered with the LiTrans lab at Ryerson University, which was working on an algorithm called End-to-End Connected Automated Vehicles (E2ECAV).
In a small portion of downtown Toronto, an AV would approach an intersection—which would have information of the queue length—and a final destination would be set. From the point of view of traffic throughput, clearance time of the network would be observed at different penetrations of zero, five and 100 percent of these connected autonomous vehicles (CAVs) at different traffic congestion levels. While it took 6,000 seconds to clear the network when considering human-driven vehicles, it took CAVs half the time to clear the network; i.e., traffic throughput increased with AVs.
Since people were moving faster and were stuck less in congestion, there were similar drops in GHG emissions, which are related to fuel consumption. However, NOx and other air pollutants did not demonstrate significant decreases—and in fact, increased in some cases. How could this be happening?
“We know that air pollutants are highly related to how people drive—how they accelerate and decelerate,” detailed Hatzopoulou. “The fact that the system was making everyone reach the intersection and then turn left and turn right at every single point, was generating a lot of accelerations and decelerations across the system that were resulting in air pollution. In our analysis, we always look at the different facets of whether the parameters are skewed towards eco driving versus aggressive driving. You’re going to see different stories in both cases. It gets more interesting when we start looking at the dispersion of these emissions.”
Hatzopoulou studied tailpipe emissions through dispersion modeling using a street canyon model (a parametrized computational fluid dynamics model). It was found that air pollution from roads often generates large turbulent eddies that become trapped between tall buildings.
“Everything that is emitted on Bay Street, stays on Bay Street, because the buildings are so tall,” asserted Hatzopoulou. “It just continues to get recirculated, and very little actually escapes to affect the rest of the city.”
To summarize, in the 100 percent CAV scenario, even though the network was clearing faster, a higher maximum of air pollution was being created. While algorithms were optimizing traffic, they were not necessarily optimizing emissions or air quality, due to the generation of local hotspots of air pollution.
Real-World Emissions
Hatzopoulou’s research group decided to measure real-world emissions based on traffic conditions. In one particular study, the researchers distributed GPS devices to a large number of drivers across the region. Using information from multiple weekday and weekend trips, drive cycles were created to represent what driving looked like in the GTHA. The data was used to determine emission levels as well as fuel efficiencies for different vehicles, to assess whether cars being driven along the region’s congested roads were meeting fuel efficiency ratings that were being advertised by auto manufacturers. The short answer? No.
In comparison to U.S. Federal Test procedures for vehicles, drive cycles for Toronto were found to be far below fuel efficiency standards. (From the areas that were evaluated, the only one worse was the New York City drive cycle.) Therefore, when it comes to highway fuel efficiency, drivers in Toronto are unable to meet manufacturers’ ratings because of the way the road network is designed, with its levels of congestion.
Hatzopoulou went on to study the emissions of ultra-fine particles less than a hundred nanometers in diameter, which are not regulated in the U.S. or Canada despite being linked to numerous health effects. (In Europe, these nanoparticles have to adhere to an emission standard called the Euro 6 PN limit.) Based on the results on Hatzopoulou’s analysis, the drive cycles for Toronto and New York City were found to be above Europe’s particle emission limits. If the U.S. or Canada were to impose a standard similar to the Euro 6 PN limit, the vehicle would fail—and not even because of its design.
Going deeper into the effect of driving behavior on emissions, the GPS data demonstrated that aggressive driving generated significantly higher levels of air pollution. Adhering to lower speed limits would improve air quality.
“Aggressive driving is really related to aggressive acceleration, deceleration,” said Hatzopoulou. “Jerky movements are the ones that are generating the highest amount of emissions.”
Hatzopoulou’s research team used a Nissan Rogue to test for ultra-fine particle emissions, for which technology had previously not been available.
“Ultra-fine particle emissions are hard to regulate because science hasn’t yet indicated what is a safe level,” illustrated Hatzopoulou. “But then there are all these epidemiological studies that are linking the particles with brain effects because they’re so small that they can cross the blood-brain barrier.”
Hatzopoulou tested the Gasoline Direct Injection (GDI) vehicle using a Portable Emissions Monitoring System (PEMS), which would take a sample of the exhaust through analyzers placed in the back of the car. When assessing for ultra-fine particles, spiky behavior was observed.
“The car would either be generating a spike, or generating no ultra-fine particles at all,” revealed Hatzopoulou. “You were either in zero, or you were in a spike. So we developed a spike detection algorithm and started analyzing the amount of emissions that are generating these spikes. We noticed that 10 percent of the time accounted for 71 percent of the total emissions, and 20 percent accounted for 90 percent. The spikes were actually associated with aggressive behaviors such as harsh acceleration and braking.”
Final Takeaways
There are many conclusions based on Hatzopoulou’s evidence. Behavior modification can lead to a measurable impact on the road. Adjusting driving behavior can result in similar reductions to GHG emissions as electrifying vehicles. There are implications to new technologies such as AVs.
“Unfortunately, our analysis has to be very layered and look at multiple dimensions,” said Hatzopoulou. “I know it makes decision-making ever more uncertain and complex, but we ought to be able to analyze and estimate these different dimensions.”
Hatzopoulou is confident that reducing aggressive driving in instrumental in the management of road networks. Considerations range from the geometry of roundabouts versus intersections, to the timing of traffic lights. There is much scope in optimizing traffic and transportation systems to be intelligent, while using multiple lens and objectives to reach an integrated solution.
Hatzopoulou believes that there is no future without electrification; however, she does not promote electric vehicles as the only solution.
“I’m not saying everyone should throw out their old car and buy an electric vehicle without considering alternatives like public transit,” maintained Hatzopoulou. “If you look at our public transit system in the GTHA, the average CO2 emission per passenger kilometer traveled is about 20 g, using your good old diesel buses. Compare that with driving a gasoline vehicle in Toronto; your average CO2 emissions are about 300 g per passenger kilometers traveled—not 250 like in other cities, because we have a very congested network. Compare that with electric vehicles. With our current electricity mix—and considering that we have a range of performance levels for electric vehicles—you’re somewhere between 10 to 15 g per passenger kilometer traveled for electric vehicles. So transit, even with your good old diesel buses, is quite competitive. If you think about greening transit even further and increasing ridership, you’re going to be on your path to decarbonization.”
During the Q&A session at the end of the webinar, Hatzopoulou was asked which zero-emissions technologies she foresaw as the most sustainable for transportation in the long term: electric, hydrogen, renewable, etc.
“At a massive scale—at the scale of countries and regions—none of them is going to save us,” asserted Hatzopoulou. “When you look at the massive deployment of this technology, they’re all going to have trade-offs. We have to just drive less. We have to consume less cars and switch our mobility needs to other modes of transportation. There is no way we’re going to be able to achieve more sustainable outcomes without that. And the one topic I’m not discussing today is the impact of active transportation such as walking and cycling on your health. That is also measurable and quite significant as well.”
The next Skule Lunch and Learn session will be held on January 13, 2021. It will be covered by UofT Professor Joyce Poon, who will discuss how innovations in internet technology are creating the potential to map the human brain. To attend the free webinar, register here.