The report Life-cycle Environmental Inventory of Passenger Transportation in the United States bills itself as “the first comprehensive environmental life-cycle assessment of automobiles, buses, trains, and aircraft in the United States.” The report, by Mikhail V. Chester of the Institute of Transportation Studies at Berkeley, goes far beyond counting the fuel consumed by vehicles. It considers the energy and materials used to build stations
, terminals, roadways, runways, tracks, bridges, tunnels and parking, as well as maintenance, heating, lighting and more. A full life-cycle accounting of travel modes has been a long time coming; it is critically needed and tremendously welcome.
The findings on life-cycle energy use are summarized in this chart:
Read on for a summary of the findings, and a discussion of how the results are affected by urban design context.
The chart information is duplicated in the table below, with figures and information links for each mode.
|Vehicle / System||Energy use||Percent for construction & etc.||Comments|
|Bus, peak||1.1||27.3||40-foot vehicle|
|California High Speed Rail||2.0||35.0||planned|
|BART||2.2||50.0||Bay Area subway|
|Caltrain||2.3||52.2||Bay Area commuter rail|
|Green Line||2.3||62.2||Boston light rail|
|Boeing 747||2.8||17.9||long-haul aircraft|
|Boeing 737||3.0||13.3||medium-haul aircraft|
|Muni||3.0||60.0||San Francisco light rail|
|Embraer 145||4.2||16.7||short-haul aircraft|
|Ford F-series||7.9||27.8||pickup truck|
|Bus, off-peak||8.8||27.3||40-foot vehicle|
One feature that stands out is the wide range in bus performance, depending on how full the bus is. During peak hours the bus ridership is assumed to be 40 passengers and energy efficiency beats all other modes handily. During off-peak hours ridership is assumed to be five passengers and energy efficiency is the worst of any mode.
In fact, the efficiency of all modes depends on how full the vehicles are. The study included a calculation of low and high ridership for each mode so that the energy use ranges could be compared:
What’s interesting here is even when the transit systems are assumed to have low ridership, they all perform better than the car and light trucks do with their average number of passengers. Similarly, when the California High Speed Rail system is assumed to have low ridership, it still performs better than the SUV or pickup with an average number of passengers.
Another aspect of the results is that the transit systems have a common characteristic. A large percentage of their lifetime energy use goes into construction, particularly the construction of underground stations built of concrete (concrete is very energy-intensive to make). This points up at least one advantage of simpler, surface-running transit: It does a better job of saving energy, compared to elaborate transit systems with large underground stations.
The role of context
As welcome as the analysis is, it has several shortcomings that derive from an important absence: There is no consideration of urban context. Context is essential to energy and pollution impacts, especially for a report that considers 50- and 80-year time horizons.
Looking at the life cycle energy results, one might be tempted to say that SUVs use three times more energy to move riders than the BART system. But modes are not directly comparable in that way, because urban context — the way street networks and land uses are configured — plays a large role in transportation energy use.
An APTA study on transit and land use quantified that effect. APTA found a “primary effect” of energy savings when people switched from cars to transit. But APTA also found a “secondary effect” of transit enabling built environments where people drove less, walked more, and used transit more. The secondary effect was twice the magnitude of the primary effect.
In other words, switching from an auto trip to a transit trip saves a certain amount of energy, but the presence of transit itself creates efficient neighborhoods that save twice as much energy. The 50-year time frame is certainly long enough for these secondary effects to develop.
Context and health
In the results of “Life-cycle Environmental Inventory,” greenhouse gas emissions (carbon dioxide, nitrous oxide, methane) closely track energy use. As a rough estimate one could say they are interchangeable. The report also calculates criteria pollutants emitted by each mode (carbon monoxide, sulfur dioxide, nitrogen oxides, volatile organic compounds, particulates). Generally speaking, however, pollutants are not interchangeable.
The harmful impacts of pollutants are felt at different scales and on different time frames. For example, carbon dioxide has global impacts that appear over centuries. Sulfur dioxide has regional impacts that appear over days and weeks. Particulates have extremely local impacts that are most powerful on the time frame of seconds and minutes. Pollutants combine and are transported in complex ways; the design of urban spaces makes an enormous difference for human exposure. This is another example of context influencing emissions.
Public health advocate Wig Zamore responded to “Life-cycle Environmental Inventory” with this comment:
The criteria pollutants used in the thesis are a good indicator of traditional regional air pollution impact. They will not help much though with the near-source air pollution impacts which have the most serious health consequences and are more akin to occupational exposures. …
There is an interesting source and receptor temporal symmetry with transportation pollution. If the pollutant’s travel time to a person is more than three minutes from the exhaust pipe, most of the health risk is greatly reduced. Transportation emissions are not only diluted and dispersed fairly rapidly, but they evolve even more rapidly. Fresh mobile air pollutants evolve furiously in the first three seconds and subsequently into much less dangerous size, composition and concentration the first three minutes after exhaust.
You can fairly easily get an intuitive handle on the scale of public health impacts of “near source” transportation pollution through consideration of four factors.
- How big is the source, as in roadway vehicles per day — VPD data is commonly available for major roadways and being next to 50,000 VPD is much worse than 5,000 VPD.
- How close are people to the source — those living within 100 meters of major congested highways or city street canyons can have occupational-scale exposures similar to long-haul truckers, urban delivery van drivers, or diesel rail engineers.
- How much of their lives do people spend in the most exposed locations — living next to a highway, going to school next to one, or working full-time in the middle of a transportation-intense environment will be worst.
- What other factors qualify the build-up and retention of pollutants that people are exposed to — as in tunnels, street canyons, local geographic traps, peak hour congestion, unfavorable meteorology, etc.
Sometimes we forget that people’s biological viability starts to be threatened within 3 weeks without food, 3 days without water and 3 minutes without air. Yet of the three, our food and water supplies are much better protected and the subject of some personal choice no matter where one lives and works. That is less true of air, and unfortunately the least-advantaged members of society tend to live and work in the highest mobile emissions locations.
This means, for instance, that a ton of particulates released from a remote concrete manufacturing plant has very different health impacts than a ton of particulates released on an urban arterial roadway with children and elderly people living close by. “Life-cycle Environmental Inventory” recognizes that in introductory paragraphs (p. 2, 234), but it lumps together pollutants regardless of where or how they are emitted.
In addition to calculating the amount of pollutants emitted from a travel mode, one also must consider the context in which pollutants are released. The report identifies this as an area of future study:
The identification of major energy and emission components serves as a tool to tackle specific issues related to the quality of performance both geographically and temporally. While emissions of a particular pollutant may be larger for the vehicle operation component in an inventory, other component emissions may be released closer to population centers resulting in a greater likelihood of exposure … The disaggregation of components presents a founding inventory which could be used in impact assessment frameworks to determine where the greatest risks exist. Additionally, the temporal aspect is important in understanding the duration of exposure for a particular population.
— Life-cycle Environmental Inventory, p. 265
The “Life-cycle Environmental Inventory” report provides critically-needed, important and useful results. Those results are best used as initial inputs for a more complete understanding of the relationships between transportation, sustainability, and urban places. The report provides a starting foundation for investigations that consider a full spectrum of time horizons and situational characteristics such as urban design context.
Chester, Mikhail V., Life-cycle Environmental Inventory of Passenger Transportation in the United States. Institute of Transportation Studies, University of California Berkeley, August 1, 2008. The author’s website includes related presentations, news coverage, and previous draft versions.
Aurbach, Laurence, APTA Study on Transit and Land Use. Ped Shed Blog, March 12, 2008.
Baldauf R., Traffic and meteorological impacts on near-road air quality: summary of methods and trends from the Raleigh Near-Road Study. Journal of the Air & Waste Management Association, July 2008; 58(7):865-78.
Baldauf, R. et al, Traffic Emission Impacts on Air Quality Near Large Roadways. TRB Transportation, Land Use, Planning and Air Quality Conference, Orlando, FL, July 9-11, 2007. Transportation Research Board of the National Academies, Washington, DC, 2007.
The Broader Connection between Public Transportation, Energy Conservation and Greenhouse Gas Reduction, by Linda Bailey, Patricia L. Mokhtarian, Ph.D. and Andrew Little, prepared by ICF International, Inc. with funding from Transit Cooperative Research Program of the Federal Transit Administration. March 2008.
Williams-Derry, Clark, Transit: A Full Ride. Sightline Daily Score Blog
, October 30, 2008.
Zhou, Y. and J. Levy, Factors influencing the spatial extent of mobile source air pollution impacts: a meta-analysis. BMC Public Health. May 22, 2007; 7:89.
Are the off-peak rider counts calculated for the private auto? In a 24 hr period, a bus might have 4 hrs of 0 riders, 16 hrs with 5, and 4 hrs with 20 for 160 rider-hrs/day. Where the auto might have 4hrs with 1 rider, and 20 hours with 0 passengers (the auto’s off-peak), for a rate of 4 rider-hrs/day. In other words, the (intuitively obvious) waste of the private auto is that it sits doing nothing for about 20 hours a day.
Well, but the difference is when the auto is “off-peak” by your definition, it is not moving. So it is not consuming energy. Stationary vehicles do emit pollutants (evaporative emissions) and the report does include those.
There is however another important aspect of cars, trucks and parking, and that is the sheer amount of land that is consumed and paved over. Private vehicles occupy more land area than transit on a per-passenger basis. Each U.S. passenger vehicle is estimated to require 4-8 parking spaces — that’s a billion or more parking spaces.
The U.S.’s 39 million acres of pavement have huge environmental effects, not only in the loss of local farms, parks, wilderness and species, but also in heat island effects, polluted runoff, and soil erosion. All of that is outside the scope of the report, but still must be included in any complete environmental accounting.
Actually, this report includes the energy cost of constructing and maintaining the parking, highway, and roadway infrastructure required to support the private automobile. It also includes the energy cost of constructing, repairing, and operating private vehicles.
The key observation, made above, is that looking at things in terms of “passenger miles traveled” misses the fact that the nature of the dominant mode itself affects the number of miles traveled.
I’m curious to know how a bicycle would fit into this equation.
I’m not sure if a rigorous life cycle analysis of bicycles and associated infrastructure has ever been done. But Umbra Fisk at Grist magazine gives it a try. That calculation uses figures from a Pacific Institute report on the energy efficiency of walking. And there are additional discussions of these topics in two Sightline Daily posts: Planes, Trains, and Automobiles and Walking: Still Better Than Driving.
With respect to the efficiency of walking and biking, a lot depends on the foods one eats and how much energy is used to produce and ship those foods. As a broad generality, the various calculations find that walking is approximately five times more efficient than driving, and biking is about twice as efficient as walking (or ten times as efficient as driving).