|
|
| Fig. 1: Bombardier R62A "1" Train arriving into 207th Street, New York (2022). (Source: Wikimedia Commons) |
Cities like New York already have extensive subway and bus networks, yet a large share of trips still happen by private car. Fig. 1 depicts a New York Subway train and the scale associated. Looking at such a large vehicle, it is understandable as to why it is widely assumed that transit through such mediums is more sustainable, but that only becomes meaningful when stated in joules per passenger-kilometer traveled. This report asks a focused quantitative question: at what average occupancy do electric mass-transit modes (buses and urban electric rail) use fewer joules per passenger-kilometer traveled than private cars, and what is the rough energy saving if a city like New York shifts a fraction of its car travel to transit?
To keep the analysis clear and within reasonable length, we rely primarily on published life-cycle assessments that report operational energy intensities of different modes, especially Chester and Horvath's comparison of U.S. passenger transport. [1] The goal is not to reconstruct every unit conversion but to put a few robust numbers on three things: the typical energy per passenger-kilometer traveled for cars, buses, and urban electric rail; simple occupancy break-even points where transit beats cars; and an order-of-magnitude estimate of annual joule savings for a New York-style city.
Chester and Horvath perform a detailed life-cycle assessment of U.S. passenger modes, including direct fuel or electricity use and supporting infrastructure. [1] For the purposes here we focus on the operational energy numbers, since these dominate in day-to-day planning. For a typical gasoline automobile, they report operational energy intensities of about 1.9-3.5 MJ per passenger-kilometer traveled (PKT) over realistic occupancies and driving conditions in Fig. 1 of the Chester and Horvath report (based on the active and inactive vehicle operation consumptions). [1] We will assume a sedan, thus taking 1.9 MJ/PKT as a representative value for a modern gasoline sedan with average U.S. occupancy. For urban buses, Chester and Horvath find operational intensities typically in the 0.5-4 MJ/PKT range, depending strongly on load factor and whether it is peak or non-peak times in Fig. 1 of the Chester and Horvath report. [1] A heavily loaded bus in peak hours can be near the bottom of that range; off-peak, with only a handful of riders, it can be closer to, or even worse than, a car. Mahmoud et al.s review of electric bus deployments shows per-vehicle electricity use on the order of 1-1.5 kWh per km, consistent with the lower end of values reported in field and test-cycle studies of battery- electric buses, which corresponds to a few MJ/km of delivered energy. [2] Combined with realistic urban occupancies, this is consistent with the 0.5-4 MJ/PKT range from Chester and Horvath. For calculations, we will assume peak hours as that is when passenger cars would usually be driven. Including a conservative estimate, 1 MJ/PKT will be used.
For electric urban rail (subways and metros), Chester and Horvath report operational energy intensities in the range of roughly 0.3-0.6 MJ per passenger-kilometer traveled for dense, high-ridership systems (fig. 1). [1] These values already reflect the fact that a single train can carry hundreds of passengers, even if not every car is full all the time. For conservative estimation, we will go slightly above the midpoint of 0.45 and use 0.5 MJ/PKT for calculations.
Putting this together, a simple typical set of working values is shown in Table 1:
|
||||||||
| Table 1: Typical operational energy intensities. [1] |
These are deliberately rounded to keep the discussion transparent and emphasize orders of magnitude rather than fine details.
The central physics of occupancy is as follows: energy per PKT = (energy per vehicle-kilometer traveled (VKT)) × (average number of passengers).
If the energy per vehicle-kilometer traveled is fixed by the technology, the only lever left is how many people share that energy cost. For an electric bus, Mahmoud et al. summarize real-world energy use in the neighborhood of 1.2 kWh per km for large urban buses. [2] This is about 4.3 MJ per vehicle-kilometer traveled. To match the cars 1.9 MJ/PKT, the bus must carry
| occupancy | = | 4.3 MJ/VKT 1.9 MJ/PKT |
≈ | 2.26 passengers |
To match the subway-like value of 0.5 MJ/PKT, it must carry
| occupancy | = | 4.3 MJ/VKT 0.5 MJ/PKT |
≈ | 8.6 passengers |
These are not extreme numbers. A bus with only three people on board is already roughly competitive with a typical car in energy per passenger-kilometer traveled. A bus with a modest crowd of nine or more passengers is in the same energy-efficiency league as high-ridership electric rail on a per-person basis.
To sketch an order-of-magnitude estimate for a New York-style city, we assume that:
10⁶ passenger-kilometers per year of travel that could plausibly be made either by car or by transit (this is a conservative scale for a large metro region, might be more like a slice of New York).
20% of that passenger-distance (2×10⁵ PKT/year) is shifted from cars to transit.
If the car uses 1.9 MJ/PKT and the subway uses 0.5 MJ/PKT, the savings per passenger-kilometer traveled shifted is ΔE ≈ 1.9 − 0.5 = 1.4 MJ/PKT. Multiplying by 2×10⁵ PKT/year yields ΔE ≈ 2.8×105 MJ/year = 2.8×1011 J/year. In energy-industry terms, that is close to 0.08 GWh per year. A larger shift, or a larger underlying travel volume, pushes this into multi-GWh territory. The point is not the exact number but the scale. After all, if a slice of 10⁶ PKT can save on the magnitude of 1011 J/year, shifting a modest fraction of urban travel from cars to well-used electric transit is certainly capable of moving 1015–1016 joules per year, not just rounding-error amounts.
In a less transit-dominant city like San Francisco, the same logic applies but the numbers shift for two reasons. First, total passenger-kilometers traveled are smaller, so the maximum possible savings are lower. Second, some bus routes and light-rail lines may run with lower occupancies than New York's subway, which pushes their energy per passenger-kilometer traveled up toward the car range. Chester and Horvaths' ranges already hint at this: the high end of bus energy intensity overlaps the low end of car intensity. [1] Future work could take agency-specific data for Bay Area systems and plug them into the same simple occupancy framework used here. Furthermore, life-cycle studies of high-speed rail, such as Chester and Horvath's analysis of Californias proposed system, imply sub-MJ-per-PKT operational intensities when their reported 13-36 kWh per train-kilometer propulsion values are combined with realistic load factors (hundreds of passengers per train). [3] This suggests that, at good load factors, inter-city high-speed trains live in the same energy-efficiency band as urban metros and substantially beat both cars and short-haul aircraft per passenger-kilometer traveled. Therefore, extending the calculations in this report to a New York-Boston or Tokyo-Osaka corridor, using published HSR energy- intensity values, would be a good direction for possible future research and work in this area.
Overall, this report has shown that a typical gasoline car in the United States occupies an energy band around 1.9-3.5 MJ per passenger-kilometer traveled, while urban electric rail in dense, high-ridership systems lies closer to 0.3-0.6 MJ per passenger-kilometer traveled. Electric buses, with realistic VKT energy use, beat cars as soon as they carry more than two passengers and approach subway-like efficiencies once they have about nine or more riders. For a New York-scale city, shifting even 20% of car passenger-kilometers traveled onto well-used electric transit is capable of saving on the order of 1011 joules per year. The key control knob is not exotic new drivetrains but occupancy: how many people share each VKT. Viewed through the lens of joules per passenger-kilometer traveled, getting people on the train is not just a slogan; it is a simple way of exploiting the physics of shared motion and highlights how modest changes in mode choice can yield substantial energy savings at the urban scale.
© Galen Xia. The author warrants that the work is the author's own and that Stanford University provided no input other than typesetting and referencing guidelines. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.
[1] M. V. Chester and A. Horvath, "Environmental Assessment of Passenger Transportation Should Include Infrastructure and Supply Chains," Environ. Res. Lett. 4, 024008 (2009).
[2] M. Mahmoud et al., "Electric Buses: A Review of Alternative Powertrains," Renew. Sustain. Energy Rev. 62, 673 (2016).
[3] M. Chester and A. Horvath, "High-Speed Rail With Emerging Automobiles and Aircraft Can Reduce Environmental Impacts in California's Future," Environ. Res. Lett. 7, 034012 (2012).
