Don’t rule out elevated rail in cities

Toronto is looking to Honolulu for transit inspiration – looking to tap into the potential for elevated rapid transit to improve the city’s transit expansion plans. However, key city officials are extremely concerned about the impacts of elevated transit to the city. Skepticism is good, any may be required to ensure that elevated rail is successfully integrated into an urban environment, but it shouldn’t be an automatic disqualifier for the kinds of improvements that make rapid transit possible. From the Toronto Star:

Toronto chief planner Jennifer Keesmaat cites the shadow that a structure like the [elevated Gardiner expressway] casts on the street below. She also brandishes one of the chief arguments for building Toronto’s LRTs in the first place.

“From a land use planning perspective, if our objective in integrating higher order transit into our city is to create great places for walking, for commerce, living,… elevated infrastructure doesn’t work so well for any of those objectives,” she said.

It’s true that making elevated rail work in urban areas is a challenge, but it shouldn’t be so easily dismissed. Of particular concern is the willingness to equate the visual impact of the six-lane Gardiner Expressway with a potential two-track elevated rail structure. The other key concern is the equivocation of grade-separated transit with at-grade light rail.

Toronto seems full of transit terminology confusion these days. Embattled Mayor Rob Ford has been pushing for subways as the only kind of transit that matters (SUBWAYS SUBWAYS SUBWAYS!) regardless of context or cost. Meanwhile, the transit agency is looking to implement a ‘light rail’ project that features full grade separation and an exclusive right of way – in other words, a subway. Ford opposes the light rail plan in favor of an actual, tunneled line with fewer stations and higher cost. Much of the rhetoric seems focused on equating light rail with Toronto’s legacy mixed-traffic streetcar network.

However, just as Ford’s dogmatic insistence of subways at any cost is irresponsible, Keesmaat’s suggestion that at-grade LRT can accomplish the same transit outcomes as grade-separated LRT can is equally misleading. Remember the differences between Class/Category A, B, and C right of way (from Vukan Vuchic, summarized here by Jarrett Walker), paraphrased here:

  • Category C – on-street in mixed traffic: buses, streetcars, trams, all operating in the same space as other street users.
  • Category B – partially separated tracks/lanes: exclusive right of way for transit, but not separate from cross-traffic. Vuchic dubs this “Semirapid Transit.” often seen with busways or light rail.
  • Category A – right of way exclusive to transit, separated from all cross traffic: This is required for rapid transit. Examples include subways/metro systems and some grade-separated busways.
Transit system types by class of right-of-way.

Transit system types by class of right-of-way. X-axis is system performance (speed, capacity, and reliability), Y-axis is the investment required.

The distinction matters because the quality of the transit service is substantially different. Service in Class A right of way will be faster and more reliable than Class B, at-grade LRT. Part of the planning challenge is matching the right level of investment (and ROW category) to the goals for the system. However, even with the need to balance transit goals with those for urban design, planners like Keesmaat shouldn’t categorically dismiss the possibility of building Class A transit facilities.

Part of the confusion might be from the technology. A catenary-powered rail vehicle can operate in Class A, B, or C right of way, and fill the role of streetcar, light rail, or metro – all with little change in technology. Consider San Francisco, where Muni trains operate in all three categories – in mixed traffic, in exclusive lanes, and in a full subway. The virtue of light rail technology is flexibility, but that flexibility can also confuse discussions about the kind of transit system we’re talking about. The vehicle technology isn’t as important as the kind of right-of-way. Indeed, many of the streetcar systems that survived the rise of buses precisely because they operated in Class A and B rights-of-way.

Keesmaat certainly appreciates the difference between the kind of regional rapid transit you’ll see in Honolulu and at-grade LRT:

“The Honolulu transit corridor project is really about connecting the city with the county…. It’s about connecting two urban areas. That’s very different from the context we imagine along Eglinton where we would like to see a significant amount of intensification along the corridor,” said Keesmaat.

At the same time, the kind of transit she’s describing and the kind of land use intensity aren’t mutually exclusive at all – quite the opposite.


Subways are nice, but require a high level of density/land use intensity. Payton Chung put it succinctly: “no subways for you, rowhouse neighborhoods.” Payton cites Erick Guerra and Robert Cervero’s research on the cost/benefit break points for land use density around transit lines. This table to the right shows the kind of density needed to make transit cost-effective at various per-mile costs.

The door swings both ways. Rowhouse densities might not justify subways, but they could justify the same Class A transit if it were built at elevated rail construction costs. Finding ways to lower the high US construction costs would be one thing, but given the systemic increase of US construction costs, using elevated transit would be a good way to extend Class A rights-of-way to areas with less density.

Instead of categorically dismissing elevated rail, work to better integrate it into the urban environment. Consider the potential for the mode to transform suburban areas ripe for redevelopment. Wide rights-of-way along suburban arterials are readily available for elevated rail; redevelopment can not only turn these places into walkable station areas, but also help integrate elevated rail infrastructure into the new built environment.

Keesmaat’s concerns about elevated rail in Toronto stem from the impact on the street:

“The Catch22 with elevating any kind of infrastructure – a really good example of this is the subway in Chicago – not only is it ugly, it creates really dark spaces,” she said.

It’s not just the shadow but the noise of elevated transit lines that can be problematic, said TTC CEO Andy Byford. If you build above the street you’ve also got to contend with getting people there, that means elevators or escalators.

First, it’s not clear what Byford is talking about: accessing subway stations also requires elevators and escalators. The nature of grade separated rights-of-way is that they are separated from the grade of the street.

Keesmaat’s concerns about replicating Chicago’s century-old Els are likely misplaced. No one is building that kind of structure anymore – and a quick survey of newer elevated rail shows slimmer, less intrusive structures. Reducing the visual impact and integrating the transit into the cityscape is the real challenge, but the price advantage and the benefits of Class A right-of-way cannot be ignored. It’s not a surprise that the Star paraphrases UBC professor Larry Frank: “On balance… elevated transit should probably be considered more often.”

5 comments to Don’t rule out elevated rail in cities

  • Malcolm

    The light rail line near my home in Los Angeles runs at-grade for most of its length, but goes above or below the major cross streets, while minor cross streets are gated or blocked. There are a few intersections where it has to wait with traffic and that does slow things down a lot, but it’s still a good compromise.

  • Alex Block

    Malcolm, that is a great example – select grade separations improve the amount of Class A and B right of way, and can dramatically improve the capacity of a rail line. Higher speed, better reliability, and higher capacity are critical elements for the longer lines in large cities like LA.

    It shows the versatility of the technology.

  • Why does the minimum required density grow faster than linearly in cost per unit length?

  • Alex Block


    I don’t know that the authors would characterize it as a hard and fast minimum. They write: Transit-supportive density thresholds need to be viewed with caution. There is no one hard and fast rule that can be applied across all projects. Regression-based models mask considerable variation.

    They’ve built a model based on cost per passenger mile, not just cost per mile – I think the chart I included in the post is extrapolating from their model:

    I also think they’re basing this off of observations of US projects, which could easily skew the results if you’re looking for a more universal relationship between the project cost and land use intensity.

    The full paper is here – would love to hear your thoughts on it:

  • Elevated rail can work but it’s applications tend to be narrow. They tend to work best along straight lines and not necessarily in the centers of streets.

    With today’s construction techniques, elevated lines need not be early 20th century-style behemoths that block out the sun. The elevated portion of Washington’s Metrorail south of National Airport is a good example. It’s high and compact enough to let light hit the surface.

Leave a Reply

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>