Air Rights Dreaming

As land available for development in core urban areas grows scarce, some cities have started to look to the air rights above transportation facilities, such as rail yards and freeways, as potential locations for growth. Unfortunately, building over this infrastructure, especially rail yards, only makes sense for high-margin projects in very built-up areas. To see why, let’s take a look at some existing air rights projects, and explore the challenges that confront the engineering for such projects.


No North American city is more air-rights crazy than New York. After all, air rights are practically baked into the city’s DNA: before zoning restricted building heights, it was common for developers of tall buildings to buy the air rights over adjacent shorter buildings to protect views from future construction.

New York is also home to two of the most widely-known rail yard air rights developments: Penn Station/Madison Square Garden and Grand Central Terminal. Given the success of those projects, it would seem to make sense that overbuilds at Hudson Yards, Atlantic Yards, and Sunnyside Yard would be viable.

However, as we’ll see a little bit later, PSNY and GCT had a big advantage in that the overbuilds were planned, designed, and constructed at the same time as the rail facility. They’re also located in very desirable parts of Manhattan, where rents justify the cost of decking over the rail yard. Atlantic Yards isn’t in as good a location, and the developer has taken stunning losses. Sunnyside Yard is worse, surrounded by low-rise residential and low-rent industrial space; there’s no way rents in the area would justify the cost of decking. Meanwhile, at Hudson Yards – the one location where the market might support private development of air rights – the city seems intent on forcing commercial development, while the market wants residential.

Update: as True Urbanism pointed out on Twitter, Madison Square Garden was of course built well after the original PSNY. Construction would have been complicated by building over an active railroad (though facilitated by the fact that the railroad itself was doing the construction and was desperate for money). Building over platforms offers something of an advantage in that you have a place to put columns.

Boston: Turnpike & NEC

Projects in Boston have not fared much better. In 2008, after wrangling significant tax subsidies out of the city and state, and fending off the predictable NIMBY challenges, the Columbus Center was set to begin construction over the Massachusetts Turnpike and Northeast Corridor east of Back Bay Station. Then the financial crisis hit, and the main investor (CalPERS, of all people) pulled out. The project has gone nowhere since, with developers in downtown Boston focusing on easy to redevelop parking lots in the Seaport District.

Further east, the South Station Tower was another victim of the financial crisis and shows no signs of life, despite being planned for during the reconstruction of South Station in the 1980s. Further south, plans have been floated for Northeastern University to bridge the gap between Back Bay and Roxbury by building over the NEC, but the university has consistently held these plans to be nonviable. And further west, the first part of John Rosenthal’s Fenway Center project is seeking subsidies as well.

The marquis air rights projects in Boston – the Prudential Center and the various cap parks on the Southwest Corridor – were all built, like PSNY and GCT, concurrent with the construction of the transportation facility. The Prudential Center was built in the 1960s at the same time as the freeway, at a time that railroads were desperate for money. When the Southwest Corridor was built, all rail traffic was temporarily diverted onto the Fairmount Line. Only Copley Place was built over active rail and after the Turnpike was open, and to be honest, I’d love to know how they pulled it off.


Out in Chicago, there’s the famous post office building, that the 290 goes straight through. The building, however, predates the freeway – it was completed in the 1920s and the post office didn’t open until 1933. The building was designed with passageways left for the road.

In New York, you have the brutal Robert Moses high-rises over the Cross Bronx Expressway, also built at the same time as the freeway. Moses proposed a similar concept for the never-built Mid-Manhattan Expressway.

Engineering Matters

So why is it so expensive to develop the air rights over active transportation facilities? A few engineering and construction challenges come to mind.

First, from an engineering perspective, you have a very unusual and highly constrained site. Next time you see a high-rise under construction, check out the amount of space between the columns. You’ll probably see it somewhere around 30’, in a rectangular or square pattern. This spacing keeps the bending moment in the beams small. (Bending moment is the force that tries to snap beams the way you’d snap a twig.)

But 30’ is only enough space for two lanes of traffic, or two rail tracks on very tight spacing. To span a four-lane freeway or four rail tracks you’d need at least 50’, again assuming no room to spare. Bending moment in a uniformly loaded beam scales with the square of the length of the beam, so a 60’ beam would have four times as much as a 30’ beam. That means you need big honkin’ beams, very strong joints to transfer the load to the columns, and big ol’ columns to take that load to the foundation. All of that costs money.

It’s also hard to fit a rectangular pattern of columns onto an existing transportation facility. Freeways tend to curve and have on and off ramps. Rail yards are worse, because if the yard wasn’t built with future air rights development in mind, it’s unlikely there’s enough room to fit columns between the existing tracks. If there’s nothing between tracks, they’re often build between 13’ and 16’ apart. At 13’ there’s essentially no useful space between the tracks. At 16’, if the rail operator is feeling extremely generous and gives you 7’ clearance, you could squeeze in a 2’ column. More likely, you need 18’ to 20’ between tracks to get columns in there. That leaves you two costly options: rebuild the rail yard before you construct the building, or go with very long spans between places where you can get columns in based on the existing geometry.

On top of all of that, you have to deal with turnouts and crossovers. Let’s say you have a column line between two tracks that are 18’ apart. A crossover between those two tracks, even a low speed one, is going to preclude any columns on that column line, probably for at least 150’. Of course, if you were planning to build over a new rail yard at the same time you built the yard, you’d modify the track design to make things easier for yourself – that’s why it was so much easier to build PSNY and GCT. But Sunnyside wasn’t planned that way, and it would be a big problem for the structural design.

For example, consider this hypothetical yard with 6 tracks and 2 ladders (crossover tracks). The red squares show column spacing. That big gap in the middle is a big money pit.


Second, after you design all of this, you’ve got to get it built, over an active transportation facility. Once the project gets above the deck level, things get easier, but building the deck itself is very costly. Consider, for example, Sunnyside Yard. You can’t be drilling and pouring foundations, or erecting columns, six inches away from tracks with trains running on them. Nor can you be hoisting and setting massive steel beams for the deck over tracks with trains running on them. That means all the work has to be done at night.

But there’s still activity at the rail yard at night, so you can only work on parts of the site at a time. Your work window might be, say, 10pm to 5am. That doesn’t mean you start working at 10pm, that means the railroad starts the process of giving you a work zone at 10pm. So the area has to be cleared of trains, and then the electrification has to be shut down and confirmed to be grounded. You work from maybe 11pm to 4am, and then you have to give the track back so that the railroad can make sure it’s ready to be put in service by 5am. And if there are service disruptions or maintenance issues, nuts to you – the dispatchers and maintainers are going to worry about restoring service first, and you might not get any work done.

A transportation air rights project also probably faces a common challenge in urban construction – lack of lay down area, or space to store construction equipment, machinery, and material. That means that every day, you have to bring that stuff in from elsewhere, which also takes time and money.

Add all of that up and you’re looking at a complicated, costly design, and complicated, costly, inefficient construction. These types of projects do make sense in some places, but you have to be able to get high returns to justify them.

What Are We Trying to Accomplish?

As always, it’s good to take a step back and look at our goals. What are we trying to accomplish, in terms of the development of the city, with these projects? There’s a saying among engineers: we can design anything you’re willing to pay to build. The question is if it’s worth building it.

In densely-developed Manhattan, it seems possible that some type of project should be able to make a go of it at Hudson Yards without tax subsidies. But Sunnyside Yard, adjacent to low-rise residential and industrial land? Projects over the Massachusetts Turnpike adjacent to parking lots in the South End or single-family housing and duplexes in Lower Allston? There’s no way those projects are economically viable, not without allowing a ton of development and providing big-time subsidies. It’s telling that in the North American city with the biggest boom in tower construction, Toronto, there are zero projects being built over Go Transit’s trunk line downtown, despite proximity to high rent areas and the lakefront.

A few days ago, Market Urbanism tweeted that the lengths to which people will go to preserve low-density residential development are incredible. And that’s what’s really going on with most air rights projects, as well as a lot of other trends in urban development. People would rather subsidize air rights projects to the tune of tens of millions of dollars – or pretend that air rights projects make sense without subsidies – than allow more development in existing neighborhoods. But that’s a really inefficient use of taxpayer money. If the goal is to create more housing, those subsidies would go much further at boring, simple construction sites. Air rights projects might sound exciting, but the costs should be left for the private sector to incur in places where they really make sense.


9 thoughts on “Air Rights Dreaming

  1. Alon Levy

    Having just visited the area around Sunnyside Yards, I think you’re selling the idea short. Large portions of the yards are now unused, having been turned into East Side Access staging areas. If they aren’t really needed, they can be redeveloped, without having to deal with an active yard. Moreover, from Queens Boulevard to 43rd, south of the mainline tracks there are never more than two yard tracks parallel, so there’s more space. To the north of the mainline tracks there are lots of yard tracks, but in the vicinity of Honeywell, there is generous space between pairs of tracks, enough for a car to access the train, and no crossovers. In the vicinity of 39th, there are no yard tracks except at the northern margin of the site.

    Moreover, there’s so much auto-oriented, unwalkable development north of the site, to say nothing of the streets crossing the tracks (Honeywell, again). Developing the non-yard parts of the site, for example near 39th, would give the area more pedestrians, creating more demand for developing the parking lots and redeveloping the auto shops north of the site.

    1. letsgola Post author

      So the idea would be to develop the southern edge of the yard (unused areas plus maybe over a couple yard tracks) from Queens to 43rd? I think that would work, though maybe not short term. There seems to be a lot of industrial/warehouse type development on the south side west of 39th. The block from 39th to 43rd would probably be the most promising. The area immediately east of 43rd seems oddly low density. (I haven’t been there personally so I’m relying on Google Earth.)

      Basically unrelated, but you wouldn’t happen to know anything about that weird mini-expressway from the Queensboro Bridge at 22nd St to Thomson Ave where it crosses the yards, would you? Offhand I’d guess it was some weird Moses scheme to connect the LIE to the bridge but as it exists, it seems totally redundant to Queens Blvd.

      1. Alon Levy

        First, Patrick O’Hara, who knows more about current LIRR practices than I do, has this map of Sunnyside Yard usage.

        Second, the point of any Sunnyside Yard redevelopment is to leverage a future junction station. This also includes non-yard space, including east of 43rd, yes. That area is indeed quite auto-oriented: Home Depot has an enormous parking lot; Northern Boulevard is mostly auto shops and has very fast traffic, whereas farther west it has local subway stops; there are former industrial usages next to the yards.

        I have no idea about the offramp; I didn’t go that far west.

  2. John Hupp

    Chicago has lots more air-rights development than just the Post Office. Everything else along the river in the West Loop is built on air rights. Millenium Park, Illinois Center and a number of other buildings at the north end of Grant Park are built on air rights.

    It’s interesting that most proposed freeway caps these days are not for real estate development but rather for parks. Take the Hollywood Freeway Cap Park, or the 101 Cap Park in Downtown LA, the Kennedy Cap Park in Chicago. Or the already completed Big Dig in Boston or Klyde Warren Park in Dallas. Basically when the main goal is to create new public space and connect divided communities, people seem much more inclined to have the government pay for it, no matter the cost.

  3. Pingback: Penn Station Elimination Followup | Pedestrian Observations

  4. Pingback: TOD: Transit-Owned Development | PlaNYourCity

  5. Pingback: Sunnyside Junction, Redux | Pedestrian Observations

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s