WMATA's plans for Metro expansion are bad

WMATA's plans for Metro expansion are bad

Slow Boring / by Matthew Yglesias / 1h

Apologies for taking up a whole newsletter with a D.C. news story, but I figure we have a lot of D.C.-area readers here, and also this illustrates some general points about mass transit. Here goes.

Last week, the Washington Metropolitan Area Transit Authority (WMATA) announced four proposals for Metrorail expansion, all of which aim to address, among other things, the urgent problem that the system can’t run enough trains to meet non-pandemic ridership on the Orange Line, one of D.C.’s most popular commuting routes. These proposals are just lines on a map without detailed engineering proposals or cost estimates. Presumably, WMATA is also conducting cost-benefit analyses, and it’s possible one will make more financial sense than the others.

But the key flaw in these proposals is that none of them are very good.

The biggest problem is that despite ameliorating the specific Orange Line issue and bringing service to new areas, the plans replicate the flaws that got us to this problem in the first place: poor network planning that doesn’t take tunnel capacity issues seriously.

The penalty for sloppiness in rail network planning is much higher than with highways. Cars go point to point, so any old thing you slap together ends up being usable. Good planning and a coherent vision let you do more with less pouring of concrete, but to an extent, you really can just do one thing and then another and keep iterating. Trains can move a much larger volume of people through a given space, but they’re much less flexible than cars, which means planning where your tracks go and why is very important.

The orange crush

Here’s the short-term issue. Some time ago, D.C. set out to build the Silver Line to connect downtown to Dulles International Airport. The project is three years behind schedule (and counting) but will allegedly open sometime next year. There have been cost overruns associated with higher commodity prices (a secondary consequence of the delays), but as usual with rail construction in the United States, the baseline costs are just very high.

Even though it’s not done, the Silver Line does exist as far out as Wiehle. And that’s a problem because as you can see on the map, WMATA is running three services (Blue, Orange, and Silver) via a single east-west tunnel through downtown.

In other words, the whole Silver Line project only involved infrastructure running from East Falls Church to points further out into Virginia. They didn’t do anything to add capacity in the core, so the only way to add those Silver Line trains has been to reduce the frequency on the Blue and Orange lines.

That’s not great planning. Reasonable people can always disagree on how much to spend on new capital projects. But if you are going to spend a bunch of money, you’d like to make all users’ experiences better — or at least not worse. But in this case, they actually made things worse for many people.

It should be said, however, that the original sin here long predates the Silver Line.

The original sin

If you ignore the colors and focus on the very core of the system, you can see that Metrorail has what appears to be a pretty conventional structure. There are three big trunk tunnels running through downtown — an east-west tunnel, a north-south tunnel, and a U-shaped tunnel — that together form what the Transit Cost Project’s Alon Levy calls a “Soviet Triangle” (probably best exemplified by Prague, which is of course not in the Soviet Union but influenced by Soviet practices), which is a good, efficient track layout. Then the north-south tunnel branches south of L’Enfant Plaza, and the east-west tunnel branches west of Rosslyn and east of Stadium-Armory.

That's all very sensible. Branching can make sense if ridership demand is lower in the periphery, especially if construction costs are also lower out there. But the Silver Line is essentially a branch of a branch, which is an unusual decision that really impairs service.

Why did they do it that way?

I’m not exactly sure, but I think the history of how Metro was built gives you a hint. In 1981, they had a sensibly designed two-trunk system with branching. Then in 1983, they opened the embryonic version of the north-south trunk and decided it would be smart to connect it to the airport.

Elites are obsessed with building mass transit connections to airports because they travel a lot and appreciate the convenience of taking transit from the airport to hotels downtown. But most people who use transit in any given city are the people who actually live there, and most people do not fly on airplanes that frequently. In fact, a very large share of the people who might take transit to the airport are the people who work there, and those people probably do not live in a hotel located in the central business district.

The original airport connection that D.C. built was actually an example of an unusually good one. Yes, it connects the airport to downtown. But it also connects it to residential neighborhoods. And it’s short enough that they could (and did) extend it beyond the airport to other reasonably dense residential neighborhoods. The best way to connect something to transit is to make it on the way to somewhere, and the original system did that. But having done that, there is no particularly pressing need to build a second airport connection that just gets you downtown faster.

The better option for the north-south trunk would have been to head southeast from downtown, entering low-income neighborhoods that eventually did get service with the Green Line.

If you’d done it my way, then by the time I moved to town in 2003, the map would have looked like this, and I don’t think anyone would have found the idea of building the Yellow Line compelling. Yes, you’d get a quicker connection from downtown to the airport but only at the cost of throttling frequency to all the Green Line neighborhoods south of downtown.

The Blue Line stations south of King Street would experience the lower levels of service you get when you branch a branch, but that’s fine. Those are peripheral residential areas.

But this isn’t really the case with the expensive Silver Line project. It requires spending a lot of money on the premise that a lot of people want to take Metro to Dulles and/or that you’ll successfully transform Tyson’s Corner into a dense, transit-rich area, but then also asking it to get by with branch-of-a-branch service levels. Alternatively, you could cut service west of East Falls Church.

Now of course we can’t go back in time to fix the Silver Line, but once you understand the nature of the problem, you can see it cropping up yet again in WMATA’s newest proposals.

A half-used tunnel

Three of WMATA’s four proposals involve digging a new east-west tunnel through downtown.

In one version, that new tunnel carries the Silver Line across downtown and over to New Carrollton.

In another, the new tunnel carries the Blue Line across downtown and over to Greenbelt.

In the third, the tunnel forms the northern edge of what becomes a somewhat odd ring line.

But all of these proposals share a fatal flaw. Right now, the signaling system in D.C. can handle trains spaced four minutes apart. Because the Blue Line shares tracks with the Yellow on one segment, that means you can at most run one Blue train every eight minutes. That’s fine on the shared segment or on a peripheral branch, but it means the new tunnel would have at most one train every eight minutes. It would be at half capacity right in the downtown core. And the Silver Line shares tracks with the Orange for a stretch, so you have the same problem with a new Silver tunnel downtown. Since under even a very optimistic scenario, a big new tunnel across downtown is going to be very expensive, it’s crazy to design it in such a way that it can only be half full.

The fourth plan, which is not quite as bad, is to build a whole parallel set of tracks so that the Silver Line can run express rather than sharing tracks with the Orange Line.

When I first moved from New York to Boston, I thought it was unfortunate that the T doesn’t have express tracks. Then I moved to D.C. and thought it was unfortunate that the Metro doesn’t have express tracks. Eventually, though, there comes a time when any young transit enthusiast learns that London, Paris, Berlin, and Moscow don’t have express tracks either, and it’s actually New York that’s weird.

And here you see why. Whatever it costs to build X miles of track duplicating the route of the Orange Line in Virginia could also be spent on building roughly X miles of track somewhere else, thus serving new neighborhoods. The only reason going from four tracks to two on that stretch seems remotely appealing is because it solves that capacity crunch in the existing east-west tunnel. But there’s a much better way.

Kill the Blue Line

As I wrote earlier, D.C. probably shouldn’t have built the Yellow Line in the first place. But it exists, and now the capacity crunch is on the east-west trunk tunnel, not the north-south one. And there is a very cheap and easy solution to this problem: stop running Blue Line trains.

The Yellow Line could branch south of King Street to serve Franconia-Springfield and Van Dorn Street. Frequency on those Yellow branches wouldn’t be great, but it’s only a handful of stations. The Arlington Cemetery Station could simply be abandoned. The change would cost approximately $0 to implement and while it would degrade service to a handful of stations in Alexandria, a much larger number of stations would see improved service.

If you want to spend a little money, you could do a construction project at Rosslyn to create the track capacity to run a Blue Line Shuttle from Pentagon to Rosslyn and reopen the Arlington Cemetery station. That’s not something you would build from scratch, but since the track already exists, investing the modest sum involved to make it work is probably a good idea.

For rail fans, this is not as exciting as digging a whole new tunnel. But the fact is that WMATA is very unlikely to get the money it would need to dig that whole new tunnel anyway. And digging a new tunnel that’s half-empty is wasteful and a little crazy. The virtue of killing the Blue Line is that it’s very cheap and easy, and it would restore Metro as a basic three trunk system with branches, a sensible starting point from which to contemplate new expansions. This post has gotten very long already, so I’m not going to delve into what those expansions could or should be here.

But one reason to prefer a minimalist solution to the Orange Line capacity problem is that this all, as usual, hinges crucially on land use policy. Discussions of these WMATA proposals display a lot of enthusiasm for the idea of bringing Metro service to Georgetown, and building nothing at all does not accomplish this. But in the next post, we’re going to compare the actual value of this idea to some other possibilities. Either way, the beginning of wisdom is to simply address the Orange Line capacity issue on its own terms, then talk about expansion rather than mixing the two together.

SHARE

VISIT WEBSITE