Saturday, June 25, 2011

Washington Metro, Improved

This is the first post of a three-post series (maybe) on improving transportation in Maryland, D.C., and northern Virginia. In the first post, I concentrate on the D.C. Metro. In the second post, I will talk about commuter rail in Maryland and northern Virginia (MARC and VRE), and the third, improvements to the Baltimore metro.

The D.C. Metro is among the most convenient subway networks in the U.S., being about equally as convenient as, say, Boston's or Chicago's (though nowhere near as convenient as New York's). However, it has two conflicting mandates: at its nether reaches, it acts as a commuter rail line, and in the city, a traditional subway. This would work a lot like New York's Far Rockaway or Jamaica Bay subways...if there were express tunnels. Instead, since the tunnels are all two-track and there doesn't seem to be a way for overtaking to be scheduled in, all trains are effectively local trains. This works well on lines, in theory, where there are connections with MARC or VRE, since it enables local/express passenger transfer--that is, passengers originating on the Metro but wishing to skip intermediate stations and go straight downtown can use the commuter rail going to Union Station from places like Alexandria, Silver Spring, or New Carrollton, or passengers originating at Manassas and going to Bethesda, or Gaithersburg going to Rosslyn. However, since the ticketing networks of MARC, VRE, and the Metro are mutually exclusive, this is made prohibitively difficult for most passengers. Therefore, the first major action needed is to unify the ticketing schemes and operational patterns across the three networks, in a way that mirrors European S-Bahns. This is, of course, organizational improvement.

Even so, there are real concrete improvements that do need to be made. The network offers poor access (service gaps) into parts of the metropolitan area to the north, south, east, and west of downtown--most pronounced all along the Potomac, in east D.C.'s poorer neighborhoods, and directly west of the Pentagon. Additionally, as the core of the system was designed with the 1970s downtown in mind, it offers an increasingly inadequate service. (See Exhibit A.)
Exhibit A. Note how current system, while adequate by American standards, has major gaps in several directions, and downtown service can be improved.
Greater Greater Washington recently ran a series on second-generation improvements to the Metro, and while some of the ideas were pretty good (such as running the Yellow Line out to Union Station via 2nd), others were merely okay (the north-running Blue Line idea and its Silver Line kin: same idea, different lines), while others were just atrocious (the how far outside D.C. would the Brown Line be extended, anyway?
Exhibit B. The new network idea.
Taking the best three ideas--the Yellow Line tunnel out to Union Station, the southerly separate Blue Line tunnel, and a modified version of the Green Line spur to National Harbor, I came up with Exhibit B. This Metro system sports two new lines (Brown and Pink), and, to complement the new tunnels, extensive new interlining. This interlining would, in addition, offer superior Generation II downtown service. A final addition is the construction of an express track along the interlined Silver/Green section between Falls Church East and Rosslyn; this track would allow for faster commutes along what is now the single longest interlined section in the network (see Exhibit C).
Exhibit C. Downtown lines and interlining.
These lines would be as follows:
Red Line would be unchanged.
Orange Line would be unchanged.
Green Line would be unchanged.
Blue Line would run a new downtown tunnel from the Anacostia River, via H Street, to Union Station, where it would curve down to the National Mall, servicing the primary jobs and cultural centers in the city. It would then curve up through Foggy Bottom to Georgetown and return to the current Blue Line at Rosslyn.
Yellow Line would (1) be extended south from Huntingdon to Beacon Mall, (2) run a new tunnel via I and 2nd Sts. to Union Station, and (3) follow a new alignment north along North Capitol Avenue to Rhode Island Avenue and thence 18th St. NE out to Langdon and Howard Divinity School.
Silver Line would be extended along a new route along North Carolina Avenue, Tennessee Avenue, and the Bladensburg Road to Bladensburg and thence across the Anacostia out to Riverdale, where it would terminate in a new transfer stop with the MARC Camden Line.
Purple Line light rail would be unchanged. (It's a Maryland project, anyhow.)
Brown Line would run from Carderock, MD, via Langley, VA, and American University through Georgetown and thence interline along the Yellow and Green Lines into Anacostia, where it would peel off and continue running south to Forest Heights and National Harbor.
Pink Line would run from Walter Rand Medical Center in northern D.C. down through Columbia Heights, where it would interline with the Green Line to L'Enfant Plaza, and then interline with the Yellow Line to Pentagon, where it would follow a westerly route into Virginia via Columbia Pike to Bailey's Crossroads and Annandale.

This network would implement service to the D.C. area's major service gaps and greatly improve downtown circulation (especially around the Mall). Combined with fare and schedule unions with MARC and VRE, making cross-platform connections infinitely easier, this network would offer the D.C. area an impressive Generation II metro/commuter rail network.


  1. You really need to look at a population density map. Like this one. There's zero justification for a new Metro line to Langley/Carderock even if it wasn't under the most difficult geology in the region. There's also very little reason to build a line past Bailey's on 244/236. The eastern extension of the Silver Line goes through a large area of low density. And so on.

    I take it you're not too familiar with the area as you have a several misspellings in your post. The issue to keep in mind is that the urban core of DC is relatively small, and the density drops below what can support a tunneled heavy rail system close in. As of now, there is no consideration being given to significantly increasing the density over large parts of the city except by a few obsessive bloggers. So the challenge in DC is to lay out the minimum amount of construction that will solve the main problems facing the core: the Orange Crush, an impending crush on the Yellow/Green south of Gallery Place, and a having only a single line serve Union Station. Beyond the core, for now, the rest of DC is best served by streetcar, which is what it was built around in the first place.

    You can see my solution to those problems here.

  2. As always, thanks for the comments. The general gist of the idea is that the best outbound lines are really not routings like the Brown Line being considered in the Greater Greater Washington post, but rather where there is relatively little Metro service in the region. I was also using Google Maps as my main place-name resource, which is publicly amendable.

    If population density dictatates terminating the line at Baileys rather than Annandale (which I picked due to Annandale being a strong park-and-ride location--see here), that's fine by me--the point was that there was a gap in regional coverage that needs filling.

    Similar reasoning drove Carderock. There's also some sort of odd building out there that I was guessing is a government job center. If not...well, feel free to modify the line beyond Langley, where I know there's a big government employment center to wherever you think is the best terminus.

    As to the Orange Crush, the best idea in the long run is probably going to be to run an express tunnel from its junction with the Silver Line through Arlington and downtown D.C. and out to where I suggest the Silver Line heads out into eastern D.C. and towards Bladensburg. The terminus I gave that line is driven by transfers.

    I don't claim my ideas are perfect. Someone with insider knowledge will know more. But what I'm asking for is that consideration of service gaps drives network expansion. Based on the current proposals I'm seeing, I just don't think that's the case.