Redundancy, engineered resilience, and subway systems: Metrorail failures will increase without adding capacity in the core

Yesterday, a fire crippled travel on the Blue, Orange, and Silver Lines of the DC Metrorail system ("Metro delays caused by cable problem similar to last year's fatal smoke incident," Washington Post). From the article:

An electrical fire in a Metro tunnel early Monday, which caused huge delays on three subway lines, involved the same type of track-based power cables that burned during last year’s fatal Yellow Line smoke incident in another tunnel, the transit agency said.

The fire, which broke out about 4:30 a.m. just west of the McPherson Square station, fouled the morning and evening commutes for thousands of riders on the Orange, Silver and Blue lines. While repair crews worked in the tunnel, the frequency of trains on the three lines was greatly reduced, and some stations were bypassed.

This event combined with a power substation fire over the summer ("Temporary speed restriction in area of Stadium-Armory," WMATA) which led to service reductions until repairs were effectuated has made me realize even more decidedly that the addition of the Silver Line to the Metrorail system without adding capacity in the core regularizes system failure on the Orange, Silver, and Blue Lines, because it has increased operational complexity without adding slack resources.

Adding the Silver Line without adding core capacity overstresses the system.  I have to believe when the decision was made to allow the Silver Line to be constructed without adding an additional crossing between Rosslyn and DC, and without a separate alignment and adding additional capacity in the core of DC, there was a failure to recognize adequately how the Silver Line addition would unbalance and stress the system in ways that would lead to routine and regular operational failure.

Adding a river crossing and creating a kind of parallel track in the core of DC were the major elements of the proposed "separated blue line" dating to around 2002.  But these plans were dropped when WMATA experienced budget problems and devolved authority for expansion planning to the jurisdictions.

-- "The "Downtown" Circulator and Rosslyn, Virginia," August 2006
-- "Blinking on urban design means you limit your chance for success," September 2006
-- "Winners and losers with the Dulles subway project," June 2007



Design and engineering constraints created by how the Metrorail system was designed.  This conclusion about the Silver Line and system failure builds on previous writings about how the engineering of the Metrorail system and the inability of any transit system to be 100% in a state of good repair means that it is unlikely that the Metrorail system will ever be able to achieve its planned maximum usage of 1.1 million daily riders.

I argue this results from how:

- each line is only two tracks, so that when a train fails or one track is degraded or out of service system performance degrades extra-normally, not to mention that additional tracks would afford more and different types of service as well as redundancy

- sharing of track between the Green/Yellow and Blue/Orange/Silver Lines reduces maximum capacity of each of the lines because of the time required for switching trains between lines; by comparison the Red Line can achieve almost 40,000 riders/hour in each direction because it doesn't share its track with other lines

- as cars age they become less reliable, and it isn't possible for the transit system to keep buying new cars (not to mention that some manufacturers are better than others and some car series function better than others)

- as the system ages, the need for maintenance and repair closes lines, and without track redundancy that means lines are shut down, which degrades service quality and leads to ridership losses.

(There are other issues too such as platform widths, the way that the platforms are offset at Gallery Place, egress capacity of stairs and escalators, failure to have 4 doors on train cars, seating configurations, failure to "harden" the system for weather in the core so that the system can operate in severe weather conditions, etc.)

The argument that additional tracks aren't financially justified may be flawed. People argue that the ridership levels of the Metrorail system aren't high enough to justify more than a two-track system, but I wonder if we have to change our assumptions about what makes sense engineering-wise for systems aiming for 1 million or more riders per day.

Note this is a moot question because rebuilding the system to add tracks is out of the question for many reasons.  But it is still worth discussing in terms of how it helps us understand transit system design and operation.

First, there is the need for redundancy as discussed above, because real not theoretical capacity is the issue.

Second, I wonder if we would have ridership increases from a more robust system, because riders would be attracted by reliability.

Third, because a more robust transit system would support even more intensive development, thereby attracting more riders and more development proximate to the stations, taking advantage of that reliability.

Tom Toles editorial cartoon about Metrorail. Washington Post.

Adding capacity to the core of the Metrorail system as a way to improve system reliability. In October 2015, US Department of Transportation Secretary Al Foxx said that Metrorail needs to focus on improving safety and reliability of the current system, rather than expansion planning ("We all want Metro fixed, and fast," op-ed).  From the article:

Metro can forget any new rail-expansion projects until it meets our safety standards. We may also require periodic closures of some Metro facilities to ensure safety measures are implemented. ...

Effective immediately, the FTA will conduct inspections, investigations, audits, examinations and testing of Metro’s equipment, facilities and operations. Safety failures we find must be corrected. There will be no new projects until Metro completes its punch list. This direct oversight will also focus on the 78 actions the FTA required Metro to complete as a result of the most comprehensive federal safety review conducted on a rail transit agency. The FTA will directly supervise Metro’s progress. The FTA has a significant enforcement tool through controlling Metro’s federal grants and will not allow federal dollars to go to Metro for any activity other than safety improvements.

While there is no question that jurisdictions, especially in Virginia, have been more interested in continuing to expand Metrorail outward rather than considering system reliability, I believe that Secretary Foxx fails to consider how "expansion" can be a solution to the system's safety and reliability problems, which result in part from lack of capacity in the core.

Polycentric vs. monocentric transit systems.  Many times I have written about how in Cities in Full, Steve Belmont argues that "polycentric" transit systems like DC's Metrorail system promote sprawl rather than land use concentration, and he contrasts Metrorail with systems that are monocentric in design.

A monocentric system provides more and better service in a smaller area, and provides intra-district as well as inter-district transit, while a polycentric system is more focused on serving riders on longer trips, primarily to school or work.


Extension vs. intensification.  We need to distinguish between extending the system outward, or intensifying the system by adding capacity in the core, because these types of extensions have different impact.  It's the type of expansion that matters, not expansion generally.

Adding capacity in the core improves reliability while extending service outward, especially without increasing core capacity, degrades service.  Eliminating expansion in general limits the opportunity to improve system reliability.

Note that in some instances, system extension serves important purposes (e.g., such as to Fair Oaks in Fairfax County, extending the Orange Line, or to Fort Belvoir, which could be reached by extending the Yellow Line) but it shouldn't occur without simultaneously addressing how extension impacts the core system.

Conclusion: One of the best ways to improve reliability and safety in the Metrorail system is to create the Separated Silver Line within DC.   Sadly, this is where my lack of ability with graphic design software shows.  Many years ago David Alpert of Greater Greater Washington was kind enough to create a graphic of an idealized Metrorail system based on my thinking at that time.


But my thinking has continued to evolve and this is what I would propose now:

Separated Silver Line and Separated Orange Line

1.  Instead of what used to be called the separated Blue Line, I would now term it a separated Silver Line.  It could also be thought of as a "Downtown Relief Line."

Instead of joining the Orange Line at East Falls Church Station, instead treating it as a transfer station, the Silver Line could continue south to Route 50/Arlington Boulevard then east to Rosslyn, crossing to Georgetown, continuing eastward to Union Station (this adds capacity to serve Amtrak's plans for expansion), and then further east to H Street NE.  (The Arlington Boulevard alignment was suggested by commenter Ryan in a thread at GGW.)

2.  This would add stations in Northern Virginia and DC.  At the very least it would add one crossing and at least 9 new stations in DC, with 3-4 more stations if an additional Silver Line leg was constructed up Bladensburg Avenue.  It would serve key activity centers not currently served in Virginia (Arlington Boulevard/Seven Corners) and DC (Georgetown especially), providing additional capacity Downtown and at Union Station.

3.  In the vicinity of RFK Stadium, I would route the Separated Silver Line onto the Orange Line alignment from RFK Stadium to New Carrollton.

4.  Perhaps a separate intra-city leg could be added along Bladensburg Road from H Street to New York Avenue and Fort Lincoln.

5.  The current Orange Line alignment would remain the same from Vienna to Stadium-Armory.  At Stadium-Armory it would assume the current Blue Line alignment to Largo Town Center.

6.  Perhaps a shuttle between the separated Silver Line and Orange Line could be created in the vicinity of RFK so that people could transfer between the lines.  The current bridge from Stadium-Armory over the parking lots to the Minnesota Avenue station would be eliminated, replaced by a tunnelized alignment along Benning Road from H Street and Bladensburg.  This would enable superior redevelopment of the RFK Stadium site.

Separated Blue Line

1.  The map above shows a truncated Blue Line ending in Rosslyn, and a "Brown Line" starting in Georgetown, extending south to National Harbor in Prince George's County and then west to Alexandria.

2.  I would merge the two lines so that the Blue Line continues from Rosslyn across to Georgetown and then up Wisconsin Avenue, connecting to the western leg of the Red Line, and turning east at some point in Upper Northwest, providing an east-west high capacity transit connection in the upper city.

3.  The Brown Line concept merges some of my ideas with Internet blogger MV Jantzen.  Now, I don't know what alignment I would propose to take the line south.  The southern alignment shown in the map above is based on Michael's ideas.

4.  This would add a number of stations within DC serving areas currently not served by high capacity transit.  Depending on the southern alignment, it would serve National Harbor, one of the area's leading new activity centers in Prince George's County, and would provide more transit stations in Prince George's County and an east-west connection to Alexandria, Virginia.

Separated Yellow Line

It's not so relevant to this particular discussion, but I'd also propose a separated Yellow Line within DC (this idea was first proposed in a blog entry by Dave Murphy) and Montgomery County.

Besides adding a station to serve Jefferson Memorial/The National Mall, the line would probably go north on Georgia Avenue from Howard University and extend outward from Fort Totten out New Hampshire Avenue to White Oak.

The DC justification of extension out New Hampshire Avenue is to interdict commuter traffic to and from DC.  Montgomery County would see additional intensification benefits also.  The trick with a leg along New Hampshire is what to do with a Yellow Line between New Hampshire Avenue and Silver Spring and beyond via Georgia Avenue.  It could be a two leg system on the north, just as I suggest a Bladensburg leg for a separated Silver Line.

It is Virginia's issue, but probably the Yellow Line should be extended south from its terminus at Huntington Avenue station in Fairfax County to Fort Belvoir, also adding stations to Rte. 1.  This should have happened anyway as a result of military base closures and the significant expansion of military activities at Fort Belvoir.

Benefits.  By separating the Silver and Orange Lines, each line would have the maximum capacity for 30 trains/hour, which is more than can be accommodated sharing the lines currently.  At least one additional crossing between Rosslyn and DC would be added.

Many more stations would be added in DC, as well as redundancy and more capacity in the core, increasing system reliability.  New areas would be served in both Virginia and DC, increasingly the ability to capture automobile trips and shift them to transit.

Note that adding legs, such as up Bladensburg Road on a separated Silver Line or both up Georgia Avenue and out New Hampshire Avenue on a separated Yellow Line would reduce capacity from the theoretical maximum because of time required to switch from the branch to the main line.

Paying for it.  As I have discussed before, by increasing the allowable height at which buildings can be constructed in Downtown, DC's property tax base would increase significantly and become large enough to fund this kind of bold expansion and intensification of the transit system within DC.

Ideally, a "transit withholding payroll tax" could also be assessed, but for this to be worthwhile, the the federal government has to be willing to pay it (unlikely in the current economic environment), therefore this revenue source may not be worth pursuing.  In the US, certain jurisdictions in Oregon and in Greater New York have such a tax.  In France, this type of tax, called the versement transport, provides a preponderance of funding for transit service in most cities.

Funding for additions within Maryland and Virginia would be provided by those jurisdictions. But planning, design, engineering, construction, and financing should be coordinated, rather than discoordinated, as occurs at present.

=====
Purple Line impact on Metrorail system reliability.  The Purple Line will be a light rail line serving Montgomery and Prince George's Counties in Suburban Maryland.  Its western terminus will be at Bethesda Station on the western leg of the Red Line, and it will connect to the Red Line western leg at Silver Spring, to the Green Line at College Park, and to the Orange Line at New Carrollton, which is the eastern terminus for the line.

The system won't interline with Metrorail so it won't impact system reliability in the way that the addition of the Silver Line has degraded service on the Orange and Blue Lines.

You could argue that by adding riders to the current system, it could add stress because of the failure to add capacity in the core.

However, by providing a missing east-west connection between the subway lines outside of Downtown DC, likely the Purple Line will "add capacity" by providing four new transfer points between subway lines, facilitating transfer between lines without requiring riders to go all the way to Downtown DC to do so, as is required today.