Posts about WMATA
To resolve capacity issues, WMATA has proposed building a new Metro loop line through downtown. But there might be a better way to resolve those issues and make the current system more efficient at the same time.
New tunnels and realigning some Metro lines could increase capacity in the core. All images by the author.
While Metro's solution will adequately address the capacity problems, it will do nothing to sort out the fundamental geometric problems with the way the system was set up to begin with.
Metro's proposal overlays multiple U-shaped or loop lines through the core, widely acknowledged as one of the least efficient ways to organize rail transit networks. Instead, Metro should build to provide routes that hit major job centers & transfer points, travel from A to B as directly as possible, and minimize the merging of routes, or interlining.
An alternative approach would be to combine new capacity from south and east of downtown with the branches of the Red Line so trains can travel all the way through the core without looping back on themselves. I believe the proposal shown in the map below shows the most effective combination of these lines.
The tunnels to create this network would be as follows:
1. Build a new approximately 3.5-mile M Street subway for the Orange and Blue lines from Rosslyn to Metro Center, with five new stations at Rosslyn, Georgetown, West End, Connecticut Avenue, and Thomas Circle
2. Build a new approximately 0.85-mile tunnel between McPherson Square and Judiciary Square to combine the existing Franconia-Springfield end of the Blue Line, the Huntington end of the Yellow Line, and the Glenmont branch of the Red Line into a new Red Line. This would include one new station under G Place between 9th and 10th streets and transfers to the Gallery Place and Metro Center stations.
3. Build a new approximately 1.65 mile tunnel under 4th and 6th Streets SW between Waterfront and Gallery Place to combine the existing Branch Avenue end of the Green Line and Shady Grove branch of the Red Line into a new Green line. This would include two new stations at Independence Ave and Constitution Ave and a transfer to the L'Enfant station.
This work solves Metro's core capacity problems, creates a more efficient system by straightening the Red Line's branches, and reduces system-wide merges to a total of three, down from six.
It also does so while building fewer stations than Metro's loop proposal, with eight instead of fifteen, and six miles of subway instead of eight. The express tracks along I-66 are not needed under either approach at current ridership, and could be built in a later phase.
Under this proposal, every branch of the network except the Greenbelt branch would pass through the city's central business district around Farragut Square, whereas under Metro's plan both the Greenbelt and Branch Avenue branches of the Green line will miss it.
If possible, the current track connections should be kept to allow for system redundancy in cases of maintenance or emergencies. The new platforms at Rosslyn and Pentagon should be designed to minimize transfer time, ideally with cross platform transfers. Because the line to Shady Grove has much higher ridership than the line to Branch Avenue, the system should be built to allow some trains to terminate at Independence Avenue and turn back north.
Metro could eventually provide additional service along I-66 and to Union Station, as shown in the long-term proposal below.
Providing fully separate tracks for the Blue Line would achieve both objectives and provide operational independence for every line. Such a subway could be built following H Street and Constitution Avenue. It would form an east-west "express" route, relieving congestion at Union Station and on both the Orange and Red lines through downtown.
This long term plan would set the system up for additional extensions in Maryland and Northern Virginia. In addition, the burgeoning streetcar network and the circumferential Purple Line would supplement this network.
Metro's proposal and the one described in this post both provide for needed expansion to the system's core and eliminate chokepoints, but do so in opposite ways. One interweaves lines together extensively, while the other reorganizes track connections to keep routes as operationally independent as possible.
This difference will have major effects on how a rider will navigate the system, where they will transfer, and how long it will take to get from A to B. I believe the proposal outlined here will create a much more intuitively navigated network with greater speed and reliability that better serves the future needs of the region.
Last week, WMATA planners released a proposal for a new Metro loop downtown to help relieve capacity issues on the other lines. How might this new line operate? Details are scarce, but we can talk about some possibilities.
The proposal is to build a new Metro line that would loop through downtown DC. Starting at Rosslyn, the tracks would tunnel under the Potomac to Georgetown. They would then follow M Street and New Jersey Avenue to Union Station. The tracks would turn south along 2nd Street to cross Capitol Hill, and then parallel the Green Line under I Street. The loop would complete itself by joining to the existing Yellow Line bridge over the Potomac River.
Metro's proposal indicates that the loop line will be fed by the current Blue and Yellow lines, which would enter at the Pentagon station, and a new "express" line in northern Arlington, which would enter near Rosslyn.
Why build a loop?
Metro's team of planners looked at a variety of solutions to the core capacity issues, including new lines. The alternative they settled on was this loop. Why did they pick a loop?
Essentially, a loop solves most of Metro's problems relatively cheaply. The primary issue facing Metro over the next few decades is core capacity, espcially in terms of train throughput. The Blue/Orange subway through downtown is at capacity, and no more trains can be added.
Untangling the Gordian knot at Rosslyn is the most complicated part of this. What is most clear is that Metro needs a new Potomac crossing near Rosslyn to increase capacity on the Blue, Orange, and Silver lines.
But Metro also needs more capacity around the southern side of downtown DC. Because the Green and Yellow Lines share the 7th Street subway, which is operating very close to capacity, each line can only be increased at the expense of the other line. In other words, WMATA can't add Green trains without subtracting Yellow trains. So Metro also needs a new subway for the Yellow Line.
However, Metro's studies found little need for a new subways outside of downtown based on the expected travel patterns in and density of those areas in 2040.
Essentially, Metro sees the need to build an east-west subway across the north side of downtown and a north-south subway across the east side of downtown. But they don't see a need for the east-west subway to continue east or the north-south subway to continue north.
For that reason, Metro thinks it makes the most sense to just connect the east-west and north-south subways at Union Station and operate them as a loop.
It's not exactly clear how the new subway loop would work, and since this project is decades away from completion much could change. But there are a few likely ways it could operate.
For now, let's only consider the Blue and Yellow Lines. We will discuss the proposed North Arlington Express line in a later post.
Operating an inner and outer loop is one of the obvious choices. In this scenario, one of the lines would run clockwise around the inner loop, while the other line runs counterclockwise around the outer loop.
So Yellow Line trains coming from Huntington would cross the Potomac north of the Pentagon, as they do today. Then they would continue east to Capitol Hill before turning north toward Union Station and going around the loop to return to Huntington via Arlington Cemetery. Blue Line trains would do the reverse.
One of the advantages of operating the Yellow and Blue lines as a loop through downtown is that the loop can actually carry more capacity. If the Blue and Yellow didn't need to share tracks with the Orange and Green lines, each could run at a frequency of 13 trains per hour (TPH) in each direction. And that means that the outer loop would have 13 TPH, as would the inner loop. Since Metro's track capacity is 26 TPH, there's actually room to add two more lines to the loop.
One of the disadvantages, though, is that riders who have a short one-seat ride in the morning have a long one-seat ride in the afternoon. Someone who commutes from Franconia to Georgetown has a pretty direct trip in the morning. But in the afternoon, they either have to face a long ride on the Blue Line via Union Station and Potomac Park or take a two-seat ride by riding the Yellow Line to Pentagon and changing.
"Transforming" loop trains would resolve that problem, though it would be more complicated and difficult to show on the map.
In this scenario, a Blue Line train leaving Franconia would run as far as Pentagon and then continue toward Arlington Cemetery. At some point on its journey, the train would magically transform into a Yellow Line train bound for Huntington. Yellow Line trains would operate similarly, becoming Blue Line trains during their journeys.
This way, a person who commutes from Franconia to Georgetown would have a short, one-seat ride on the Blue Line in both directions. The same would be the case for Yellow Line riders.
For anyone waiting for trains at a station on the loop, trains on the outer loop would always be bound for Franconia, but would have come from Huntington. On the inner loop, trains would be bound for Huntington, but would have come from Franconia.
To avoid confusion, trains bound for the loop would just be signed with their color and a destination of "Downtown." The change of color and destination on the loop wouldn't matter for the passengers on board, since the train would continue around the loop. This is, incidentally, what Chicago L trains do as they arrive at the Loop: they change their headsigns from "Loop" to whichever destination they're headed back to.
Alternatively, trains don't have to "loop" all the way around the loop. Instead, the Blue and Yellow lines could just be interlined on the new tracks.
In this scenario, Blue Line trains would operate onto the new tracks for a certain distance. On the map above, I've shown trains going as far as 4th & Eye, but they could stop at any point along the line (Union Station, for example). Then trains turn back around and run over the same tracks back to Franconia. Yellow Line trains would operate similarly.
Anywhere the two lines overlap would max out the capacity of the new line, just as the other lines in the core are currently topped out at 26 TPH. The reason this is true for the interlined scenario but not for the loop scenarios is because in the loop scenarios, trains run around the loop once. In the interlined scenario trains run over the tracks twice, once inbound and once again outbound.
North Arlington Express trains
As noted above, the new loop would also carry trains from the North Arlington Express line. We haven't discussed those trains yet, but we're going to cover how they might work with the loop in another post soon.
Metro's vision for the future is still decades away, so we have no idea what the final product will look like exactly. It might look like one of the operating patterns shown here, but then again, lots can change in 25 years. But Metro's core is approaching capacity, and expansion is desperately needed. Metro's new vision will set the stage for building the system's next generation.
Come spring, Boston's transit system, the Massachusetts Bay Transit Authority (MBTA), will start offering late-night service on all subway and 15 major bus lines. Like Metro, MBTA may charge higher fares for it. Could this discourage transit use?
Currently, the MBTA shuts down shortly after midnight seven days a week, leaving revelers and workers of unusual hours with no recourse but automotive ones. Under the new schedule, which is a one-year pilot program, Boston will match DC's practice of staying open until 3am on Saturday and Sunday starting this spring.
The announcement was a big deal to native Bostonians like myself, who for years have been frustrated having to choose between staying out late and being able to get home. However, the MBTA is also considering emulating WMATA in another, less desirable way: charging higher fares for late-night service.
The MBTA may consider charging $3 or $3.50 after midnight instead of the usual $2 fare for a train ride. (Boston has a flat fare that does not increase, even during rush hour.) It's not hard to understand why WMATA charges rush hour fares during the wee hours, and why the MBTA might want to follow suit. Late-night public transit is a niche service that only a small subset of the population uses.
Rather than spreading the cost of providing it across the whole transit-using population of greater Washington, late-night riders should have to pay a little more to support their customized service, right? Put another way, if WMATA is expending a constant amount of resources for fewer-than-usual users, each user needs to pay more than usual in order to meet budget.
The problem is that Metro does not apply this logic evenly. If you accept this premise, then fares should actually be lower during rush hour, when huge ridership will never have any problem sustaining even elevated frequency of service.
Instead, the correct pricing principle is one that conforms to supply and demand. Metro rightly charges peak fares during rush hour precisely because that is the busiest time of day; it knows most commuters don't have the choice to be scared away by sticker shock then, and if they are scared away it knows it can absorb the blow.
Higher fares serve to some degree as crowd control; if we have to discourage transit use (which higher fares necessarily do), we ought to do so when transit use least needs to be encouraged. And, most elegantly, people rightly pay higher fares when they are causing the most strain on the system, helping to offset the wear and tear caused by rush-hour crowds.
The flip side of this, of course, is that WMATA, and the MBTA that seeks to emulate it, should charge its lowest fares when the system is least crowded. These are the times when transit use needs more incentives, and of course entrance fees are one of the most surefire ways to manipulate that.
By charging peak fares between midnight and 3am, Metro is creating a deterrence, even a small one, to people taking public transportation. Crucially, peak fares after midnight also do not come with the benefit of extremely frequent trains that accompany rush-hour peak fares.
Of course, there is always the chance that fare manipulation may not have a huge effect on ridership after midnight. In that case, by charging off-peak fares, WMATA would give up revenue it currently relies on. However, it's dubious to think that the laws of pricing dynamics cease to apply after midnight.
Perhaps Metro volume is fairly inelastic during rush hour, when many people have to commute to work no matter what, and when many people feel that they have no other choice but to take the train because of DC traffic and the cost of car ownership. But people certainly do have a choice about whether and how to travel late at night. Lower fares after midnight would not only result in that many fewer Uber trips, but more importantly, they would entice more people to go out.
The upside-down fare system is unfortunate in DC, but in Boston, it could be fatal to its experiment of late-night service. MBTA officials will only make late-night transit permanent if enough people use the service during the one-year trial period. If higher fares deter people from taking the train, the MBTA may very well determine that there is not as much demand for the service as it thought.
This is a realistic worry; $3.50 doesn't seem like much, but it's a 75% increase on the fare Bostonians are used to paying; and for a group of four friends out at a bar, a taxi ride would only need to cost $14 to be an equal or better deal. Indeed, excessively high fares were one reason Boston's previous foray into late-night service, "Night Owl" buses that cost up to $4 one-way, went under. (The Night Owl's failure to attract riders is more proof that late-night ridership is not inelastic.)
Thankfully, because Boston is just reacquainting itself with late-night service, there is still time to avoid these mistakes. DC has fulfilled an important role by serving as the Boston's likely model for late-night service; a fellow cash-strapped system that needs all the overnight maintenance it can squeeze in, WMATA showed the MBTA that late-night service was still possible.
It's probably no coincidence that Boston is adopting the same weekend schedule, but that doesn't mean it should copy DC's methods wholesale. At least in the way late-night fares are structured, Boston can and should do better.
After Metro announced last week that it hopes to build a new loop line between downtown DC and Arlington, many people immediately started comparing it to the other cities that have transit loops. How would Metro's line compare to the others?
There are actually a variety of loop types. Each type tends to operate differently and have different characteristics.
Very few lines operate as a true circle. Glasgow is one of the exceptions. Trains operate in both clockwise and counterclockwise directions. It's a more urban-type circle. Moscow's Koltsevaya line is another urban example. Berlin's Ringbahn also operates as a circle, but it's got a larger radius, and is more of a suburban-style circular line.
Other lines tend to act more as urban distributors, offering a convenient way to turn trains around and to serve larger areas of the central business district. Examples include the Chicago L Loop and loops in Sydney and Melbourne, which serve commuter trains.
Miami also has an urban circulator called Metromover. It operates in a very small circle downtown and mainly serves to distribute commuters from Metro to destinations around the CBD. Two other Metromover "loops" (not shown in the graphic) operate via the downtown loop, but are more linear in nature, continuing to the north and south of the CBD.
Next, we'll look at how a DC Metro loop might work.
To relieve congestion on the Orange and Blue lines and support future growth in the region's core, Metro is proposing a loop line between downtown DC and Arlington. They've just created a map of what the service might look like.
Detail of Metro's proposed downtown loop from PlanItMetro.
The loop is part of Metro's Regional Transit System Plan, which lays out a vision of how the transit system should expand over the next three decades to accommodate predicted regional growth. It incorporates previously studied ways to expand Metro in downtown DC, including new Blue and Yellow lines.
The loop line would go to areas that don't have Metro service, like Georgetown, while adding new connections to existing transfer points like Farragut Square and Union Station. It's unclear how Metro's service patterns would change to serve the loop. Right now, the map shows the Blue, Orange, Silver, and Yellow lines all running on the loop.
WMATA planners are also considering an express line on I-66.
Metro's also looking at a new express line along I-66 between Rosslyn and East Falls Church, which could give the Silver Line an alternate, faster path to downtown DC. This isn't a new idea, either.
What do you think of Metro's loop line?
WMATA thinks talking about reliable buses is boring, asking "Can't we just talk about shoes?" Instead, many riders are talking about how sexist the agency's new ad is.
WMATA placed the ad highlighting its Metro Forward rebuilding campaign at Metro Center. Capital News Service correspondent Lucy Westcott first noticed the ad, which then appeared on DCist. Backlash to it has been fierce, with many Twitter users and anti-sexism group UltraViolet calling the ad sexist and offensive.
A WMATA spokesperson told DCist that "The point of the ad is to get people talking about Metro's massive rebuilding effort by juxtaposing technical facts with a variety of light responses in conversation between friends."
The ad certainly has people talking, but not for the reasons Metro intended.
Yesterday afternoon, subscribers to WMATA's MetroAlerts across the region got an alert advising them of dire service impacts that could affect their commute: a discount on tickets to a basketball tournament.
As you can imagine, many found the spam message irritating, and Twitter lit up with snarky responses. MetroAlerts is, of course, a tool meant to alert riders to disruptions in service. These alerts can be valuable in helping riders choose an alternate route or leave earlier or later to avoid delays. What riders don't expect are spam emails, which have absolutely nothing to do with alerting customers to potential problems, and are a waste of time for Metro's customers.
One of the reasons yesterday's "alerts" message generated some angst on Twitter among riders is probably that many riders, especially on the Red Line, faced some delays during the morning rush. Many riders complained that alerts were slow to come out. And then a few hours later, WMATA sent them an "alert" they didn't need.
No one who visits the MetroAlerts website would expect these spam emails. According to the webpage, MetroAlerts sends out the following information:
- Major Metrorail and Metrobus delays and service disruptions
- Metrobus schedule changes and detours
- Metrorail advisories specific to your line or frequently used stations
- Other changes or enhancements to Metro service and facilities
Why spam riders?
It's not entirely clear what Metro thinks riders get out of this spam. Clearly some MetroAlerts subscribers are basketball fans. Maybe they'll appreciate the discounted tickets to the BB&T Classic. But most riders don't care about the Classic in the least.
These alerts aren't targeted toward anyone, like basketball fans, for example. It's just mass advertising sent out to a large group in the hope that someone will find it helpful.
It doesn't appear Metro gets anything out of it, either. I asked spokesperson Dan Stessel whether the organizations promoted in the alert compensate Metro for the use of their email lists and alert system. He says Metro doesn't get paid for these. According to Stessel, "these are in-kind promotions for the benefit of riders; generally barter only."
Perhaps what Metro ends up getting out of this is some additional ridership. Some of those riders that got the alert that do like basketball might buy tickets (at a discount!) and then take Metro to the event.
But using untargeted alerts like this seems like a very crude way to promote ridership or create value for riders.
After all, what percentage of MetroAlerts subscribers are basketball fans? Even if it's a high number, like 25%, what percentage will actually buy tickets to this event? For the rest of the subscribers, this is an irritant. And it's more likely to make them unsubscribe from MetroAlerts.
I also asked Stessel what discussions Metro staff had about the appropriateness of using the alerts system for promoting unrelated events. He didn't answer directly, but he did point out that people have to opt in to receive the promotions.
It's fairly easy to opt out. Riders just need to log into their account and uncheck the box labeled "promotions." But it's not a matter of opting in. The "promotions" box is checked by default (along with the "alerts" and "advisories" boxes).
Anyone who registered for MetroAlerts before the addition of "advisories" and bus alerts would have registered before Metro added "promotions." Stessel said that subscribers were notified by email of the change at the time.
Alerts are for alerting, not advertising
In the world today, we have many different forms of communication for emergency alerts. Most state departments of transportation have electronic signs, especially in urban areas. They're frequently used to broadcast messages about travel times, traffic accidents, construction, and Amber Alerts. But they're not used to advertise discounted basketball tickets.
Spamming riders with promotions unrelated to Metro service reduces the value of MetroAlerts and it wastes riders' time. It also creates the perception that Metro is out of touch with what riders want. After all, if people liked spam, email services wouldn't have created spam filters.
In the future, transit riders need timely alerts that actually help riders during disruptions so they can have a smooth commute, rather than junk mail.
Following a week of terrible Red Line Metro service, WMATA is offering refunds to riders who were inconvenienced. That's a nice gesture, and perhaps it's appropriate as a special measure when the same customers are hit with so many unexpected delays in such a short time. But in general, refunds are not the answer to Metro's woes.
Yes, customers deserve to receive the service they've paid for, and yes, it's satisfying to hold WMATA financially accountable when it doesn't meet expectations. Those are the arguments in favor of refunds, and admittedly they're compelling.
But if WMATA actually gave refunds every time there was a breakdown (which as we all know is unfortunately frequently), what would the side effects be? As always, follow the money.
For one, it would slow Metro's reconstruction efforts, by diverting funding away from maintenance work. This painful period of single tracking and service interruptions would likely last years longer.
That's bad enough, but consider WMATA's likely response the next time there's an unexpected rush hour breakdown. Instead of allowing thousands of passengers to continue entering the system and paying fares after the breakdown starts, only to be due refunds later, Metro would probably start to simply shut down entire lines. After all, you can't demand a refund if you were never allowed to enter in the first place.
So on days like Wednesday, instead of bad service Metro riders might be left with no service at all. I don't think that's better.
At some point WMATA does need to be held accountable, so perhaps it was appropriate to offer refunds this week. But Washingtonians should neither demand nor expect a refund every time we have to wait for single tracking. Ultimately, improving Metro service back to the point where these delays don't happen so often is what we really want. Frequent refunds wouldn't help that goal.
Cross-posted at BeyondDC.
- How might the new Metro loop work?
- Want the urban lifestyle? DC's best corner is...
- More roads won't solve traffic on I-95 in Northern Virginia
- Can Metro add capacity without a downtown loop?
- See what Montgomery BRT could look like
- Can we build up around MARC stations?
- What's the oldest continuously named street in DC?