Greater Greater Washington

Posts about WMATA


Fifteen big shutdowns, and many smaller ones, will get Metro repairs on track

The plan is finally here. Metro is launching a plan it calls "SafeTrack" to replace significant portions of the system's rails, fix numerous safety problems, and bring the system closer to a state of good repair. Riders will face weeks-long periods of single-tracking and station shutdowns for the next year.

Photo by Elvert Barnes on Flickr.

The Metro system is 40 years old and needs massive repairs. Today, trains run in service 135 to 168 hours each week, leaving little time for comprehensive maintenance. The SafeTrack plan will squeeze in work which would otherwise take three years to do at current rates, including clearing some urgent NTSB safety recommendations by the end of this month and others by the end of summer.

When the actual shutdowns will happen

There will be 15 work "surges." Some involve shutting segments of the system down entirely for a week or two. Others involve continuous single-tracking for multiple weeks (as much as 44 days), which means single-tracking even during rush hours.

June 4-19Franconia to Van DornContinuous single-tracking
June 20-July 3Greenbelt to College ParkContinuous single-tracking
July 5 (10 pm)-July 12National Airport to Braddock RoadFull shutdown
July 12-19National Airport to Pentagon CityFull shutdown
July 20-31Greenbelt to College ParkContinuous single-tracking
August 1-8Takoma to Silver SpringContinuous single-tracking
August 9-19Shady Grove to TwinbrookContinuous single-tracking
Aug. 20-Sept. 6Eastern Market to Minnesota/BenningFull shutdown
Sept. 9-Oct. 21Vienna to West Falls ChurchContinuous single-tracking
Oct. 9-Nov. 2NoMa to Fort TottenFull shutdown
November 2-12West Falls Church to East Falls ChurchContinuous single-tracking
Nov. 12-Dec. 5East Falls Church to BallstonContinuous single-tracking
December 6-24Pentagon to RosslynFull shutdown
Jan. 2-March 7Friendship Heights to Medical CenterSingle-tracking after 8 pm
March 6-14West Falls Church to East Falls ChurchContinuous single-tracking
April 16-May 8Braddock Road to Huntington/Van DornContinuous single-tracking

These "surges" will affect riders beyond these zones. Outside the single-tracking or shutdown areas, the capacity of each line will still be reduced, so there will be fewer trains on any lines that run across that segment. You can see a detailed list of affected sections in the full presentation.

The "surges" also aren't the only piece. Metro is trying to minimize full shutdowns by also doing more work at nights and on weekends. Right now, when a segment is shut down for a night or weekend, that starts at 10 pm or midnight; now, it will start at 8 pm, like on the Friendship Heights to Medical Center segment in early 2017.

Metro will stop doing maintenance during July 4, the Presidential inauguration in January, and next year's cherry blossom season, but nothing else. Even if there's a street festival or other special event in an area scheduled for maintenance, Metro will stick to its maintenance schedule.

Photo by Hannah Rosen on Flickr.

Why such long windows?

Shutting down a piece of track for weeks is the only way to do some maintenance that WMATA has never done since it opened. Metro will completely replace rails, ballast, and the substructure which the rails lay on top of.

A significant number of the wooden rail ties in the system are original from when the system opened. Some have been replaced here and there as they deteriorated too much, but the new program means the agency will be able to replace large numbers of the ties as well as fasteners, rail, and other track equipment.

WMATA prime mover. Image from WMATA.

The only way to fully rebuild a section of rail is to keep trains off it for a long time, which is why single-tracking would have to last for weeks.

Other maintenance will address NTSB findings that the "boots" on the power-supplying third rail aren't always properly fastened and that watertight seals need replacing.

How people can get around in the meantime

Metro plans to have a dedicated fleet of up to around 40-50 buses once the track work starts, in order to help move passengers around closure and single-tracking areas, to minimize impact on surrounding roads. Even still, passengers will certainly see delays.

Dedicated traffic control officers, dedicated bus lanes, and other changes to the traffic pattern could help keep the buses moving. Those are up to the state, county, and city departments of transportation. Metro will depend on jurisdictions to step up, said Barbara Richardson, Metro's chief of external relations, "because we need to think creative and differently about how to move people throughout the region."

What Wiedefeld announced today is just the draft plan. He's releasing it to the board, the local DOTs, and the public for input. He'll finalize it next week, and maintenance will start in June.

Photo by brownpau on Flickr.

In the long run

After the year, the system will be in better shape, but that doesn't mean no more shutdowns, said WMATA spokesperson Dan Stessel. Riders got used to having no single-tracking and no shutdowns when it was brand new, but that's because the necessary maintenance wasn't being done.

What exactly the maintenance plan will look like beyond a year isn't certain, said Stessel, but riders shouldn't expect no shutdowns at all.

Metro is also exploring whether to permanently roll back weekend hours to midnight instead of 3 am. Stessel pointed to BART, which doesn't have late weekend hours, and said that policy has preserved longer maintenance windows for them to keep up with repairs. Whether Metro makes this change or not will be up to the board.


Metro's track work schedule dug the system into a hole

Metro is currently facing a huge backlog of work on aging and broken tracks, partly because its way of scheduling the work isn't working. The agency is announcing a new strategy for fixing its tracks on Friday, meaning there's a huge opportunity to get track work right.

Silver Spring pocket track. Photo by Ben Schumin on Wikipedia Commons.

There currently isn't enough time to do maintenance

In 2011, Metro started MetroForward to fix the system's infrastructure. The plan was to spend a large part of the project's $5 billion budget to fix the rails themselves. Five years and $3.7 billion later, the track work backlog hasn't gone anywhere

WMATA employees, the Tri-State Oversight Committee, and the Federal Transit Administration have all said that doing track work only during overnight hours isn't enough. The FTA also found that there is less time than before to perform track inspections that could find defects like cracked rails, defective fasteners, and third-rail insulator issues.

WMATA General Manager Paul Wiedefeld is now mulling over options to fix the way the agency does track work, hoping to cut the current backlog and bring everything up to a state of good repair.

This chart from WMATA, made in 2011, shows why there isn't enough time to get all the maintenance work done overnight. There are three periods—5-hour, 7-hour, and 56-hour work windows—in which work could be done. Green signifies the usable time and red signifies time lost to setup/teardown and other non-work activities. The shorter the maintenance period, the less productive the actual work time is.

Track usage production ratios. Image from WMATA

Metro's current track work strategy doesn't work

Metro usually performs single-tracking in several places each weekend while workers take a track out of service. Trains often run less often because of this, sometimes every 24 or 26 minutes on average (in reality, trains may come as quickly as every couple of minutes on one side of the single-tracking area, or as infrequently as every 40 minutes on the other).

Weekend single-tracking is rough. It's caused a vast drop in weekend rail ridership, with some saying the system is unusable on the weekend. The reduction in service and increased unpredictability due to single-tracking is a big reason for this.

Full weekend shutdowns may be the way to go

As opposed to a single evening that provides 2.5 hours usable trackwork time (or 4.5 hours if the work starts at 10pm instead), a full weekend of track work from 9 pm Friday to 5 am Monday can provide up to 48.5 hours of usable track work time and allows access to both tracks, increasing the useful time ratio from 50 percent to 86 percent. The drastically-increased amount of unbroken time that workers have available on the track means that much more work can be accomplished.

Shutting down sections of track on the weekend to perform maintenance work not only is more convenient for passengers (rip the band-aid off instead of a slow pull), but also more beneficial to the maintenance personnel performing the work. Instead of being constrained to one track to work in, workers can spread out and move around the site easier, there's no need to watch out for trains moving on the opposite track, and there's a good potential more types of work can get done through better utilization of the work area.

Buses cost money though, and they would be needed for sectional shutdowns to ferry passengers across. Full shutdowns cost somewhere around 15 percent more than single-tracking according to one project. But if one method has a chance of retaining ridership and will accomplish more work in a single weekend than the other might in two or more, it may be worth it. Cost is no longer the driving factor for when and where track work occurs, as new Wiedefeld has seemingly made clear.

Not all stations and track can always be shut down for a weekend at a time though, so the agency may have to revert to other less-efficient track work strategies:

Short periods of track work are incredibly inefficient

During the week, the Metrorail system closes at 12 am, and opens at 5 am. While this provides five hours of "off" time, it really only gives maintenance personnel 2.5 hours or less to spend on the tracks. Setting up and tearing down track work is incredibly time consuming, as the process can include assembling workers and equipment, moving any heavy track machinery (prime movers) into the area, making sure the track is clear, setting up "shunts" to mimic a train in the area so the central rail routing system knows not to send trains into the work area, and more. An hour or more can quickly be used just to set up the work area.

If everything is set and placed, work can start. But when things don't go as planned, time can be easily eaten up, as the FTA noted in their Safety Management Inspection review published last year:

On the March 23 [overnight] shift, for what should have been a simple T-bar replacement...the contractor was still not able to get access onto the work site until 2:00 a.m., and the contractor had to start clearing the site at 4:00 a.m., leaving only two hours of productive time...The impacts of this limited work window can be exacerbated by communication and logistical challenges. For became clear that the Power branch work crew had the wrong size T-bar...and there was confusion whether the ATC department had been consulted in the planning of the work.
One hour before the crews need to wrap up (i.e. 4am in the morning if the system opens at 5am), they stop working so everything can be cleaned up so passenger trains can use the track. When this work runs over, passenger-carrying trains can be disrupted and you may see a tweet saying that there has been "late-clearing track work."

Mid-day track work is not productive and causes passenger frustration

Based on the track work production numbers from the chart, mid-day track work is incredibly time-consuming, disruptive to passengers, and doesn't provide much benefit. A mid-day maintenance window might be from 10am to 3pm (5 hours), which provides 2.5 hours of usable track time in the best case scenario. But if there are any issues in the morning and rush hour runs over past 10, the track work may wait, and that eats into the usable time.

What starts out as 2.5 hours of usable time could get whittled down to 1.5 or 1, or even less, on an especially bad day. While this work is going on, trains carrying passengers get to single-track around the area, waiting at least 10 minutes on either side of the stretch before going through.

Start track work earlier in the evenings

One option that helps provide more track time is to start track work "early outs" earlier in the evening at possibly 8pm, and end them later, possibly up to 6am. Current early outs usually start at 10pm and end by 5am. This provides 7 total hours of time for work, or around 4.5 usable hours of track time. The two-hour-early start provides two more hours of usable track time, which is valuable. Starting even earlier and ending later could provide up to 10 total hours, or 7 hours of track time, over double what crews would get with a typical overnight work session.

While early outs let track work start earlier, they require a delicate balance as well. If they start too early or end too late, they could severely impact peak rush periods and cause delays and rider frustration. More work could get done, but if done poorly could continue the ridership decline.

Longer shutdowns are an interesting possibility

Metro has never performed a shutdown longer than a weekend in recent history, but it seems Wiedefeld is headed in this direction. A shutdown from 9pm on a Friday to 5am the Monday a week later would provide about 200 track hours of usable time of 224 available, meaning 90 percent of the week is usable for track work. 200 hours of track time would be the equivalent to over 4 weekend shutdowns, 44 nights of work with 10pm early outs, or up to 80 regular overnight work sessions. Track work could be completed easily and quickly for both tracks in the work area.

At the same time, shutting down a station or two during the week is also hard. Buses and drivers would be needed to shuttle passengers around the shutdown area, more than Metro may have available. Replacing a rail car that can comfortably hold 120 passengers with 40-50-person buses means lots of road traffic, buses, and drivers. Extended shutdowns requires planning with jurisdictions to ensure alternate transportation for Metro riders goes as smoothly as possible.

This assumes the work is being done properly and coordinated well

Good use of track time means that workers are well-trained for what they need to do, equipment is available, and contractors are prompt and responsive. Audits have found that this is not always the case, and that new workers are liable to receive deficient training. At other times equipment has not been available or wasn't brought to the work site, resulting in large amounts of time wasted while waiting for the necessary parts.

Similar to how some electrical crews are receiving specialized training to better handle heavy-duty power cables, perhaps some track and structures crews may require the same training to verify that work is being done properly or to help ensure training of others. Metro's safety and track departments need to step up their quality control and assurance game, too: track work needs to be inspected by independent analysts who know exactly what they're looking at. Independent workers need to be in the tunnels checking that inspections are being performed and equipment is being installed properly. Otherwise, performing the track work is a waste of Metro's time and our money if the equipment has to be ripped out and replaced again.

Wiedefeld is releasing his track work strategy tomorrow, but as always, communication with customers again will be key to make this plan successful. Riders not only want to know what work is being done, but why the work needs to be done, what specifically is being done backed up with photos or video, and that the work is being done properly. Working with all interested parties, especially riders, is the only way the upcoming trackwork program can be successful while not alienating the very riders paying for a large portion of the agency's budget.


Here are the answers to whichWMATA week 82

On Tuesday, we posted our eighty-second photo challenge to see how well you knew Metro. I took photos of five Metro stations. Here are the answers. How well did you do?

This week, we got 27 guesses. Nine got all five. Great work Peter K, JamesDCane, Justin..., Chris H, AlexC, Dillon the Pickle, FN, Solomon, and Stephen C!

Image 1: Tysons Corner

The first image is the view of Tysons Corner station from the plaza adjacent to the Vita building. Given the unique Silver Line architecture, you should have been able to easily narrow this down to the two "Gambrel" style stations. But Tysons Corner station isn't in a median, like Greensboro is, so this can't be Greensboro. All but one of you got this one right. Great work.

Image 2: Fort Totten

The second picture shows the top of a memorial plaque in the mezzanine at Fort Totten. The "Y OF" that is visible is part of the phrase, "in memory of," and is a memorial to the nine people killed in the 2009 Metro crash just north of the station.

Additionally, the windows and angles here are indicative of the mezzanine shape of Fort Totten. From this vantage point, we're looking toward the two escalators connecting the mezzanine to the Green/Yellow platform.

Twenty knew the correct answer.

Image 3: Van Dorn Street

The third image shouldn't have been too hard if you know the architecture of the system. The presence of three Blue Line trains on the PIDS tells you that this is almost certainly a Blue-only station, of which there are only three in the system. This picture was taken on a Saturday, but even though you didn't know that, there are times (weekends and off-peak), when Van Dorn and Franconia are not served by the Yellow Line.

Two of the three Blue-only stations can be easily eliminated. Arlington Cemetery is depressed in an open cut (so the trees wouldn't be visible) and doesn't have a canopy at all. Instead, it's only covered where Memorial Avenue crosses it. Week 8 gives you a sense of what that looks like. Franconia/Springfield, on the other hand has a very different canopy and the PIDS signs have a different format at terminal stations (BLUE LINE | LARGO CENTER | LEAVING 3 MINS).

But even if there had been only one Blue Line train on the board, you still could have solved this. That's because the canopy visible here is a "Gull I" design. And the only Gull I station served by the Blue Line is Van Dorn Street.

Twenty-five figured it out.

Image 4: Wheaton

The fourth image was a little trickier, and required you to take a second look to figure out that this was Wheaton. Many of you went with your first instinct, but closer inspection should have revealed this to be a twin-tube station. One clue is the presence of "can" lights hanging from the vault, which aren't present at the higher single-vault stations.

The perspective here is clearly looking through the doors of an elevator. Some downtown stations do have side platforms with the elevator in an alcove off to the side like this. But all of those stations have the "waffle" design, not the taller coffer "arch"-style. None of the "Arch I" or "Arch II" stations have side platforms. And that means this has to be one of the twin-tube stations.

It can't be Forest Glen because, as several of you noted, the elevators there all land in a common lobby and are farther from the tracks. At Wheaton, however, the solitary elevator lands not in the escalator lobby, but in an alcove at the far northern end of the Shady Grove platform.

Ten came to the correct conclusion.

Image 5: Capitol Heights

The final image was even more challenging, but there was enough information to figure out that it's Capitol Heights.

Like with image 3 above, you can tell that this station is served only by the Blue and Silver Lines (since the Orange Line isn't listed on the sign). There are only two underground stations that are served by the Blue and Silver, so even without additional information, you could have made a guess with a 50/50 probability of getting the right answer. Some of you did that and got lucky.

But there was a way to be 100% certain, and it involves knowing that while Benning Road and Capitol Heights are nearly identical, they're mirror images of each other. In week 56 we also ran a set with a similar signage clue and noted in the answers post the difference between the two stations.

At both stations, the single mezzanine is at the far end of the platform. At Benning Road, the mezzanine is at the east end. At Capitol Heights, it's at the west end. That means that when you descend to the platform at Capitol Heights, you're facing east. And if you're facing east, trains going eastbound to Largo are on your right, and trains going west toward downtown are on your left.

One final note: The reason you know this sign is at the bottom of the escalator when you arrive is because this is a fairly typical application of WMATA's signing in this case, since this is a decision point and because anywhere else on the platform, the column would also include a strip map of farther stations reached on the appropriate track.

Nineteen came to the correct conclusion.

Thanks for playing! We'll be back in two weeks with our next quiz.

Information about contest rules, submission guidelines, and a leaderboard is available at


The feds aren't helping on Metro safety, says DC transportation chief

Metro has to do better on safety, local and federal governments, riders, and many others all agree. But while the federal government is pushing for safety, some say it's also a part of the problem.

Unhelpful image from Shutterstock.

At a regional "summit" on Metro's future on March 30, Leif Dormsjo, director of the District Department of Transportation (DDOT), blasted the Federal Transit Administration's relationship to WMATA and other transit systems around the nation.

Nobody can call Dormsjo a WMATA "apologist." From the moment he took over DC's transportation agency, he's been calling for reforms at Metro. A year ago, he said, "WMATA needs to hire and fire better, manage its capital projects better, follow accounting principles better, and communicate with the public better."

Dormsjo also has called for Metro to focus on safety and reliability. At the summit, he praised new General Manager Paul Wiedefeld for emphasizing just those two factors and getting "back to basics." Last year, when former safety head Jim Dougherty chafed at FTA oversight, Dormsjo rebuked him, defending the FTA's involvement.

But when it comes to fixing problems, he said in March, the FTA is not acting like a partner.

"Every other mode of transportation has better safety oversight relationship with the federal government than transit and the FTA," he said, listing freight, aviation, highways, and other forms of transportation, each of which has its respective "modal administration" inside USDOT. Those departments oversee safety but also are "partners" in fixing problems, said Dormsjo.

Not FTA. Instead, he said, FTA sits back and criticizes transit agencies for missteps but doesn't try to help find solutions. "It's always easier to knock someone down than pick them back up," he said, and FTA is not a "collaborative partner."

"I wish the current administration would extend their hand to the nation's transit system," said Dormsjo. He suggested FTA help work out solutions to problems and then apply them to other transit properties across the nation.

Photo by Gen Kanai on Flickr.

This is a common complaint

Other transportation officials have said similar things privately before, and about more than just safety. FTA also oversees the procurement procedures of transit agencies and monitors their actions to make sure they meet federal rules for grants.

One official, who spoke on condition of anonymity because he wasn't authorized to speak publicly, relayed the story of a transportation conference where the FTA's representative on a panel kept lecturing agencies on how to achieve compliance with federal rules. That FTA speaker, he said, never talked about helping agencies reach that compliance.

If the FTA were less "rigid" in its interpretation of regulations, said Dormsjo at the March 30 panel, that could save transit agencies a lot of money in things like buying vehicles.

A former transportation executive in the region, who also wasn't willing to speak on the record, said, "I've never found the FTA to be helpful. They are custodians of money, not advocates for projects. They never act like they are trying to help you; they usually act like you're being difficult."

Sometimes it seems as though the FTA treats the nation's transit more like a reality TV show than a vital public service. They sit at the judges' table, watch the agency's performance, and lob criticisms.

The FTA put WMATA in a sort of financial penalty box, called "restricted drawdown," in 2014 after discovering weaknesses in financial controls. But over two years later, the penalty hasn't been lifted.

Maybe that's for good reason, if Metro's internal controls aren't satisfactory, though that's not clear from public information; board chairman Jack Evans has claimed WMATA has improved and deserves to be let out of the penalty box. An Inspector General's report found that "inconsistencies in the way [FTA] regional offices enforce the rules ... meant [WMATA] faced longer delays in receiving reimbursements than the other two transit agencies examined."

Even if WMATA isn't ready to go off restricted drawdown, the FTA ought to make it a high priority to help WMATA get there, in whatever way it can. Perhaps FTA really is doing that (it's not in the public record), but if it is, based on off-the-record statements, that's not typical.

Nobody questions that Metro has to do better on safety and management procedures, and it's right for FTA to push for improvements. Problems, like the fire last week near Friendship Heights, are Metro's failings and Metro's responsibility to clean up. However, to do so requires teamwork from every other stakeholder as well, and the federal government needs to see its role that way, not just as a judge but a partner.


Think you know Metro? It's whichWMATA week 82

It's time for the eighty-second installment of our weekly "whichWMATA" series! Below are photos of five stations in the Washington Metro system. Can you identify each from its picture?

Image 1

Image 2

Image 3

Image 4

Image 5

We'll hide the comments so the early birds don't spoil the fun. Please have your answers in by noon on Thursday.

UPDATE: The answers are here.

Information about contest rules, submission guidelines, and a leaderboard is available at


Metro needs to do a better job fixing rail cars

Blue-chip consultants at McKinsey & Company recently released a preliminary report on Metro performance. It dives into paratransit, the HR back office, parking, and other issues, but the one that most impacts commuters each day is rail car maintenance. If Metro doesn't do a better job of fixing its rail cars, it's going to keep losing riders.

Metrorail on-time performance. On-time performance can vary and drop more when there are fewer rail cars available for service. Graphic by the author, data from WMATA.

Get better at fixing the rail cars, improve daily service

According to the report, 63 percent of all rail line delays are due to railcar failures. 36 percent are caused by trains not being dispatched, and 27 percent are caused by trains being removed (offloaded) from the line. These delays and car failures lead to fewer trains available for passengers to board and more crowding, making the system less enjoyable—or functional—to ride.

The five most common rail car failures, according to the study, were propulsion, door, HVAC, brake, and pneumatic subsystems. Any one of these failures might cause a train car not to be dispatched from a rail yard or could lead to a train being offloaded and taken out of service during the day. Since rail cars operate in married pairs, a problem with one means that both in the pair would have to be pulled off the line.

One of the problems identified by the report is the high number of repeat failures in rail car components. McKinsey stated causes of this include below-average quality control/quality assurance processes, repair protocols that don't fully determine the root cause of the issue, and maintenance staff that are less experienced than in the past due to turnover. Any one of these issues identified would be a problem, but the three put together mean that rail car reliability is likely nowhere near where it should be.

Graphic showing repeat failures of rail car components. Image from WMATA via WAMU.

Metro is building a new two-mile test track near Greenbelt station primarily to ensure the new 7000-series rail cars work as expected when they're received from Kawasaki, but it could also double as a certification track for cars post-repair. As it stands today, rail cars do not undergo a "checkout" period to verify that the issue they were in the shop for was actually fixed.

Once the test track is finished, rail cars that have had their brake or propulsion issues worked on at the various yards could be checked out for additional issues and for an extra quality assurance step. Verification testing to check if the work was done successfully could be one way to help find recurring issues when they happen, before the cars are placed back into revenue service.

Railcar availability issues caused 36% of delays

The report noted that over a third of all rail delays were caused by not having enough cars available to run in service. Whether the cars are out for work or sidelined because WMATA doesn't have the parts needed to fix them, the lack of cars means fewer trains run for passengers to use.

The graph below shows how many of Metro's trains didn't operate in each month of 2015, as well as some of the top delays noted in Metro's daily service reports. The number of trains not dispatched only dips below 100 a few times, around October and November. Given each 6- or 8-car train can hold anywhere from 800 to 1000 people each, a train that's not dispatched from a rail yard leads to less capacity to move people and creates gaps in service causing passengers to end up waiting longer.

Stacked graph representation of the top 5 delay-causing rail issues according to WMATA's Daily Service Reports. Graphic by the author, data from WMATA.

The rail dispatching issue is relatively new, having only started becoming a more significant issue in 2014 and 2015. One reason the report points out that might cause this is the sometimes-significant delays in the parts ordering process, in part due to FTA requirements. At times there have been up to around 70 cars out of service awaiting work since the parts were not in stock.

Last 4 years of delays for the top 6 causes of rail delays, according to the Daily Service Reports. Graphic by the author, data from WMATA.

One of the main priorities the report wants Metro to focus on is to get better at figuring out the root-cause of issues when they occur, and use that information to help fix and prevent it from recurring on other rail cars. Getting better at fixing cars the first time would get them in and out of the shop bays faster, and keep them from coming back. More focus here would help keep the related issue of mechanic experience under control.

The track side of the system has had a spotlight pointed on it in recent months, primarily by the FTA with hundreds of inspections being performed to make sure the tracks are in good order. But Metro needs to be just as focused on keeping their rail cars in working order - not just to say that they did, but to reduce the number of delays that passengers feel, and make the experience of riding Metro better. Focusing on improving rail car maintenance won't be quick, but the dividends it will pay by providing better service to passengers will make it worth it.


Metro's new bus prediction system is pretty accurate, but leave yourself some extra time

Metro recently swapped out its seven-year-old bus predictions system for a new one called BusETA. Last week, fourteen Greater Greater Washington contributors and staff audited BusETA for accuracy. Overall, the system performed well enough, but buses sometimes came earlier than predicted, and "ghost buses" are still real.

Image from WMATA.

Real-time departure information has totally changed bus travel, as any rider with a smart phone can now triangulate the real location of the bus. That means the ability to minimize wait times or choose a travel mode with more confidence; we're no longer reliant on a printed schedule that can be shredded by congestion, incidents, events, breakdowns, or weather.

WMATA's predictions come via a combination of real-time data from GPS transponders on the buses and computer models that predict bus arrivals using historic data about traffic patterns. Once they're made, WMATA publishes predictions on its website. The marketplace for independent prediction apps like Transit Tracker and Citymapper is also pretty crowded, as WMATA publishes tools that any developer can use.

A recent change in who provides WMATA's prediction technology means NextBus, the widely used but proprietary system, is out, and BusETA is in.

Bus lines included in the audit: F4, H8, 16G, 16H, 70, 64, P19, S9, E4, 74, 96, 52, 53, 54, C21, C22, 30S, 31, 33, 37, N2, J3, X2, 4B. Map by the author.

Our contributors recorded how BusETA did one morning

For the audit, which we did on Thursday, April 14th, each participant went to a specific bus stop and called up the prediction for the stop by the stop's unique ID number. While BusETA will also give predictions by bus line, we audited only the bus stop interface. Each participant took a screenshot of the prediction and then recorded the actual arrival times of the predicted bus(es). Our participants audited a total of 27 buses covering 24 lines at 17 locations from 6 to 9 am.

Here's how close each bus came to arriving when BusETA said it would:

A positive error value means the bus was early. The variance of the error clearly increases as the prediction time increases (the further the bus is away, the worse BusETA is at predicting the arrival time).

However, the latest bus was an F4 bus that Gray Kimbrough watched from across the street: It was nine minutes late as it passed through Silver Spring, heading away from downtown. It is notable that this bus was audited at a location near the end of its run. It could be that the further into the bus' run the stop is, the less accurate the prediction is because there have been more opportunities for the bus to encounter delays.

Buses were early as often as they were late. One contributor missed his bus because the prediction told him he had five minutes, and he actually only had three.
Anecdotally, it seems like BusETA might under-predict bus arrivals more often than NextBus did (i.e. the bus shows up 'early').

If so, this is a major problem, because when you miss a bus by only one or two minutes, you have to wait the entire headway of the bus line for the next one, which is the worst-case delay scenario. Based on the results of our audit, I'd recommend factoring in a three minute margin of error when using BusETA.

There were still ghost buses (either buses that were not predicted, or predicted buses that didn't show). Jonathan Neeley got an H8 prediction, but when he refreshed it two minutes before it was supposed to arrive, the bus had disappeared (the actual bus did arrive a moment after, however). Steven Yates read that the 74 was 11 minutes away...just as it pulled up.

Overall, though, BusETA worked more often than it didn't. Brendan Casey said that for his commute, BusETA is far more accurate than Transit or Citymapper.

BusETA prediction

The BusETA technology is different, and likely better

NextBus had a lot of accuracy problems. WMATA's switch to BusETA means it has joined the open source OneBusAway project, which is also used in Atlanta and New York City. That means that all the old apps that used the NextBus standard don't work for DC any more.

The switch to an open source standard via BusETA should promote innovation and help interested parties understand how and why various prediction apps are working. Anybody can contribute back and improve the OneBusAway project since the code is freely available as an open-source project.

This potentially makes the software quite powerful: If someone wants to write a feature in, they can pull the freely-available code, edit it, and publish it back for approval. This freedom allows the application to be more feature-rich than it might otherwise be, and be developed faster than a typical commercial application.

In the long run, an open-source standard will hopefully mean more and better apps for DC.

Thank you to Gray Kimbrough, Chris Slatt, Abby Lynch, Jim Titus, Sebastian Galeano, Steven Yates, Brendan Casey, Bryan Barnett-Woods, Ronit Dancis, Andrea Adleman, Angela Martinez, Jonathan Neeley and Sarah Guidi for participating in this flash audit.


Here are the answers to whichWMATA week 81

On Tuesday, we posted our eighty-first photo challenge to see how well you knew Metro. I took photos of five Metro stations. Here are the answers. How well did you do?

This week, we got 50 guesses. Sixteen got all five. Great work JamesDCane, AlexC, Peter K, Eric P, Stephen C, ajw4, Solomon, Sand Box John, FN, Chris H, Andy L, dpod, DavidDuck, merarch, Travis Maiers, and We Will Crush Peter K!

Image 1: Medical Center

The first image shows a platform pylon at Medical Center. There are five stations in the system with "center" in their name: Federal Center SW, Largo Town Center, Medical Center, Metro Center, and Mount Vernon Square/Convention Center.

However, Metro Center and Federal Center have waffle-style vaults, Largo is outdoors, and the "Center" in Mount Vernon Square wraps onto a second line of text on the pylon. That leaves Medical Center. Additionally, you can see that the vault here has only four coffers, making it Arch I, present only on the Red Line's Shady Grove end.

Forty-four knew this one.

Image 2: Branch Avenue

The second image shows the southern terminal of the Green Line, Branch Avenue. There are only four high peak stations, which narrows the field considerably. The next train indicator is a hint that this station is a terminal station. It's true that these signs are also present at some stations that used to be terminals, however none of the high peak stations are former terminals.

That leaves the two current terminals, Franconia/Springfield and Branch Avenue. However, Franconia is not in an open cut with retaining walls on either side. Additonally, the trapezoidal caps on the columns supporting the canopy are only present at Branch Avenue. That attribute was featured in week 66.

Forty-three got the right answer.

Image 3: Cheverly

The third image shows a view looking northwest from the Vienna platform at Cheverly. This image was harder than I anticipated, given the clues present.

The presence of single-level Amtrak equipment narrows this down to one of the stations along the Orange Line between Cheverly and New Carrollton, the Blue Line between Braddock Road and Franconia, and NoMa. The only Amtrak service along the Red Line is the Capitol Limited, which uses double-decker equipment.

But another clue is that the station has side platforms, which you can tell from the perspective. Finally, it's an outdoor side-platform station with a mezzanine above the tracks. Only Cheverly fits the bill.

Only 30 got the correct result.

Image 4: Vienna

The fourth image shows a view from the platform escalator into the mezzanine at Vienna. The main clue is the sign overhead, which directs passengers to the north and south side access roads, with buses and parking on both sides, a situation only present at Vienna and Wiehle Avenue.

The skylights are indicative of a general peak station with a mezzanine above the tracks. And the design here is clearly older than the more modern touches on the recently-opened Silver Line.

Thirty-nine guessed correctly.

Image 5: King Street

The final image shows a station pylon at King Street. This one was a little tricky, I'll admit, which is why I put it last. I order the five images so that they increase with (what I believe to be) difficulty from first to last.

The entrance pylon at the corner of King Street and Diagonal Road was never updated after the Blue Line was extended from National Airport to Van Dorn Street in 1991 (the Yellow Line was extended from National Airport to Huntington, including King Street, in 1983).

Many of you were either familiar with the absence of the blue stripe on this pylon or correctly deduced that the letter just below the yellow stripe was a "K". King Street is the only station that starts with the letter K.

I know it was tempting to guess Huntington, since the first letter clearly wasn't an E, and only Eisenhower Avenue and Huntington are served only by the Yellow Line. However, this erroneously labled pylon is distinctive, and I left enough of the K, I thought, for this to be fair game.

Twenty-four of you figured it out anyway. Great work!

Thanks for playing! We'll be back in two weeks with our next quiz.

Information about contest rules, submission guidelines, and a leaderboard is available at


Batteries on tracks could power trains and save Metro money

Metro trains can sometimes generate energy when they brake, but right Metro doesn't have a way to capture and use it. A recent WMATA study says the agency would need to buy less power from Pepco and Dominion if it installed battery systems along its rails.

A traction power substation. All images from WMATA.

In 2010, WMATA staff presented the idea for an "energy storage demonstration project" to the agency's board of directors. Metro spends nearly $50 million dollars per year on power to make trains move, so any savings could be substantial. The study was performed and tested around 2013, but the FTA only recently published the results.

A brief background on train braking

Metro trains have two types of braking systems: dynamic and friction brakes. Dynamic braking on a train car is similar to what hybrid cars do when the driver steps on the brake: A motor turns when the brake pedal is pushed, and essentially acts as the brake to slow the car down while also generating an electrical current. That current gets fed back to the car's battery, which lets you drive farther.

Dynamic braking is the focus of Metro's demonstration project. When it's used, the energy generated gets "dumped" back into the third rail, which is what transmits power for trains up and down the tracks. Usually, that energy put back into the system is lost and rendered useless, except when another train nearby is accelerating at the same time.

This is where batteries come in. Instead of wasting this energy, the testing system uses it to charge a large bank of batteries nearby. This energy can then be stored for a later time for when other trains are passing through and need to accelerate.

If Metro could save a significant amount of money with the battery system, then the agency could be able to pay less each year for electricity and put its money to better use elsewhere.

Metro tested how large-scale batteries might integrate with the train systems

In order to test and see if batteries would work, Metro bought and installed a Battery Power System (BPS) from Kawasaki. The system, made up of over 550 smaller battery cells, is rated at 2 megawatts, or 378 kilowatt-hours. This isn't incredibly huge—the larger Metro power substations are rated up to nearly 10 megawatts—but it's large enough to get some results, which is what the study aimed to do.

The system was activated and installed in June 2013 outside of the West Falls Church station. The left and right-most rows of equipment shown here include all the battery cells making up the overall unit and the center row has extra equipment needed (circuit breakers, cable terminations, etc.).

Battery installation at West Falls Church.

The battery system was tested in a few different scenarios

To test a few different real-world configurations, the battery system was tested with and without a 6 megawatt transformer (Scenario A), a 3 megawatt one (Scenario B), and none at all (Scenario C). Scenario C represented a case where the battery system is installed at a Tie Breaker Station, which are typically located between the numerous transformer substations along the railroad. The TBS' are used to help balance how much power is available to all the nearby tracks.

Testing in each scenario consisted of running trains near where the battery system was installed, and measuring how much power was saved by having the batteries active. The scenario's success was measured by comparing how much power was consumed with the battery system installed, versus how much was consumed without it. The three scenarios tested are shown on the left, and the amount of power used is graphed.

Battery installation effects on peak power draw.

The chart below shows a modified version of the one above; that is, the power saved by having the battery system installed is what is graphed here (to calculate this, subtract the "Peak power with BPS" bar above from the "Peak power without BPS one").

Amount of power saved at peak draw.

As shown in the two graphs, the most power was saved in Scenario C, followed by Scenario A and then Scenario B. Scenario C's results make the most sense: a TBS is located between power substations, meaning any sort of power draw at that location would be the most "expensive" since it's furthest from a power substation. Also the fact that the TBS doesn't produce its own power means that any saved by the battery system would be the first used when a train needs some, which helps increase the amount of power "saved."

When applying the power-saved numbers to moving trains, the impacts found in the study start coming into focus. The 2MW BPS could reportedly move 25 empty or 19 fully-loaded trains 2800 feet - a little over half a mile. The same BPS could move 17 empty and 13 fully-loaded trains 4000 feet (75 percent of a mile). If the BPS is fully-charged, then all of this power could be coming from what you've reused. This starts adding up for some serious savings.

If WMATA was able to save $888,289 over 10 years in the first scenario and $1,863,229 scenario C as the study claimed, what gives? Basically, if the cost of installing and maintaining the battery system is lower than what the agency could save over the life of the system, they come out ahead and are able to use that saved money in other ways for other things that need to get done.

Possible Benefits

The FTA gave WMATA a $300,000 grant to study this one method of capturing and storing the energy created by trains, and to figure out what the savings and benefits of such a system could be.

The installation of one of these Battery Power Systems can provide a number of benefits with seemingly few if any downsides. One of the big possibilities it provides is to act as an emergency power source (like a UPS) if there's a commercial power outage. If the commercial power feed goes down in the area, there wouldn't be nearly as much a worry of stranding trains, and passengers, on the tracks. The batteries could keep the trains moving until they reached a station where passengers could get off safely. Currently if power were to fail, a train in the middle of a tunnel would be out of luck and couldn't move.

With the study complete, WMATA doesn't have any public plans to expand the usage of batteries to supplement the power system. But the results of this study suggest that the battery system could save WMATA some money, which would be beneficial in today's financial environment.

Support Us