The Washington, DC region is great >> and it can be greater.


Ask GGW: Why can't Metro keep more trains on the Blue Line?

With Silver Line opening soon, Metro is reducing the number of Blue Line trains running between Pentagon and Rosslyn. Reader Daniel W. wants to know why Metro is doing this.

Photo by Adam Fagen on Flickr.
Could GGW offer some insight into why Metro is increasing headway on the Blue Line instead of clearing room in the tunnel by running fewer Orange Line trains but running them with eight cars? WMATA can service the same number of Orange Line riders with one third fewer trains by simply running trains at system capacity.

The current situation is fairly complex. Each Metro line has a capacity of 26 trains per hour (TPH). Right now, the combined Blue/Orange line between Rosslyn and Stadium/Armory is operating at that limit. When the Silver Line starts operating later this month, both the Blue and the Orange lines will see fewer trains to make way for the Silver Line. That much is inevitable.

But the Blue Line reduction is more severe for two primary reasons. First, there's simply more demand for trains coming from north Arlington, so Metro wants to give the majority of the slots to the Orange and Silver lines. The other reason is that the line coming up from Alexandria and southern Arlington splits at Pentagon. Blue and Yellow line riders have two options for getting into downtown. But for the Orange and Silver lines, the only place they can be routed is via Rosslyn.

Many Blue Line riders are understandably upset about having longer waits for a direct train to the western part of downtown. Now, it may often be faster for riders to take the "rush plus" Yellow Line from Franconia and transfer at L'Enfant. But without a separated subway for the Blue Line, there's not much WMATA can do to increase capacity.

Comparing capacity

What does the current breakdown look like?

View peak service levels: Pre-Silver Line   With Silver Line
Note: This graphic was originally created in 2013. Since then, Metro has announced there will be 5 TPH per hour on the Blue Line once Silver opens, rather than 6 as shown here.

Right now at Rosslyn during peak hours, there are 19 inbound Orange Line trains and 7 inbound Blue Line trains each hour. That means there's an inbound train about every 2 and a half minutes.

Of the Orange Line trains, about 40% are 8 cars long and 60% are 6 cars. All of the Blue Line trains are currently 6 car trains. That means that each hour at Rosslyn, there are roughly 42 inbound Blue Line cars and 130 inbound Orange Line cars. This means a total of 172 inbound cars per hour.

Once the Silver Line opens, the distribution will change. Starting on July 28, the mix at Rosslyn will be 11 Orange Line trains each hour, 10 Silver Line trains per hour, and 5 Blue Line trains per hour.

All of the Silver Line trains will be 6 cars in length. The proportion of Orange Line cars is not expected to change, so it will remain about 40% 8-car trains. The Blue Line will operate with half of its trains as 8 car sets.

That means that at Rosslyn, there will be 35 Blue Line cars inbound each hour, a reduction of 7 cars. The Orange Line will also see a reduction, with only 74 cars per hour. The Silver Line will have 60 cars inbound each hour. Now, since the Silver and Orange share in northern Arlington, riders at stations like Clarendon will see 134 inbound cars each hour, which is a very slight increase. Overall, at Rosslyn, that will mean 169 cars inbound each hour (which is a slight decrease).

So, even though the transit agency is reducing Blue Line frequency in order to fit more trains coming from northern Arlington, the change in the length balance of those trains means that essentially passenger capacity is staying the same. It's only increasing by 4 cars per hour. And of course, it's dropping by quite a bit for the stations on the Orange Line west of East Falls Church (from 130 today to 74), though that section of the line is less crowded than the parts closer to the core.

Alternate solutions

Daniel's question is specifically whether WMATA can have the same capacity by increasing the length of Orange Line trains rather than reducing frequency on the Blue Line.

If WMATA kept the same number of Blue Line trains as there are today—7 TPH each 6 cars in length—the inbound Blue would have 42 cars per hour. The number of Orange Line trains would still have to change to accommodate the Silver Line, so let's assume the remaining 19 slots are distributed 10 to the Orange Line, with all 8-car trains, and 9 to the Silver Line, with all 6-car trains.

There would then be 80 inbound Orange Line cars per hour and 64 inbound Silver Line cars per hour, for a total of 144 cars per hour coming through northern Arlington. That would mean a total of 186 cars per hour inbound at Rosslyn, which is significantly more than today.

However, this solution would actually require Metro to use more cars in daily service, and right now, there simply aren't enough. The Silver Line requires more cars to run, since Wiehle Avenue is so far out. The opening of the new line is already stretching Metro's fleet. There will be more wiggle room sometime soon, once the first 64 7000-series cars arrive, but that milestone is probably more than a year away.

The other issue is that Metro still needs upgrade power systems on all the lines to enable more 8-car trains to operate. Right now, even if Metro had enough cars, there isn't enough traction power capacity to run all 8-car trains, though the agency is slowly working to upgrade traction power substations throughout the system.

Metro does anticipate running 100% 8-car trains in the near future. But being able to do so is dependent on being able to finance more 7000 series railcars. Right now, the local jurisdictions—who pay for capital upgrades—haven't ponied up enough money for that to happen, but there's still time.

WMATA expects ridership to continue to grow. More people ride the Orange Line in northern Arlington than ride the Blue Line between Pentagon and Rosslyn, and so the transit agency is trying to send enough trains that way. Officials see a Blue Line reduction as the only way right now, and since the Yellow Line bridge offers a relief valve, Metro can still give Blue Line riders just as many inbound trains each day, but some riders who used to have a direct trip may find themselves making a transfer.

The only way to resolve this capacity crunch is, in the short term, for local governments to fund more railcars and more power stations, and eventually build another Potomac crossing. Without a new river crossing, there can't be more trains on Blue, Orange, and Silver at the same time.

Matt Johnson has lived in the Washington area since 2007. He has a Master's in Planning from the University of Maryland and a BS in Public Policy from Georgia Tech. He lives in Greenbelt. Hes a member of the American Institute of Certified Planners. He is a contract employee of the Montgomery County Department of Transportation. His views are his own and do not represent those of his employer. 


Add a comment »

Nice map -- this simplifies things quite a bit.

So, after the Silver Line opens, riders originating from the Vienna, Franconia/Springfield, Huntington, Wiehle Avenue, and New Carrollton branches will all see 10TPH into the core. Sounds fair to me.

Calling for more Orange Line trains to be converted into Blue will worsen service on both the Vienna and New Carrollton branches (impacting 8 originating stations) while only benefiting riders coming from from 2 Blue Line stations. Looking at Metro's ridership statistics (, New Carrollton actually serves *more* riders than Franconia/Springfield. Why call for a change which will hurt many more riders than it will help?

by jms on Jul 10, 2014 2:20 pm • linkreport

I understand why Virginia Blue Line customers are upset but honestly, there is little WMATA can do right now. Why is this so hard to understand? If they want more service, they should ask Richmond to pony up money for a new tunnel and all that goes with it.

by Randall M. on Jul 10, 2014 2:21 pm • linkreport

Each Metro line has a capacity of 26 trains per hour (TPH).

Where, exactly, does this number come from? I mean, what is it based on, what's the specific constraint?

by TimK65 on Jul 10, 2014 2:23 pm • linkreport

Weren't Metro's switches built for 90 second service? Isn't that 40 tph?

by Mike B on Jul 10, 2014 2:28 pm • linkreport

"26TPH/Where exactly does this number come from?" Headway The ATC System provided shall enable the METRORAIL System trains to operate at 90 second headways in either direction on either mainline track, unless otherwise specified. The ATC System shall, wherever economically practical, enable the METRORAIL System trains to operate with a sustained 10-minute headway in each direction during single track, run-around service between successive interlockings made necessary by maintenance or emergency conditions.

by BO on Jul 10, 2014 2:30 pm • linkreport

The 26tph figure is based on an engineering-led study that deemed 26tph the maximum level trains could be operated reliably without minor problems/malfunctions causing outsized effects along the rest of the line.

This is what Bowersox had to say distinctly:

There was an engineering-led study about 12 or 13 years ago that looked at this issue. The study found that the Metro system can reliably operate a maximum of 26 trains per hour (one every 135 seconds) on each track. While there are certain components of the system, such as certain types of switches, that have a slightly higher design standard, as a practical matter, any more frequent than a train every 2 min 15 seconds would be too tight to ensure reliable service.

by Rob K on Jul 10, 2014 2:31 pm • linkreport

Actually Matt, there are 5 trains per hour on both Yellow and Blue in peaks, each on the 12 min frequency. There is an 11th train in the height of the peak hour on the Orange Line.

by Lord Baltimore on Jul 10, 2014 2:33 pm • linkreport

My understanding is that 24tph is a self-imposed limit; ATO, specifically SelTrac nowadays, can handle 90sec headways without switches and 2min with switches in place. Pretty stupid if you ask me...London runs 34tph on the Victoria line just fine.

by Phil on Jul 10, 2014 2:35 pm • linkreport

People keep saying that the 7000-series cars will result in this wonderful future of 8-car trains....but, most of the purchased 7000-series cars are to retire the 300 cars from the 1000-series, thus closing a longstanding NTSB safety recommendation. Metro isn't going to put 7000-series into service (beyond a token) without removing 1000-series cars, for good, to close that recommendation.

Also, unless Metro elects not to remove the 4000-series cars, also scheduled to be replaced by the 7000-series, than the system really doesn't move up in capacity that much as the 7000-series cars arrive....ABSENT Metro being able to purchase additional 7000-series cars through as of yet unused options.

WMAA only paid for 128 7000-series cars 128 (i.e. 16 trains) to staff the Silver Line once it opened. And, since cars cannot be dedicated to one line, they won't necessarily actually service the Silver Line.

by JDC on Jul 10, 2014 2:37 pm • linkreport

Great Explanation but I still think blue riders have a legitimate issue

Trains are going to be reduced to every 12 minutes which is unacceptable in rush hour

Adding just one more blue line train per hour cuts the wait down to 10 minutes and is much more effective than adding another orange and reducing the wait time by 30 seconds or less

The silver line parallels the orange line in North Arlington for the heaviest usage so coverage shouldn't be a problem

by thanks but on Jul 10, 2014 2:39 pm • linkreport

The solution is quite simple. Don't open the Silver Line. I call for a protest. Purposefully make all Silver Line trains go "out of service" by blocking the doors.

by Centrist on Jul 10, 2014 2:39 pm • linkreport

Since the switches on Rosslyn are an issue could they schedule things to be 2 trains for Orange / Silver followed by one blue and alternate as such. That way you aren't switching for every single train. Would that help?

by GP Steve on Jul 10, 2014 3:15 pm • linkreport

Randall M gets it entirely right, the solution isn't to fight over who gets what capacity in the existing tunnel, but to build the second tunnel. The benefits for economic development are significant throughout the anticipated route from Rosslyn to Union Station.

by Will on Jul 10, 2014 3:18 pm • linkreport

Localities that build next to a Metro corridor and allow more development get more service. Look at the orange line, now look at the blue line. Orange line (Rosslyn-Ballston), I see 10-12 story buildings, dense development, few to no surface parking lots. Blue line (Van Dorn, Franconia-Springfield), I see 3-4 story garden apartments, surface parking lots, light industrial development including a waste incinerator plant. The pedestrian environment is constrained on the north by railroad tracks and on the south by a multi-lane highway.

It's no surprise to me that the Blue line is not getting its service upgraded like the Orange line and Silver line. It was built in a low-density environment and is constrained in how many people it can serve.

by Michael Perkins on Jul 10, 2014 3:29 pm • linkreport

@ Michael Perkins - but the parking garage at Franconia-Springfield is (I believe) the largest in the Metro system. So yes, it might not have all of the TOD that exists in the R-B corridor, but Franconia-Springfield serves as a huge commuter catchment and a lot of buses are also geared to service Van Dorn.

by JDC on Jul 10, 2014 3:36 pm • linkreport

@JDC, it has about 5000 parking spaces, which is about a third of the daily ridership of the Ballston metro station alone, and WMATA does not have to build/maintain/operate a parking garage at Ballston to get equivalently as many people, they arrive by foot/bike under their own power.

Agreed that underperforming stations can be helped by building parking garages and bus routes, but all that costs money and is not as efficient as just allowing local land owners to build things that will use the transit that is close by.

by Michael Perkins on Jul 10, 2014 3:46 pm • linkreport

On that point, At Springfield (and King Street and Crystal City) you also have the VRE.

More expensive but it'll get you to Union Station on one seat.

But to all my friends in Southern Fairfax, Alexandria and Arlington. Start sending an annoying amount of letters and emails to elected officials asking to get started yesterday on the blue loop.

by drumz on Jul 10, 2014 3:46 pm • linkreport

Several years ago, I wrote about changing the Metrorail cost sharing formula to take into account the differences in number of riders (and riders' fares) that stations in low vice high density areas bring into the system. When I later performed calculations using this idea, I found that Arlington and DC should essentially get Metrorail service for no additional subsidy due to their dense development, and PG and Fairfax county should be paying a lot more because of very low development.

by Michael Perkins on Jul 10, 2014 4:03 pm • linkreport

Sorry, link to the post here:

by Michael Perkins on Jul 10, 2014 4:04 pm • linkreport

Hopefully this issue will help push for building the second tunnel sooner. Having a second tunnel is a priority that needs to be addressed today not way in the future. However I'm curious as to how much VA is going to pay for this tunnel because all but one of the new stations in metros loop plan are in DC. And I don't see DC paying the majority of the 26 Billion dollar price tag that's estimated.

by Ervin on Jul 10, 2014 4:39 pm • linkreport

Hopefully this issue will help push for building the second tunnel sooner. Having a second tunnel is a priority that needs to be addressed today not way in the future. However I'm curious as to how much VA is going to pay for this tunnel because all but one of the new stations in metros loop plan are in DC. And I don't see DC paying the majority of the 26 Billion dollar price tag that's estimated.

by Ervin on Jul 10, 2014 4:39 pm • linkreport

We're not getting another tunnel for at least 20 years. It is more realistic in the mid-term to push for a new Rosslyn station that will allow some trains to link Alexandria and Arlington. That could realistically happen in the next 15 years.

by movement on Jul 10, 2014 4:49 pm • linkreport


The District's MoveDC transportation plan includes the current WMATA proposed loop option for the Blue line. While funding for could be considered a new trunk line is very expensive (DDOT estimates $7.8 billion for the DC section pg. T-52) they see it as a way to reduce automotive traffic from Virginia, increase property values and provide an alternative transportation mode for District residents and visitors.

When Virginia is serious about finding money for its side, DC will likely find money for it over the next 30 years.

by Randall M. on Jul 10, 2014 5:15 pm • linkreport


To a small degree, the arrival of the 7000 series cars will result in additional 8 car trains because they can't run in any other configuration. But in the short term, this is actually a bad thing because it reduces flexibility in when an abnormally high number of cars go out of service. Currently WMATA can break 3 8-car trains in to 4 6-car trains. While this doesn't help capacity, it does help maintain service levels.

Also, because of current yard capacity constraints, WMATA cannot simply decide to hold on to any of the 1000- or 4000-series cars once their replacements arrive. There's simply nowhere to store them.

Finally, Metro won't be running any 7000 series cars on the Silver line for a long time, if ever. They've already announced all Silver trains will be 6 cars long. And MWAA downgraded the substations along the Silver Line in a change order so they don't currently meet the power requirement to support 8 car trains. That to me is the most infuriating part of all this: this is inevitably going to need to be corrected, and the other jurisdictions will end up footing part of the bill.

by dcmike on Jul 10, 2014 6:10 pm • linkreport

@Michael Perkins

You compared vastly different portions of the orange and blue lines. Everything from East Falls Church to Rosslyn, down the National Airport (and depending on where they land location wise, Potomac Yards station) are all in Arlington.
You've chosen to choose only the ballston-rosslyn corridor for north arlington, and a station in fairfax county and one just barely in the limits of alexandria city.

Pentagon City and Crystal City are fairly comparable to ballston and rosslyn, almost respectively since one's the neighborhood with the mall and other shopping, the other is the place where more people work than live.

When you choose the last two stations on the blue line in VA you should probably compare that to two stations at the end of orange in VA. While there are large dense populations near the end line stations of both, neither currently has much other than you've described for blue line. The difference is, those south of the pentagon on blue/yellow are very likely to be busing in straight to the pentagon. Since it isn't all about getting downtown, many of those people also work or attend school on the orange line in va, rather than how this article refers to saving time by just transferring to yellow to get to western dc.

When train times to get from south of rosslyn to west of it only run every 12-16 minutes it makes linking up with those buses very difficult. When buses run anywhere between 10-40 minute intervals depending on time of day and route, just trying to get to work on time could have a 15-20 minute variation just trying to transfer, and heading home waiting 12-16 minutes between trains could mean a 60 minute longer commute by bus.

by ThisGuy on Jul 10, 2014 6:27 pm • linkreport

I've heard the explanation as to why you can't run more trains through the tunnel, but it seems like a poor reason and simple thing to fix. I have a hard time believing that spending billions on digging a new tunnel is a better solution to upgrading the switches so you can run more trains. I mean, seriously how hard or expensive can it be?

by Switch on Jul 10, 2014 7:02 pm • linkreport

I honestly think the people this impacts are making far too much noise.

The impact is 4 minutes a day each way maximum (assuming no delays). This means if you don't bother to plan when you get to the station your mean time offset is 2 minutes each way, or 4 minutes total.

My commute like many requires a connection via a bus, that runs twice an hour... this means I have to make sure I waste at least what these commuters increased headway is just to make the connection!

I can understand if you purchased property on the line to commute to Foggy Bottom to be annoyed that your services are decreased, but please... anyone can find 4 minutes of productive activity on a smart phone, writing an extra email or reading something you would have read anyway.

I can drive to work in 25 minutes, instead I take Metro and the bus for an hour every day each way, because sitting behind the wheel just pollutes and I cannot do anything useful. I don't look at my 2 hours on trains and buses as lost time just time I need to use with a book or doing something on my phone.

by Brian on Jul 10, 2014 7:58 pm • linkreport


The control system for the subway uses blocks. If a train is in a block you can't put another train there or big bad accident. Think following distance in your car. 26 trains an hour is one just over every two minutes.

Increasing the throughput on the tracks would require you to decrease the block size for the entire system otherwise the "wider pipe" in the tunnel wouldn't do any good. If you could shrink the block size from Rosslyn to Stadium then yes you could push more trains through the system. Without that you just move the bottle neck one station away.

Moreover, think about how long a train must stop in a station to board passengers. I haven't timed it for high traffic stops like Rosslyn but it could easily be more than a minute. With 12 stations in the Orange/Blue/Silver core 12/60 minutes are probably just loading and unloading. Plus by road it is 7 miles and change between Rosslyn and Stadium. The top speed of the subway is only 75. Most of the time it is less than 60 MPH. So that's already about 20 minutes. So that means no matter what you do with a single track you aren't going to do better than about 30 trains per hour. They do 26.

If you want more trains you need another set of tracks and less interlining. That's just the bottom line.

by Brian on Jul 10, 2014 8:12 pm • linkreport

Give the blue line a fixed published schedule and right of way over other lines the way Amtrak has right of way over Marc and VRE to meet its schedule. That way the median wait time will not increase for those able to time arrival to a station.

by JimT on Jul 10, 2014 8:44 pm • linkreport

Correction I made an error. The calculation would fall to zero trains per hour if the line got infinitely long. I should have divided the transit distance time by the number of stations to get the average time between stations. So it would be about 1.5 minutes in the rough estimate per train max. That is 40 TPH. Its a very optimistic back of the envelope calculation though. Since I am now curious I will time the actual ideal travel time in off peak and the unload time in peak to get the real values some day...

by Brian on Jul 10, 2014 8:53 pm • linkreport

@ThisGuy: I chose to make those comparisons because for a long time the R-B corridor was Orange only, that a big decision was made by Arlington to put it downtown in a developable corridor rather than in a highway median, and the Jeff Davis corridor stations are also served by the Yellow line.

So it's not a surprise that the stations served by two metro lines for three decades are seeing a lot of growth. I wanted to compare two areas that had been served by only one metro line for the same period of time. In that case it was the R-B corridor compared to the end(ish) of the blue line.

I agree that the comparison is not perfect, but part of the reason the comparison is so skewed is that we made different decisions with what to do with the Orange line compared to the Blue line. And those decisions are leading to the Orange line getting upgraded service while the Blue line is getting cut.

by Michael Perkins on Jul 10, 2014 9:35 pm • linkreport

Why call for a change which will hurt many more riders than it will help?
This is normally the part of this circular argument where I come in and start talking about equity - about how the system needs to work for everyone, and how the harm to the fewer riders on the Blue Line is disproportionate compared to the benefit of the many more Orange Line riders.

Normally, here's where I talk about the fact that we have a responsibility to try and treat everyone fairly, where nothing good ever comes of doing something that directly harms a group of people just because a larger group benefits somewhere else.

But since nobody cares about the fair treatment of all Metro riders anymore, I've decided to get on board with this whole attitude that only sheer number of bodies moved matters. So, yeah, let's do it! Cut more trains from the Blue Line! Hell, why do they even need five??? Those cars would surely carry more people as Orange or Silver.

It's long past time we just close the line entirely. The people who were formerly riding it can learn to love transferring, or just start driving. Who cares? They were taking seats away from Orange riders, and that's unacceptable.

Maybe once Silver really takes off, we can revisit this whole second tunnel idea - devoted to a line that actually carries enough people, of course!

I honestly think the people this impacts are making far too much noise.
The great many people involved in building urban freeways probably thought that the people whose houses they were knocking down were making too much noise, too.
Give the blue line a fixed published schedule and right of way over other lines the way Amtrak has right of way over Marc and VRE to meet its schedule. That way the median wait time will not increase for those able to time arrival to a station.
Once again, this is a meaningless feel-good gesture that accomplishes nothing. Wait time will still increase for all riders regardless of their ability to time arrival to a station because there's no opportunity for Blue trains to overtake Orange or Silver, no third track, no passing sidings, and 26 trains per hour on each of the two tracks it is using.

The best you can ask for is for Orange and Silver trains to be held arbitrarily outside of Rosslyn for Blue to clear, so that it can promptly get stuck behind Orange and Silver further in the tunnel.

It's pointless, it's a waste of everyone's time, and it's even more of a slap in the face to Blue Line riders that you're going to publish a schedule for them (and only them) because you can't be bothered to service them like a real Metro line.

But then again, we've already established that they're not a real Metro line. They don't qualify, because not enough people ride. So do whatever you want to them, or just close the line already - it doesn't matter! In fact, since more people can be moved if those are Orange trains - keeping the line open is the wrong thing to do. Right?

by Ryan on Jul 10, 2014 9:38 pm • linkreport


Youre assuming that the blue line train is not jam packed and has room for you to get on.

Getting between north arlington and alexandria and a major problem, which is why i always drive. Its much quicker to drive and its relatively easy to find parking.

by Vance on Jul 10, 2014 9:42 pm • linkreport

The other question is why the Silver Line is starting, from day one, with 10 trains an hour. That seems awfully high. Would reducing service to 8 trains an hour (7.5 minute headways) as is the case on many other trains to less dense areas help, and then gradually reducing trains while the Blue Line adds more cars and the traction power systems are upgraded?

by calwatch on Jul 11, 2014 12:13 am • linkreport

I posted the following comment (actually part of a much longer comment) on the PlanItMetro website in response to the "Silver and Blue Line Changes: Over a Decade in the Making" blog from June 30th:

"The upcoming reduction (yet again) in Blue Line trains to Rosslyn and through DC to Largo when service starts on the new Silver Line simply highlights the fact that a second Rosslyn station solely for the Blue Line is urgently needed. This station would be an interim terminal and the initial step in constructing a separated Blue Line through DC. Planning for this station should begin immediately without waiting for DC to decide the routing of the line within the city, or whether the proposed loop line is the best way to enhance Metro capacity in the area’s core (which I don’t think it is, but that’s the subject of a different comment)."

Arlington, Alexandria, Fairfax and the Commonwealth of Virginia (perhaps with the aid of a focused tax district in Rosslyn) need to start the process now so that a second Rosslyn station can open around 2025 or so, not in 2040 or later. Personally, I think this would be a better use of transit funding than the Columbia Pike streetcar (which, in my humble opinion, should be a Metro line in that heavily-traveled corridor extending to Skyline and Mark Center).

by Mike in Tysons on Jul 11, 2014 12:17 am • linkreport


Yes. I am assume there is breathing space on the train. Even on the 4th of July I had no issue getting on the first train I just had to be in the last car. But you are quite right if the cars are completely full then its a compounded issue. Yellow service is perfectly viable routing with minimal time loss as well since the loss of blue is offset by Yellow you have no worse a wait really for one or the other.

Considering yellow...

From King Street to Foggy Bottom which is the worst case would be 26 min (yellow + orange/silver/blue), vs 20 min (blue). We have to drop the arrival penalty for reduced blue here because there are many more yellow trains than there were blue anytime recently to make a fair comparison (plus the total trains on the Blue line remain the same with rush plus Yellow). Then you have to add the transfer wait. The time for that is (60/26)/2 on average. Or an average random wait of 1.2 minutes to transfer. For a net loss of time per trip of 7.2 minutes on average over best case historical times.

For all parts that share Blue/Yellow the offset will be the same as King St save for total reduction in time.

So with the transfer: 7.2 for Foggy Bottom, 3.2 for Farragut West, 1.2 for McPhearson. Everything else is better to take the yellow line.

If you feel both the Yellow and Blue lines are both too full to use then it would seem the other lines have it worse as per WMATA ridership figures.

The reduction of Blue service is the effect stated. Riders who go to Foggy Bottom should probably try catch blue line trains everyone else can opt for either a blue or yellow line train at a loss of less than the time to read a couple GGW blog posts per trip.

Let's presume that we have a hypothetical rider at King St who wants to go to Foggy Bottom. They just arrive randomly at the station. Before the change in service they would wait an average of 4.3 minutes for a blue line train to arrive. It would take them 20 minutes to reach their end station. Total time 24.3 minutes. If they opted to take a Yellow train they would wait on the platform for 2.3 minutes, and have a transfer of 1.2 minutes to switch to the other line. Then they would have a ride of 26 minutes. Total time 29.5 minutes.

Now they do the same trips take 26 minutes or 29.3 minutes (or 29 minutes if there are 15 yellows not 14 with silver service).

My calculation before was based on the idea that headway would increase by 4 minutes, which was the worst case. Here I am taking the true average which is what you would get with random arrivals for a long period and assuming 5 blue trains. I note the map above in incorrect with 6 blue marked whereas all the press says 12 minute headway and the silver line website says 5 trains per hour.

In any case unless you get off at Foggy Bottom you are certainly under 4 minutes per trip. Listen to one song, read one blog post or 1000 other things...

If you can get the first blue train the loss is 1.7 minutes average per trip in the worst case station routing.

by Brian on Jul 11, 2014 12:35 am • linkreport

Please note, the graphic was created in January of 2013 (over a year ago) for this post:

At the time WMATA had not said definitively that there were going to be 5 BL TPH. That's why the graphic does not match the text. David (the editor) added the graphic yesterday after I had written the post.

by Matt' Johnson on Jul 11, 2014 7:19 am • linkreport

Citing a lack of cars seems like a false argument. WMATA could give Blue six 6-car trains (36 TPH) by taking one train from Silver and making 3 SL trains 8-car.

Capacity per line matches their current plan, the number of 8-car trains per hour on the OR/SL corridor is less than now so no power issue, and train requirements are about equivalent to the current plan.

But by giving blue one more tunnel slot and slightly balancing headways across the lines, it can save riders tens of thousands of hours per year:

by Save The Blue Line on Jul 11, 2014 8:19 am • linkreport

@ Save the Blue - if you read above, dcmike has explained that 8-car trains CANNOT run on the Silver Line. When Metro designed the construction of the line, they did not include enough power to move an 8-car train. So the Silver Line will only be 6-car trains, until the power stations are upgraded.

by JDC on Jul 11, 2014 8:32 am • linkreport


"Give the blue line a fixed published schedule and right of way over other lines the way Amtrak has right of way over Marc and VRE to meet its schedule. That way the median wait time will not increase for those able to time arrival to a station."

Once again, this is a meaningless feel-good gesture that accomplishes nothing. Wait time will still increase for all riders regardless of their ability to time arrival to a station because there's no opportunity for Blue trains to overtake Orange or Silver...The best you can ask for is for Orange and Silver trains to be held arbitrarily outside of Rosslyn for Blue to clear, so that it can promptly get stuck behind Orange and Silver further in the tunnel.

Actually, you could ask for the Orange and Silver trains to be held so that the Blue trains make it to Foggy Bottom on time. After that, no need to be giving it better service. The idea is to make up for the greater headway by making schedule more predictable over the stretch with the 12 minute headways.

Similarly, in the other direction there is a fair amount of flexibility and predictability allowing dispatchers to call something blue or silver and if necessary change names well before Metro Center. Agree this is not perfect, neither is Antrak. The point is merely that it is possible for dispatchers to create a bias in how they organize train traffic and that can make up for alot for people who need to take Blue Line.

by JimT on Jul 11, 2014 9:06 am • linkreport

Why do people assume that everyone taking the the blue line is only going from Van Dorn or Springfield into western downtown? Lots of people use it even through the shared yellow stations to get into North Arlington.

As more of those people start to drive, I'm sure the VA response will be to widen roads rather than pony up money for more buses or metro improvements.

by Sarah on Jul 11, 2014 9:19 am • linkreport

Actually, WMATA and ART bus are adding new bus service that will connect Crystal City to Rosslyn/Courthouse. Plus the new bus lanes in Alex./Arl. should help.

Art 43

And the 10R and 10S from WMATA (not new exactly but still).

by drumz on Jul 11, 2014 9:24 am • linkreport

@ Brian

Why is King Street to Foggy Bottom the worst case? Plenty of people go west on the orange line after going north on the blue line (e.g., anyone working in the R-B corridor). I travel frequently for work, so I make the trip between National Airport and East Falls Church twice weekly. Yellow line routing is not a suitable substitute for the blue line on that trip. With the service cuts, I’m far more likely to drive / take a cab instead, especially if it means that the handful of blue line trains remaining are completely packed (hopefully less of an issue with 8 car trains if that happens).

Similarly, plenty of people that live in the R-B corridor or further west take the metro to work at places along the blue line like Crystal City or Pentagon City. The metro isn’t just a commuter rail system for getting people into DC. Local Virginia-only trips involving some time on both the blue and orange lines with a transfer at Rosslyn are very common, and the yellow line is not a good replacement for those trips.

I totally understand why Metro made the change, but blue line riders have a right to complain about the service reduction here, especially when coupled with increased fares. I definitely think it’s time for people in Virginia and DC to start seriously pushing for a separated blue line running though the core north of the current alignment. It’ll take billions of dollars and decades of time to happen, but the sooner the process can start, the sooner it can become a reality. Since extensive tunneling would be involved, this would be a much bigger project than the Silver Line. You would keep Rosslyn as a single transfer station for Orange / Blue / Silver, but then run a tunnel / bridge into Georgetown. Continue underground through the city to connect to Green / Yellow somewhere around U Street, connect to Red at Union Station, and eventually reconnect to the current alignment somewhere around Eastern Market. Additional stations can be added throughout the line in the spots that make the most sense. Since Largo will now be served by the Silver Line, maybe you could get some funding from Maryland too by splitting the Blue / Silver in PG County. Virginia would be more likely to contribute funding if they also got an extension of the blue line past Franconia / Springfield.

by Jason on Jul 11, 2014 10:02 am • linkreport

@calwatch: "The other question is why the Silver Line is starting, from day one, with 10 trains an hour. That seems awfully high. Would reducing service to 8 trains an hour (7.5 minute headways) as is the case on many other trains to less dense areas help, and then gradually reducing trains while the Blue Line adds more cars and the traction power systems are upgraded?"

It's not about the silver line service levels per se, it's the incredibly busy and dense ballston-rosslyn corridor. There are also expectations that a good bit of the bus load at Vienna will switch to the silver line. If the projections there don't pan out then it would be appropriate to look at the allocations.

@Mike B: "Weren't Metro's switches built for 90 second service? Isn't that 40 tph?"

@Phil: "My understanding is that 24tph is a self-imposed limit; ATO, specifically SelTrac nowadays, can handle 90sec headways without switches and 2min with switches in place. Pretty stupid if you ask me...London runs 34tph on the Victoria line just fine."

@Switch: "I've heard the explanation as to why you can't run more trains through the tunnel, but it seems like a poor reason and simple thing to fix. I have a hard time believing that spending billions on digging a new tunnel is a better solution to upgrading the switches so you can run more trains. I mean, seriously how hard or expensive can it be?"

No train system in the world schedules their service to max out the peak capacity of their signalling system. If you do that, and anything at all delays a train (e.g., someone gets stuck in the doors) you cannot possibly recover and the delays rebound across the system. So, yes, metro has a 90s theoretical peak headway, giving 40tph, but they intentionally schedule enough slack to try to keep 24tph moving giving normal expected problems (a train can just go faster to make up time if it gets delayed at a station). That's extremely typical for the industry. So how does the Victoria line run 34tph? They upgraded their signalling system to one with a 60s theoretical peak headway (60tph). On a percentage basis, they're actually using a lower fraction of their theoretical capacity than metro is--and for the same reason, to keep routine problems from making the system collapse. Also, you can't upgrade the signalling in just the tunnel, you need it on the entire track (at least the parts with shared lines--the single line endpoints will have lower peak requirements for obvious reasons). Transport London spent $1bn pounds on the Victoria line upgrades (about $1.7bn), so it's not all that much cheaper than a new tunnel (worth doing, but not a quick or easy fix). Metro has been doing signalling upgrades ever since the red line accident, but I think (not certain) that's just to get everything back to the original spec after too much deferred maintenance, not to upgrade the capacity.

by Mike on Jul 11, 2014 10:29 am • linkreport

@Mike That's all well and good, but it's not a question of completely maxing out Metro's capacity. Even just boosting the tunnel to 30 tph would allow for 6 minute headways on all three lines, and that still leaves a great deal of buffer room for recovery time.

And there's no reason we can't do both signalling upgrades and separate the Blue Line - we shouldn't just be adding infrastructure, we should be making the most of what we already have.

by Low Headways on Jul 11, 2014 10:54 am • linkreport

Also, I think the speculation that the silver line is unable to run 8 car trains is incorrect. Yes, the power was downgraded, but that was tied to a change that required 135s minimum headways (26tph) vs 90s (40tph) for both the traction power and the signalling. This actually makes total sense: using metro's existing formula for capacity utilization (about 60%) results in an a practical limit of 15tph on the silver line. But remember: the silver line needs to share trackage with the Orange line after EFC and won't ever see much more than about half the shared track's capacity (12tph). This is exactly the kind of planning we want to see--instead of overbuilding the silver line through blind adherence to the existing design standards they looked at the system and saved a couple million bucks where it made sense. They also left in a requirement that traction power upgrades be possible without taking the line out of service if they decide that they do need more power in the future (e.g., if they get another tunnel and split the orange & silver line, or if they shut down the orange line west of EFC and give all the capacity to the silver line). That possibility is remote enough that simply leaving the option open seems reasonable.

FWIW, there have been 8 car test trains seen on the silver line.

by Mike on Jul 11, 2014 10:54 am • linkreport

@Low Headways: I'm somewhat more inclined to trust the extensive modeling and pervasive industry practices than relying on internet assertions that "it must be easy enough to just run a couple more trains". Metro is having enough trouble maintaining its schedule now, I have no trouble at all believing that taking more slack out of the system will have negative consequences.

by Mike on Jul 11, 2014 10:56 am • linkreport


Train to bus transfer is not as easy as you make it sound.

Lets say someone is coming from Clarendon and headed to Pentagon to catch a bus:

They leave work and get to Clarendon, they've got a max of 5 min wait for a train no matter what time they get there within rush hour. That will get them 2 stops over to Rosslyn in 8-10 minutes. From here, lets assume they've got somewhere between 6-12 minute wait for the Blue line train. Lets also assume they've built in enough travel time to accomodate for waiting for both Orange and Blue lines to be 5 minutes early for their bus. (this all asssumes their job is flexible enough to leave when they wish).

Now, what often will happen is that since those blue line trains south from rosslyn are often so packed, the person may not be able to board the train. They are now going to have to wait between 18-24 minutes total at rosslyn rather than 6-12 minutes. They then get to the pentagon (or wherever they should catch a bus) 6-12 minutes later than planned. They've now missed their bus, which may not come again for another 30-60 minutes depending on route.

Now this is two train lines and a bus, but it's pretty similar for anyone commuting from Rosslyn south via metro and bus. What happens is that instead of (8 min ride +5 min wait) from Clarendon -> Rosslyn and (12 min ride +6 waiting) from Rosslyn -> Pentagon and Pentagon -> home via bus (20 min ride + 5 waiting), which should come out to ~55min

Because of overcrowded trains that only run every 12 min, they now have something more like this

Clarendon -> Rosslyn (8 min +5 wait), Rosslyn -> Pentagon (12 min + 18 wait), Pentagon -> Home via bus (20min ride + 30-60min wait) = 90-120min

Missing one train when they only come every 12 minutes can have a VERY big impact.

by ThisGuy on Jul 11, 2014 11:03 am • linkreport

@Low Headways: also, again, it's not just a question of upgrading something in the tunnel, you need to upgrade the signalling for the entire track segment from Rosslyn to Stadium-Armory. In real terms, you probably need to upgrade everything but the red line because mixing different signalling technologies on shared lines seems like a recipe for disaster. As far as there being no reason not to do both: sure there's a reason--the money doesn't exist to do either. If I had to fight for one or the other in a 15 or 20 year plan, I'd push harder for the tunnel because it results in significantly more real potential capacity and because there probably not much chance that any of the business guys want to push for another round of signalling upgrades before the current round is depreciated.

by Mike on Jul 11, 2014 11:05 am • linkreport


It's too bad there isn't passenger movement data out there. I doubt there are very many people taking a bus to Pentagon to then take a Blue to Rosslyn and Orange to Clarendon. It's much more likely that they take a local bus to somewhere along Wilson Blvd.

by MLD on Jul 11, 2014 11:12 am • linkreport

MLD it was purely for example, but I can ensure that there are lots that go from Pentagon to Rosslyn. Pentagon, Rosslyn, and Pentagon City are the 3 busiest stations in VA for both morning and evening commute. Pentagon is a major bus depot not only for metro bus but local buses as well. You can take clarendon out of my equation entirely and you still have the same situation. You miss one train in the afternoon due to crowding and you miss a bus transfer you're screwed.

WMATA has the data for entry and exit locations as well as busses people have used (so long as it's all smartrip).

by ThisGuy on Jul 11, 2014 11:24 am • linkreport

IIUC they are adding buses from Pentagon to Rosslyn. Also its not a terribly long bike ride I think, for those who might consider that.

by AWalkerInTheCity on Jul 11, 2014 11:31 am • linkreport

Mike in Tysons has it right.

Virginia should take it upon themselves to accelerate the schedule of a blue line terminal station connected to the existing Rosslyn station. No more problems with tunnel capacity for Virginia-only riders -- simply ask for more Blue line trains that stop at Rosslyn. Seems only fair that Virginia pays to get this more quickly, too, since the main benefit are to those traveling between different points in Arlington, Alexandria, and Fairfax.

If DC takes forever to get the separated Blue line / planned loop going, they may even be able to experiment with a Silver/Blue hybrid line in the interim.

by Vinnie on Jul 11, 2014 1:42 pm • linkreport

We can double capacity with more new rail cars, and no new tracks, basically ending discussion.

We have 5 to 11 trains per hour, yet 26 trains per hour tunnel capacity, (data from comments above,). More new rail cars is the short and medium term solution, for longer trains and more frequent trains. Subway disruptions would become even more serious, so back up express bus routes duplicating Metro rail lines with stops only at Metro Stations would be the highest priority back up service.

Long term is another story. Building parallel tunnels and tracks along the currently shared portions of tracks, for separating service color into their own track lines, would be a smart move for several reasons, such as emergency back up capacity in the most congested corridors, especially if one parallel line is disrupted. This means Yellow and Green would become adjacent tunnels, but completely different tracks and platforms, and likewise Orange line and Blue line would be separated. This would make for much greater n-stage capacity through downtown DC, and greater Metro rail reliability if any individual service color line was disrupted for any reason.

Digging parallel tunnels would have to be part of a new tunnel master plan, including logical plans for eventually building express Metro lines, copying Paris RER and NYC Subway style express trains and their separate express service tunnels/tracks, to bring people from edge cities into downtown with greatest speed, or counter commute to new edge cities. These express trains would likely go from edge cities to downtown DC, and stop at all metro line intersection stations (big dot on map stations), and suburban commuter rail and suburban Amtrak transfer stations, while skipping many stops on the way to downtown.

This Express lines would be lines such as
Red Line Express, from Rockville through downtown DC to Silver Spring,
Blue-Yellow-Green Express Fraconia-Springfield to Alexandria through downtown DC to College Park to Greenbelt, and
Silver-Orange Express, from Tysons Corner through downtown DC to New Carrolton.

Each Express line would stop at most downtown stations, and all downtown stations would have to built for emergency and future flexibility sake, but opt to regularly skip during typical daily commutes McPherson Square, Judiciary Sq, Smithsonian. The suburbs likewise would have many stations skipped, one or two at a time, between shared express and local stations. Awkwardly like Paris RER express trains, the express lines would likely have to charge a higher fare box rate, by having different turnstiles to charge the different fare rates by local or express, to pay for the expensive new Metro system express service, and reflect the greater sustained value of express trains to people living or going further away.

by npendleton on Jul 14, 2014 1:57 pm • linkreport

There are two very significant non-Metro rail solutions which might close the Rosslyn to Crystal City Metro Blue line service gap, and two Metro rail solutions. Each is expensive.

1) using VRE/Amtrak/AirportExpress Rail
2) a new Streetcar line,
3) Blue line from Francoia-Springfield to Rosslyn optionally terminating in an expanded Rosslyn station each train during rush hour without blocking Silver-Orange platforms, nor blocking traditional Blue line through service to DC, requiring quickest walking distance transfers between new Blue line platform and Silver-Orange platforms,
4) a captive short line running from Rosslyn using concept (3) option terminus, but also new tracks and platforms at Pentagon direct to new platforms at Crystal City Metro and VRE/Amtrak/AirportExpress station, skipping conflicts with Yellow line, and skipping Pentagon City to maximize each lines capacity.

by npendleton on Jul 14, 2014 2:37 pm • linkreport

@npendleton - I've never laughed so hard in my life until I read your manifesto! Hilarious! The region already cannot find enough money to build a single additional tunnel under the Potomac or buy more than a handful of new 7000s... but you call for the double tracking of the entire 135-ish-mile system, with absolutely no concern about how to pay for it all.

Where did you buy whatever you took before writing all of that, so I can get some too?

by Dave on Jul 15, 2014 11:59 am • linkreport

[This comment has been deleted for violating the comment policy.]

by npendleton on Jul 15, 2014 4:48 pm • linkreport

@Dave, one other thing, separating each service (line color) into separate tunnels and tracks only needs about 20 miles of tunnels, to go from Rosslyn to the east bank of the Anacostia River, and L'Enfant Plaza to Ft. Totten. 20 miles is not your 135 miles. Building long term these 20 miles of parallel Metro tunnels would double the capacity of the network downtown and drastically expand service on each suburban Metro rail spoke of the downtown, especially the under served lines East on the Blue line to Largo, and South East on the Green line to Branch Ave.

by npendleton on Jul 15, 2014 4:58 pm • linkreport

Express tracks mirroring the current Metro tracks would be a waste of those funds. Better to use that money to build new lines to cover more of the DC metro area. Or put that money into MARC/VRE to make those services more RER-like.

The early NYC Subway lines have express tracks for several reasons - the stations are closer together (skipping some brings higher benefit), top speeds of trains are lower, acceleration of trains is lower, etc. None of these apply to Metro. The time savings from an express track on Metro would be minimal.

Your other ideas are the long-term plan - upgrades necessary for full 8-car operation (power and rolling stock) and more separation of lines so more places can have full service. Imagine where 106 miles of additional Metro could go instead of going to duplicate the existing system!

by MLD on Jul 15, 2014 5:03 pm • linkreport

@MLD, long term my Express Metro trains concept has its place, especially with separate tunnels and tracks, because it backs up the existing Metro rail corridors, with walkable density around each station, most of which corridors are a very long way from VRE and MARC rail corridors, or far from any other old heavy rail surface corridor that might be converted into a surface commuter rail type service.

As for what I might do with an entirely new Metro line, see my proposal map for my "Skyline" proposal. The map looks strange for several reason, 1) it includes my guesses about other useful possible new Metro lines might go, and it assume the city only has budget for one single new Metro line, and further it assumes that building up downtown reliability of Metro service is the highest priority. I connect as many legacy walkable communities and existing apartment towers in DC that I could find, and provide single seat commuting service to as many key downtown rail hubs as possible, such as the new streetcar stop north east of Union Station. I also specify where I believe DC should zone and encourage building the tallest commercial and residential skyscrapers and towers, to create the unifying downtown that best leverages the wider regions rail systems.

by npendleton on Jul 15, 2014 5:17 pm • linkreport

[This comment has been deleted for violating the comment policy.]

by npendleton on Jul 15, 2014 5:31 pm • linkreport

@Dave, [Deleted for violating the comment policy.] "including the key fact or actually having read what I said, thereby completely discounting any careful thing I just wrote, such as more rail cars and emergency back up express bus service only stopping at subway station, as I said above ends the discussion for the short and medium term, because these can be done in affordable and reasonable steps."

"I made very clear and logical long term master planning solutions suggestions for when there is no more existing Metro rail track capacity. My solution suggests which right-of-ways to protect to permit a future time when we must expand the number of tunnels and where to dig them. So what is your brilliant long term solution? You don't have one." (three sentences deleted.) "So be it. But someone must think ahead, and take the effortless risk to express a good idea."

"Any other suggestion you might have in fact costs the region more forever to do basic commuting, especially if the regions population doubles like the nation in the next 55 to 80 years. Do you think we should use more automobiles per person per commute? More automobiles per mile of road lanes per hour? More fuel per person per commute? Any way you cut it, your predictably myopic financing complaint is hubris, because I just stated the lowest total cost of commuting in terms of both time and energy."

by npendleton on Jul 15, 2014 5:53 pm • linkreport

@npendleton - Everything you write is just total pie in the sky, which is why I find it so funny. Sorry if you think your plans are fiscally sound and/or would receive full funding in the next 25 years, which is about as far out as any Federal or state agency budgets out for.

You're correct that building Metro capacity results in "the lowest total cost of commuting" for the region as a whole... if those costs were borne completely by a single entity called The Region that had full control over all the region's GDP. Since those costs are in fact borne by individuals all making separate spending plans with their tiny slivers of the region's GDP... it is not going to be clear to everyone that they should simply suck it up and spend all the region's transportation dollars on a single goal of expanding Metro capacity (despite that being the "perfect" solution that helps the most number of people at the lowest possible cost).

If that's "myopic" thinking, it's also fully cognizant of typical human motivation and behavior when it comes to appropriating what humans see as "their" income streams.

by Dave on Jul 21, 2014 2:52 pm • linkreport

[This comment has been deleted for violating the comment policy.]

by npendleton on Jul 21, 2014 4:27 pm • linkreport

Add a Comment

Name: (will be displayed on the comments page)

Email: (must be your real address, but will be kept private)

URL: (optional, will be displayed)

You can use some HTML, like <blockquote>quoting another comment</blockquote>, <i>italics</i>, and <a href="http://url_here">hyperlinks</a>. More here.

Your comment:

By submitting a comment, you agree to abide by our comment policy.
Notify me of followup comments via email. (You can also subscribe without commenting.)
Save my name and email address on this computer so I don't have to enter it next time, and so I don't have to answer the anti-spam map challenge question in the future.


Support Us