Since there are only two tracks, this'd require near perfection in scheduling to really save much time, wouldn't it? If a train is stopped at a station for too long, the next train that would skip that station would still have to wait for that stopped train to move.

Additionally, isn't skip-stop service more appropriate for residential ends-of-lines? These Arlington stations (Ballston-Rosslyn) are morning office destinations as much as residential boarding sources . . . with lots of commuters coming from both directions.

by Joey on Mar 4, 2009 4:11 pm  

I've never liked skip-stop operations. Confusing to casual riders, and the time benefits are really quite slim. It's a useful interim solution, but the long term problems require more capacity in terms of more rail cars and eventually, a separated Blue Line.

by Alex B. on Mar 4, 2009 4:16 pm  

One thing I haven't seen metro use is converting trains to temporary "express" service. By that I mean when a train has fallen behind schedule, it tends to exacerbate, because stations get more crowded and boarding takes longer. Meanwhile the train behind catches up.

Why doesn't Metro then send the slow-moving train ahead a couple of stops? Sure, it would inconvenience a few people, but it would be announced one or two stations in advance. E.g., at Metro Center they would announce a train would stop at Farragut North, but then not stop until, say, Cleveland Park.

by ah on Mar 4, 2009 4:22 pm  

Joey ... I had the same question that you did. How is it possible to skip stops ... wouldn't you need a third track?

And AH ... your idea is great but not as need on Metrorail as it is on Metrobus. When bus bunching occurs, buses should just start skipping stops.

by Tim Shea on Mar 4, 2009 4:26 pm  

@ah

I am pretty sure they do sometimes do this when there are extreme delays, I have definitely been on red line trains that have skipped stations in the Upper NW before when the delays were particularly bad.

by Nate on Mar 4, 2009 4:33 pm  

@Tim Shea -- definitely true for buses. It's even easier to see the problem then. Just yesterday I saw two buses on the same route right behind each other. Odd thing was that the first one stopped to pick up passengers. Maybe he had to drop (although it wasn't a stop where people would likely get off in the morning), but I would have thought protocol would be to have the second one pick up.

by ah on Mar 4, 2009 4:36 pm  

If it can work there, why not here?

Lack of effort.

@ah: I second your "express" trains. Hear hear. But it's never going to happen. WMATA does not have the vision to run things intelligently.

by Jasper on Mar 4, 2009 4:37 pm  

since the problem on the orange line is the Rosslyn tunnel, how would this help? Maybe the trains would be less crowded, but I don't think the current arrangement would allow for more trains.

by charlie on Mar 4, 2009 4:41 pm  

@ah ... i have been writing a post (that i have yet to finish) on my own blog for quite some time regarding bus bunching and why the drivers can't figure it out. It was fueled by WMATA saying they are hiring three additional line supervisors for the S1/2/4 during rush hours to help with bus bunching issues.

Supervisors? What ... are you kidding? It just baffles me that the bus drivers don't take action into their own hands. I don't know if they are not allowed to, but it seems to me that if I were a bus driver and saw that each stop was packed and there were two or three buses behind me - I would start skipping stops.

WMATA is just ridic.

by Tim Shea on Mar 4, 2009 4:41 pm  

I'm not sure if it's worth it, although I'd be up for a trial. (Of course, I don't commute on the Orange Line, so I probably shouldn't have a vote).

"ah",

Metro does "express" trains. I've seen it several times and been on a few. I've only ever seen it on the northern section of the Green Line. Stops skipped usually are West Hyattsville OR College Park (but not both). One problem with this is that operators aren't always audible. If the train suddenly goes through your station without stopping it can be a bit shocking.

"Tim",

You don't necessarily need a third track. Philadelphia's Market-Frankford Subway Elevated (MFSE) only has two tracks (whereas the Broad Street Line has express and local tracks), but the MFSE is the line used for A/B Skip-Stop service. However, without the third track, I'm not sure you'd actually gain any time savings. It might alleviate some overcrowding, however.

Another way we might be able to help Arlington riders get a seat would be turnback service between Stadium-Armory and Ballston. This would require the construction of a pocket between Ballston and East Falls Church, but the I-66 median seems wide enough.

by Matt' on Mar 4, 2009 4:41 pm  

I agree something has to be done. I live in Fairfax and work near Foggy Bottom. The trains take far too long to get through the crowded, close-together stops in Arlington.

I've often wondered about a "suburbs express" option that goes from Rosslyn to East Falls Church w/o stopping. The problem is the damn two-tracks. If they could find a way to squeeze a third track through, the expresses could become a reality. Of course there's a huge need for that, which goes way beyond expresses to the overall efficiency of the system (see today's Orange Line mess).

Oh to dream...

by Andrew on Mar 4, 2009 4:47 pm  

New York City ended up discontinuing their skip-stop service, by the way. 1 and 9 trains alternated stops (with a few "all stop" stations mixed in) after 137th Street in Upper Manhattan up to 242nd in the Bronx at the end of the line. The J and Z trains also did skip-stop after Myrtle Avenue in Brooklyn out to Jamaica in Queens.

They ultimately discontinued the 1/9 skip-stop. Turning trains around at 242nd took time, and when trains bunched up, they'd have to wait to get into 242nd. Since the 1/9 ran frequent rush hour service, other trains would back up farther down. The trains were basically having to stop at stations and not open doors. The J/Z skip-stop is still running, but for budget reasons the MTA is planning to cut the Z and have the J make all stops.

Both of these lines were quite long. The J and Z each skip seven stations along their route. That lets the time savings add up. I think this works best for very long lines with a lot of stops and low ridership. I suspect it wouldn't do much for the Orange Line, but I'm not sure.

by David Alpert on Mar 4, 2009 4:51 pm  

Supervisors? What ... are you kidding? It just baffles me that the bus drivers don't take action into their own hands. I don't know if they are not allowed to, but it seems to me that if I were a bus driver and saw that each stop was packed and there were two or three buses behind me - I would start skipping stops. WMATA is just ridic.

Tim, do you know what street supervisors do (in any agency, not WMATA) - or anything about transit operations?

by A on Mar 4, 2009 5:32 pm  

A ... as much as the tone of your commentary makes me want to not respond.

While I don't consider myself an expert in transit operations or policy ... my understanding is that a field/street supervisor is essentially a field dispatcher. Observing what is occurring with the line, knowing where the buses are, and making the call regarding bus movement/changing of schedules on a real-time basis

I look forward to your explanation though...

by Tim Shea on Mar 4, 2009 5:38 pm  

@ several posts -- I guess I have the good fortune not to need to take Metro out to its extremities, so wasn't aware that they do "express" trains sometimes. I haven't personally encountered it on my occasional rides to the outer stations. Good to know someone's thinking.

by ah on Mar 4, 2009 5:48 pm  

Part of the problem here is that MARC and VRE aren't nearly good enough. Far too many people who would be served much better by decent commuter rail take Metro. Frankly, it's pretty ridiculous for anyone to use local-stop Metro to go all the way from (say) Shady Grove to downtown DC. A larger portion of that market could and should be captured by express commuter trains.

by BeyondDC on Mar 4, 2009 6:05 pm  

10 years ago I worked out at Twinbrook and in the morning some trains going that way skipped the Twinbrook stop. (redline towards gaithersburg). I can't remember if there was skip in the evening going back into DC.

by Bianchi on Mar 4, 2009 6:10 pm  

I rode Philadelphia's Market-Frankford El (blue line) for a number of years between 69th Street Terminal and Center City. Skip Stop service works!

by Alfred Carr on Mar 4, 2009 6:12 pm  

CTA in Chicago also ran "A" and "B" trains, I think until around 1990. Some of the older station signage still indicates whether a station is an "A" stop or a "B" stop.

A lot of this comes down to design philosophy. Older subway systems aren't that far removed from streetcars. Historically, they ensured safety by moving slowly. CTA and New York City operated under "stop on sight" rules, which meant that one train could conceivably come within a carlength or two of another train. (I think that this is still the case.) They can squeeze more trains into the same track space. Low-tech, but resilient.

Metro is closer to a mainline railroad by design. Its signalling system and operating rules enforce much greater space between trains, because the system moves heavier trains at higher speeds. It allows and, indeed, requires strong central control. Unfortunately, that means less flexibility in train movements.

It is possible to increase train density by rewriting rules and by altering signal systems. Rebuilt signal systems are how the British can operate high-speed trains without constructing new trackage. A more serious obstacle is that Metro has comparatively few crossovers between main tracks. A mainline commuter railroad, by comparison, would give trains the ability to pass one another at pretty much every station.

by David Ramos on Mar 4, 2009 6:27 pm  

Regarding express trains:

Maybe by using them in the downtown core would help allieviate(sp?) bottlenecks. While the Rosslyn tunnel sucks I've been on many a orange line between foggy bottom and metro center that is stuck in a bottleneck. My understanding is that this is caused by loading and offloading taking too long. If a train that was having this problem recognized it (and that it was holding up trains behind it) why couldn't it skip a station downtown.

Ex. Overcrowded train approaching foggy bottom (it was the first orange line to come in 4 minutes while 3 trains came at 1 minute intervals behind it- I feel this happens all the time). Conducter announces we will not be stopping at Farragut West- if that is your stop get off at Foggy bottom and get on the train directy behind this one.

Do this a few times downtown and people will adjust. If it actually works and allieviates some bottlenecks people might appreciate it.

by mafiosa on Mar 4, 2009 6:41 pm  

I'm all for skipping stops after a delay, but I don't think it will save much time...and lead to lots of confusion among passengers.

Part of the problem with doing this on the Orange Line is that headways are already too long (generally 6 minutes during rush hour). That means you're talking a 12-minute wait during rush hour if you're at one of the skipped stops. It's one thing if the trains come every minute or two, but not if you're paying peak fares to wait more than 10 minutes.

You don't really add any capacity, since we're already maxing that out, and probably don't end up saving much time because trains tend to be backed up already.

by Adam F on Mar 4, 2009 6:55 pm  

I took a look at track diagrams. The two-track portions of the CTA have crossovers at each station, more or less, though this is complicated by the junction trackage. Metro has fewer, maybe crossovers every two or three stations, and those stations are farther apart.

by David Ramos on Mar 4, 2009 7:07 pm  

The only thing that can help the orange line is a separate parallel silver line with a new crossing over the Potomac into Georgetown where it hooks into a rerouted blue line.

by Adam L on Mar 4, 2009 7:16 pm  

The "L" in Chicago could run locals and express trains because the distances between stations were often small, compared to what is typical on much of the Metro, so getting on the wrong train wasn't always a big deal. Moreover, only a small number of stops were skipped.

The small amount of time saved on the Orange would not create economies of time to make up for this and trains might become even more prone to bunching. The Rosslyn tunnel is a rate limiting step and this will worsen with the Silver Line. Even as a thought exercise, it's not difficult to see how unworkable this idea is.

by Rich on Mar 4, 2009 8:22 pm  

I like the idea in general, although the two track vs three track issue is a legitimate challenge. A possible alternative skip-stop would just be to alternate stations after Ballston (inbound). I think every train should stop at every underground station (Ballston onward) since those are common 'destination' stations. But having the above ground commuter oriented stations beyond Ballston served by every other train during rush hour seems reasonable.

by PJ on Mar 4, 2009 9:50 pm  

How about express buses going nonstop into DC via I-66, leaving from Vienna, WFC, and Ballston? I live near Ballston, and I would definitely take such a bus. (In fact, I used to take the 3Y [Lee Highway to Farragut] until it got too full of people and started stopping every 100 feet on Lee.) That would get a lot of people off the Orange Line, leaving those still left a lot happier. Or there could be a 3Y express that would stop just a couple of places on Lee and then cross the Key Bridge into Georgetown, where riders could catch a Circulator bus into downtown--or just hoof it.

From my experience, having your stop skipped is infuriating. Like Bianchi, I also used to take Metro to Twinbrook. A couple of times, they skipped Twinbrook--ostensibly because the train was "behind schedule." As if there's a schedule that people follow in waiting for the train. I couldn't believe it. (It was also by far the least logical stop to skip, given the many huge employers nearby, but that's beside the point.)

That said, if it were 100% bankable that skipping a couple of less-used stops (e.g., VA Square) would actually result in less crowding, I'd be open to it on a trial basis.

But first, let's try some true-blue I-66 express buses!

by Sloop John B on Mar 4, 2009 10:28 pm  

So what about the

"small minority of riders traveling to one of the skipped stations also experience increased times waiting for their train. In a few cases, they may actually have to change trains if both their boarding and deboarding stations are served by different trains"

What do we do about them should they receive a discount of some sort maybe 40 cent off the fare and paying around $0.95 instead of the 1.35 or whatever it could be since that would be a big FU to people who only need to go 1-4 stops; I know for sure if I had to go through that crap I would be driving.

Doing stuff like that just turns the system into a Marc/Vre like commuter rail and is that's the case than all stations aside from Vienna, Ballston and Rosslyn should not be there

by Kk on Mar 4, 2009 11:31 pm  

So what about the

"small minority of riders traveling to one of the skipped stations also experience increased times waiting for their train. In a few cases, they may actually have to change trains if both their boarding and deboarding stations are served by different trains"

What do we do about them should they receive a discount of some sort maybe 40 cent off the fare and paying around $0.95 instead of the 1.35 or whatever it could be since that would be a big FU to people who only need to go 1-4 stops; I know for sure if I had to go through that crap I would be driving.

Doing stuff like that just turns the system into a Marc/Vre like commuter rail and is that's the case than all stations aside from Vienna, Ballston and Rosslyn should not be there

by Kk on Mar 4, 2009 11:31 pm  

I wrote a long post on why the benefits cannot possibly outweigh the drawbacks, but let me just summarize

("It's not worth the hassle")^∞

by Squalish on Mar 5, 2009 12:11 am  

The real problem is that the core route can only handle 20 trains an hour. If the London Underground can handle 28 trains an hour through the sub-surface lines at Liverpool street, then you should be able to here. That route in London has three lines sharing, that have multiple branches and all the routes access the core route via flat junctions.

The solution of course will cost a lot of money as it requires more frequent signals to shorten headways and more trains to take advantage of it. In theory you could rune even more trains as the tube used to run 34 trains an hour on this route until the stripped out a lot of signals as cost cutting measure in the 1970's.

If you could get the core up to 28 trains an hour then the branches could have a train every 4 minutes in the peak.

by Rational Plan on Mar 5, 2009 5:34 am  

Rational Plan,

The ATC system is capable of running trains every 90 seconds, meaning 40 trains per hour, not 20. The problem is that WMATA simply doesn't have enough train cars to do that.

by Alex B. on Mar 5, 2009 8:06 am  

Alex: you sure on that? I'm pretty sure I've read WMATA literature which states the minimum headway possible is 135 seconds, which would mean 26 trains per hour.

by Froggie on Mar 5, 2009 8:20 am  

I think I "did the math" awhile back and calculated the number of man-minutes saved vs. lost using station ridership estimates as well as the amount of time saved by skipping a station vice waiting for the next train. Overall, I think the result was about an average 30-second loss in time per person on average, with the time gains by further out riders being almost balanced by the close-in riders losing time.

The other idea to consider is that the time savings accrue to people that require subsidized parking and buses (riders from Vienna and WFC) to reach the station, while the people that lose time are in walkable neighborhoods (Clarendon and Virginia Square).

How's that for an inflammatory statement?

by Michael Perkins on Mar 5, 2009 8:49 am  

I may be mistaken on the exact number, but I know the maximum number of trains per hour is more than 20.

Anyway, skip stop service doesn't solve any of those problems, anyway. It's a pain to implement and would probably make the timing issues worse as trains 'zipper' together at Rosslyn.

by Alex B. on Mar 5, 2009 9:30 am  

seems to me the crowding at the inner stations can be easily solved by having trains that *originate* at those inner stations. it would inconvenience far fewer people than "skip stopping."

by AJ on Mar 5, 2009 9:48 am  

@Alex B. According to the presentations I've seen, it's 26 per hour per direction maximum for combined lines (Blue/Orange and Green/Yellow) because you need a little play to zipper them together, and something like 30 per hour for the Red line because you don't have to deal with any merging.

@AJ: That would be nice if it were possible, but there aren't tracks to add trains except at West Falls Church. Any addition of empty trains at Court House, for example, would be stuck behind trains servicing all the stations between West Falls Church and Court House.

by Michael Perkins on Mar 5, 2009 9:58 am  

I found the document I was looking for, I was wrong about the Red line:

http://wmata.com/about_metro/board_of_directors/board_docs/021408_BlueLine.pdf#11

Document states minimum headway of 135 seconds, and maximum trains per hour of 26.

by Michael Perkins on Mar 5, 2009 10:05 am  

And, breaking news: WMATA's FAO Committee has adjourned today's scheduled meeting, where the Committee was to take up the issue of service cuts. Additionally, the Board Meeting scheduled for today where the Board would have approved those cuts for circulation and public comment, was cancelled. Those meetings will likely take place next Thursday.

by Michael Perkins on Mar 5, 2009 10:09 am  

My mistake on the headways.

Anyway, Michael's Blue Line document shows the real problem, which is 'zippering' the trains together at Rosslyn. Skip Stop service does nothing to help that, and the scheduling of skip stop trains might actually make it worse. It certainly wouldn't help much along that relatively short stretch of the Orange line.

Short term, they need to add more rail cars and run more 8 car trains. Medium term, they need to consider the Blue line split to add more Orange trains and simplify the switching movements. Long term, they need to plan for and add the New Blue line via M street or some other downtown alignment.

by Alex B. on Mar 5, 2009 10:29 am  

As for the 90 second/135 second headway discussion, the confusion comes because WMATA *can* run trains every 90 seconds. *Except* at switches. When switches have to be realigned, it takes 135 seconds between trains.

This means that by running the service pattern "OR, OR, BL, OR" is more efficient than "OR, BL, OR, BL". This is helpful on the Orange/Blue subway because the ridership is so much higher on the Orange Line. If it was more evenly split, they wouldn't be able to run two Oranges for every Blue.

See: http://tracktwentynine.blogspot.com/2008/11/understanding-blue-line-reroute.html

by Matt' on Mar 5, 2009 10:47 am  

I was about to say. I know the software and hardware is technically capable of running trains at 90 second headways, but the practical limitations prevent that.

Sand Box John, to whom I'll defer on all technical questions regarding Metro, says he's seen the system running at 90 second headways on a test once (on the Red line, I think?). It can be done, theoretically.

by Alex B. on Mar 5, 2009 11:08 am  

I lived in Santiago, Chile, in Fall 2007, and the Santiago Metro system incorporated a variation of this proposition on the L4 line.

The Santiago Metro did "skip-stop" on trains headed in the "reverse commute" direction (e.g. those headed out of town). This moved trains more quickly to the end of the line so that they could re-enter the inbound train stream during rush hour to handle the crush of commuters.

It was very helpful for people headed out of town, and added much needed capacity to inbound commuter service without requiring additional equipment. Metro might want to consider trying a pilot of this in the future.

by Carl Seip on Mar 5, 2009 1:03 pm  

Headways don't tell the whole story about rail system capacity. Time separation is useful for understanding operations planning, but for safety purposes, the real consideration is physical separation of trains. A moving train must have enough space to stop, at the speed it is going. A train moving under manual control at 5 mph requires much less space than the same train at 60 mph.

Safety is, to a large extent, a matter of procedure. Also of signal density, braking capacity, and train weight and speed. Distances that are "safe" on another system might not be "safe" on Metro.

by David Ramos on Mar 5, 2009 3:52 pm  

Steve Offutt weighing back in.

Thanks all for your ideas and comments. I'm glad this got people thinking. Here are some of my thoughts based on your comments:

- I understand the problem at Rosslyn may not make shorter headways possible, but until Metro has more cars, it may allow for more 8-car trains, allowing more capacity with the same frequency, which should not exacerbate the problem at Rosslyn; they can run on the exact schedule they do now, but the one or two trains that are freed up can distribute their cars to the existing 6-car trains.

- Since each A or B skips exactly two stations, they shouldn't bunch up. In theory the trains ought to be at least two stations apart at any time, so unless there's a disruption (a different problem altogether!), I don't anticipate a bunching problem. When I lived in Philadelphia I don't recall that ever happening.

- @ Squalish "It's not worth the hassle." I'm not sure if you are referring to hassle for riders or hassle for the system. I'm not sure exactly how many riders use the Orange Line, but let's say it's 20,000. If each saves on average 45 seconds, that's 62,000 person-hours saved per year. At $15/hr, we're talking $1 million dollars in saved time. That doesn't take into account the slight improvement in crowding, too. In fact, I suspect time savings will be higher, because of all the people at Courthouse who have to wait for 2-3 trains just to get on. A little improvement in crowding would get some of those people in sooner.

- Also, Metro is so stressed at this point, that little improvements ARE (IMHO) exactly what they need. The big improvements may be years or decades out, so making small operational changes are exactly the sorts of things they need to continuously investigate while we await (and lobby for) the big ones.

- (Along these lines, you may have seen my previous posts about the invisible Farragut tunnel. Something they ought to do now, since a real tunnel is a long way off.)

- Waiting time. Keep in mind that average waits are 1/2 the headway (sometimes you just catch it and other times you j..u..s..t miss it, so it averages out). So a 12-minute wait is the max, not the average. The average wait is 6 minutes on 12-minute headways. I like the idea, though, of offering a reduced fare. Why not?

- Also, Metro riders are enterprising. Some who live close enough to Ballston or Courthouse may choose to board there if they find they are not willing to add the extra 3 minute wait.

- One of the problems I have seen with transit systems is that they are so concerned about not upsetting a few people that they miss opportunities to make it better for the cast majority. The problem is that the many who save 1-2 minutes are not going to scream like the ones who lose 5 minutes. Hence the reluctance to express trains after breakdowns (I may post on this issue later; I have some good thoughts). Thousands of people get there somewhat faster, but if 20 have their station skipped, then they complain and Metro doesn't have the spine to stand up for the more effective solution.

- I don't think skip/stop is really the best term, because it's not every other stop, it's skip two/stop at five. That's why I like the A/B designation better.

- I used to live near VA Square, and it is primarily a boarding station during rush hour--not a destination. I rarely see anyone get off (nor get on going outbound in the PM). Similar at Clarendon, but slightly less so. (@Joey outbound trains do not have to skip in the morning, nor inbound in the PM)

- @AJ You are so right! The people who live in closer are punished every day trying to squeeze on, while those way out get a comfy seat. I think it would be great to have trains originate in closer. My wife tells me of the ghost "special" train that rolls into Ballston in the morning with no one on it. I've only seen it once, so maybe it doesn't actually exist and it was just a hallucination (I got on it, so it was pretty realistic).

Thanks again, and we'll meet again soon.

Steve

by Steve Offutt on Mar 5, 2009 6:07 pm  

If you want to buy some other things which are not listed in our website:

runescape gold

runescape accounts

runescape money

runescape power leveling

runescape gold

runescape accounts

runescape items

runescape gold

runescape accounts

runescape accounts

runescape money

runescape accounts

runescape items

wow cd key

please contact with our customer service with on-line chat or add our Email,we will stock it for you in a short time after you told us what you needed.

by runescape accounts on Mar 5, 2009 9:40 pm  

Interesting suggestion and good comments, but I agree with all those longing for real express tracks. Other than Matt's "airporter" suggestion on Track Twenty-nine, was there any serious consideration of adding express tracks on the silver line? I think Matt's got the right idea--a loop around Tysons with a stop at Whiele Avenue to pick up the Tysons crowd and then express to Dulles. With separate platforms Metro could even charge an enhanced express fare and run cars with luggage racks, maybe even sell the rights to that revenue or license the service to someone like First Group of Veolia who would handle marketing. I think many would willing to spend an extra $10 each way to save 15 minutes on that ever stressful trip to the airport and the tired and cranky ride home. It's going to an impossibly long ride from downtown (though still better than driving at rush hour).

by TK on Mar 5, 2009 10:22 pm  

All skip-stopping does is create more rolling stock - and not that much, at that. The real capacity issue is congested platforms and the 30-trains-per-hour limit at the platforms.

We don't need new tunnels or crossings - we just need more platforms at the bottleneck stations.

The reason Metro can't run more than 30 trains in the tunnel isn't that the tunnel or drivers can't handle it, it's that there's nowhere to go on the other beyond waiting for the platform to clear. Add platforms and you can double capacity without digging new river crossings.

Downtown, some stations are close enough that you'd end up with a parallel double track, but outside the core there's a big difference between 400m of extra rail and tunnel/ROW per station for a platform and switches and trying to double track the whole thing.

by Ethan on Mar 6, 2009 6:48 am