If we wanted to cover the entire deficit with a fare hike in order to maintain service levels, how much higher would fares have to be?

by BeyondDC on Dec 23, 2009 10:31 am • link • report

I heard a nickel on every fare would raise $30M? So raise everything by thirty cents to get $150M?

Somebody check this, please, It's fuzzy. The numbers seem about right.

by Michael Perkins on Dec 23, 2009 10:37 am • link • report

Maybe we could charge rush hour rates after midnight on Fridays and Saturday nights? Somehow, I doubt that would fill the gap, though.

by Cavan on Dec 23, 2009 10:38 am • link • report

Of course, you'd want to design your fare increase better than this.

by Michael Perkins on Dec 23, 2009 10:39 am • link • report

@Cavan: They already charge rush after 2am, I think.

How do you communicate this with the riders? The Metro map is burned into the minds of most residents and many visitors. How do you do this without special posters?

by Michael Perkins on Dec 23, 2009 10:42 am • link • report

One thing they should do now, and especially if they reduce the number of trains, is hold late night trains for connecting trains at transfer stations. They should do this for buses too.

by Maggie on Dec 23, 2009 10:45 am • link • report

I'm not a fan of eliminating the direct link between the Pentagon and the Rosslyn-Ballston corridor. As a resident of Pentagon City, I frequently travel that segment of the Blue line in order to get to where I'm going in North Arlington, even in the evenings or on weekends. Having to cross over into DC, switch to an Orange line train at L'Enfant Plaza, then ride all the way through downtown doesn't seem like the best idea. Not to mention all the folks who take the bus to the Pentagon and are looking to go into North Arlington, or everyone who lives south of me in Alexandria wishing to do the same.

Arlington Cemetery might be a station that isn't used much, but that section of the Blue line is, I think. Altering routes in the system is a good idea, but I think we need to look elsewhere on the map.

by Steve on Dec 23, 2009 10:46 am • link • report

@Matt, I think you meant Arlington and Alexandria instead of Arlandria. Arlandria is a specific area of Alexandria between Four Mile Run and Del Ray.

Mostly good ideas. I agree with Steve that going from 1 stop to get from Pentagon to Rosslyn to 8 stops and a transfer is rough, but if you wait 10 fewer minutes to get your train, it mitigates some of the problem. For every station west of Rosslyn, you'd require a transfer, anyway, so the people that are most affected are those going from Pentagon and further South to stops from Rosslyn to about Metro Center on the current alignment of the blue line.

Signage is a big problem here, but I believe it costs less than $1 million to replace it system-wide (I think when Wash National Airport went to Reagan it was somewhere around $500K or less). You could have an inset for late night schedules, but I don't know what you'd do about the maps of specific lines on the brown pillars around the station.

by Nick P on Dec 23, 2009 11:09 am • link • report

I like the thinking behind this map. Sending all the Va traffic from Springfield and Huntington across the L1/L2 bridges would do wonders to ease congestion in the Rosslyn tunnel, especially when the Silver Line comes on line. My only comment would be to extend the Blue segment to West Falls Church to take advantage of the switches there to turn it around.

How to deal with the Cemetery is problematic, but the configuration now contributes to a lot of congestion in downtown. It would be nice to see how this map provides service overall, as kind of a test for possible changes once the Silver is complete.

by Lou on Dec 23, 2009 11:10 am • link • report

I like this map, I think it makes a lot of sense. I also agree with charging rush-hour fares after midnight on weekends, if needed. Late night and weekend service needs to remain with reasonable headways; the economic and safety issues that would result would be disasterous. For example, people may be less likely to go to a late dinner, a show, or drinks if Metro's not running 10-15 minute headways. I think people would be willing to pay higher fares, however, if Metro could guarantee reasonable service.

by Adam L on Dec 23, 2009 11:26 am • link • report

What about having the trains from Franconia-Springfield (which on this map cross the bridge to Mt. Vernon Square) terminate at Rosslyn instead? That would solve Steve's problem, and it looks like it would make the overall route a bit shorter. If the Orange and Blue line trains are only coming through Rosslyn every 15 minutes, it seems like there would be plenty of time for a Yellow line train to pull in, turn around, and go back out.

But what worries me about this whole idea is that it might stop the death spiral in some parts of the system by accelerating it in others. If a large enough fraction of Metro users aren't feeling the pain of the reduced service, Metro might decide that 30-minute headways for the outer parts of the system are a swell idea to continue even after this budget crisis passes.

by Johanna on Dec 23, 2009 11:27 am • link • report

@Lou: Late at night when Metro is considering 30 minute headways, there are no tracks or tunnels that are limited by congestion. A tunnel can hold about 25 trains per hour each direction, and we're talking about four or six.

@Matt: What's the turnaround capability at Ballston? Obviously they have something ( a crossover before or after Ballston station ) because we operated underground-only service there during the weekend.

Once we have the Silver line, this idea is less useful, since the R-B corridor (and East Falls Church) would enjoy 15 minute headways even at the system's worst.

by Michael Perkins on Dec 23, 2009 11:28 am • link • report

This is a really important contribution. Thanks.

by Richard Layman on Dec 23, 2009 11:28 am • link • report

@Michael Perkins

My comments were less about evening operations, and more about how this map may indicate future changes in thinking during peak and normal operations, with Silver Line trains added to downtown. I think testing this configuration on off peak times would be great and might help planners think about future changes.

Also, there is a crossover in between Ballston and VaSq.

by Lou on Dec 23, 2009 11:34 am • link • report

@Nick P:
David charges by the letter, so I try and shorten words where possible. Would "Alexington" be a more acceptable or less confusing term?

@Lou:
First off, congestion is not a problem at night, as Michael P explained above. Secondly, this "solution" would not work during the rush hour periods when congestion is a problem, because the 7th Street subway doesn't have room to add the Blue Line at its current 6 minute headways. It could only handle the Blue Line at 12 minute headways (during rush hour) unless one sacrifices the Yellow or Green Lines's frequencies. I examined this issue here:
http://tracktwentynine.blogspot.com/2008/11/understanding-blue-line-reroute.html

@Michael P:
There is a crossover between Virginia Square and Ballston, so a train could turn back there. This would be similar to Yellow short-turns at Fort Totten, where there is no pocket. At 30 minute headways, there would be no problem turning on the mainline.

@Johanna:
Trains cannot turnback at Rosslyn easily because there is no crossover present. They would likely need to travel as far as Foggy Bottom before turning back. Again, in terms of budget, we have to cut something. Keeping the Blue Line traveling through Arlington Cemetery make the Blue Line's route *longer* than it is today, because it would run from Largo to Ballston *and* Franconia to Foggy Bottom. That is basically the status quo with 30 minute headways.

by Matt Johnson on Dec 23, 2009 11:43 am • link • report

@Matt

From reading your link, we agree the problem is the Rosslyn tunnel. You say that eventually the Silver Line will need its own tunnel through downtown, well, "eventually" is pretty much here and Metro is not going to dig another tunnel as far as I have heard. Something needs to be rethought, and to me seeing this map reaffirmed something I have felt for a while, that the Blue line service to Springfield may need to find another way into downtown. That is not a one-step solution obviously, it has impacts on Yellow, etc.

by Lou on Dec 23, 2009 11:56 am • link • report

Just to nit-pick the map:

It seems to me that the map shows solid lines for state lines and dotted lines for city/county lines. But the northern border between Virginia (Fairfax County) and Maryland (Montgomery County) is dotted.

It shows one red line train going from Shady Grove to Glenmont, then another from Grosvenor to Silver Spring. Don't rush hour trains usually split it up, so one goes from Shady Grove to Silver Spring and the next goes from Grosvenor to Glemont? This may have been deliberate, though, so I don't know.

by Tim on Dec 23, 2009 12:43 pm • link • report

@Tim:
The dotted line separating Fairfax County and Montgomery County is dotted because it's dotted on the Metro rail map from the website. You are correct though, it probably should be solid. Good catch.

And no, during rush hours, WMATA operates services between Glenmont and Shady Grove and also between Silver Spring and Grosvenor. Off peak, WMATA operates trains between Glenmont and Shady Grove and also between Silver Spring and Shady Grove.

by Matt Johnson on Dec 23, 2009 1:13 pm • link • report

While we're nit picking Metro's map, it always gets me that they refuse to show that part of DC lies outside of the Beltway. That would blow up too many paradigms I guess.

by Lou on Dec 23, 2009 1:21 pm • link • report

Lou, which part are you talking about? The tiny bit of water on the Potomac where the WW Bridge is? Because no part of the District is really outside of the beltway at all.

by Alex B. on Dec 23, 2009 1:46 pm • link • report

Yeah, the little bit by Jones Point. Hey, I've got a few free drinks with that little bit of trivia.

by Lou on Dec 23, 2009 1:59 pm • link • report

@Matt--

Does this save approximately as much money as Metro's proposed cut to 30 minute headways? It seems to me like your proposal involves more total service-- essentially, swaps out the current blue line for two separate services, one from Franconia to Mt. Vernon Square and one from Ballston to Largo.

by Josh B on Dec 23, 2009 2:24 pm • link • report

Josh B: This proposal includes the cut to 30 minute headways. So if Franconia-MV plus Largo-Ballston are the same total running time as Franconia-Largo, then it should be about budget neutral.

by David Alpert on Dec 23, 2009 2:26 pm • link • report

I think dropping the Cemetery stop is a really good idea. Occasionally the Trip Planner actually calculates that going from National Airport to Ballston via downtown/L'enfant *is* the fastest option (probably due to long Blue headways and transfers).

The map shows how the re-routings would work, but aren't routes that change color by the time of day kinda confusing? Would it be any less confusing to have trains change colors? E.g. a Blue line train from Largo could turn Orange at Stadium. Passengers would have five stops (and announcements) to realize the train isn't Blue anymore.

@Tim: Isn't the thin solid line just an outline for the District? All non-District boundaries are dotted.

by JM on Dec 23, 2009 2:33 pm • link • report

@David A.--

By my count, Franconia-Mt. Vernon should be a 33 minute trip and Ballston-Largo 44 minutes, for a total of 77. Franconia-Largo is only 64 minutes. Additionally, each circuit involves four turnarounds instead of just two, which presumably adds a bit of extra time.

This might be partly offset (I think?) by the fact that Matt's proposal terminates every other Red Line trip at Grosvenor, instead of running them all to Shady Grove as is currently true for late nights.

by Josh B on Dec 23, 2009 2:35 pm • link • report

@Michael -- assuming your figures are correct, don't forget that a fare increase, particularly a 30c one, is likely to reduce ridership. In other words, one would have to know the elasticity of demand to determine the overall revenue effect of a fare increase.

I suspect elasiticity is fairly low for peak travel, but much higher for off peak travel.

by ah on Dec 23, 2009 2:46 pm • link • report

@ah: Metro assumes an elasticity of about -0.3 for determining demand changes (I tried to obtain this information via PARP, but Metro claimed they did not have records for it -- I call shenanigans). I believe those figures given are net, rather than gross. Still, we're talking less than 50 cents (admittedly, that's a lot for bus, but it still keeps us below $2 which is the going rate in a lot of cities).

by Michael Perkins on Dec 23, 2009 2:51 pm • link • report

This proposal recognizes that the suburban commuter rail aspect of Metro should to be treated differently than the urban subway aspect. This is a fantastic idea - especially for non-peak service. Half hour headways are pretty good for the kind of long suburban trips served by the outer Metro IF you know when the train is going to leave. A true half hour schedule that is consistent and reliable ("on the :30s from Metro Center") combined with timed transfers (because what really sucks is waiting 2 x 29 minutes because you have to transfer) is probably better than what we have now.

by egk on Dec 23, 2009 4:37 pm • link • report

I still think an easy solution to this is a time-based fare increase for the weekends. By cutting services, WMATA assumes that all of its customers value their time over money (for a small fare increase). Let riders make that choice on their own.

If WMATA can get me to my destination within some X amount of time, then I'll pay an additional fare increase. If I choose to skip a train and take a longer time getting there, I won't get charged that additional fee. You can bet that not many people will be doing the latter. Depending on the fee, the plan could increase revenue, and will not discourage people from using Metro on the weekends.

Since this feature would only work with SmarTrips, the plan would have a fringe benefit of encouraging more people to use SmarTrips (since paper pass users would be required to pay the increased fare).

Instead of trying to find route permutations, let's keep the system simple - either you choose to wait, or you don't.

by SDJ on Dec 23, 2009 4:39 pm • link • report

Instead of running service which allows track maintenance to proceed in 29-minute sections, why not break up the system into 20 or so sections that can be selectively single-tracked with overlapping end stations, and *tightly* manage scheduling so that short connections are practical & everything stays safe?

by Squalish on Dec 23, 2009 5:21 pm • link • report

Forgot the obvious part of my statement - That would increase, not decrease, the operation labor expenses. what would something like this save on maintenance expenses?

by Squalish on Dec 23, 2009 5:24 pm • link • report

Change the fare system; they will to do eventually why not do it now go to a zone based system and not change it at all for the next 5 years.

Once the purple line is done the current system was be shot to hell; try explain and justify the fare from Silver Spring to Bethesda or New Carrolton to College Park when there is a purple line the distance will change which would require changing the fare especially if its integrated within the other lines.

The routing change of the blue and yellow lines would suck for in Arlington & Fairfax going to Alexandria I suggest if we have an increase in fares that trips originating in the western orange line and end on the Virginia yellow or blue line could receive a 1-5 cent discount.

Another problem is how will the times at which trains start using the night routing work. Lets say it starts at 9pm. Would it be that trains leaving at either end at 9 travel the route or any train that is operating after 9 go to the new route it would be confusing to riders if they get on a train going somewhere and mid route changes where it is going: thats a problem now with red lines and it pisses people off that needs an answer.

One problem I see with that is it could end up on all the lines and at that point a map would be useless if the trains are all changing routes traveled.

It might be better to just use a different color or different name for the train such as

Blue line A operates normally XX hours and after this time all Blue Line A trains in route will travel former route and all trains leaving Largo or Springfield after said time will travel this route.

Blue line ends at 9pm and after that the Navy Line and the Sky lines will operate until so and so time when the Blue line will operate.

What about using some of the old switch tracks in the system to redo some of the lines such as the track from the red line to the blue/orange line tracks near farragut west. Use the track to create silver lines to union station or any other combination.

by kk on Dec 23, 2009 6:01 pm • link • report

If Metro does implement something like this, they need to ensure that trains run on time and spaced correctly.

One thing I noticed during recent track work on the Orange and Blue lines is that Metro would platoon their trains together on the weekends in order to squeeze them through the single-tracked segment faster - that is, instead of having OR trains on 20 min headways and BL trains on 20 min headways as well with the trains evenly spaced (so that in the common track area, headways were 10 minutes), they'd bunch the trains together so you'd have a 2 minute wait between one train and then an 18 minute wait for the next. That's not softening the blow at all, and if that were to be the result of this, then it's not a real solution.

There's no reason to think that it wouldn't work, so long as there's not any track work - but still, worth considering. With the situation I described above, Metro clearly wanted to avoid having people wait in trains holding at stations to clear the single tracked area, and did so by sacrificing their average headway.

by Alex B. on Dec 23, 2009 6:23 pm • link • report

MARTA is a lousy role model, although the late night switch for the Northeast line is one of their few good ideas. Unfortunately, they have very infrequent service at night and long waits for switches to the East-West line from the North/South line; harmonized schedules are no brainer except in a place like ATL. The one thing I liked was the piped-in classical music at some stations. Supposedly a crime reduction measures (there's some data supporting this), it creastes a more pleasant ambiance at stations.

by Rich on Dec 23, 2009 9:00 pm • link • report

This shouldnt even be discussed.

The nations capital should not see 30 minute headways. It should not even be discussed. The discussion should be "we need more money" not "what can we do without money".

Does the beltway make a profit? Why not close it from 10pm-4am to save on wear and tear and maintenance costs?

by J on Dec 24, 2009 6:01 pm • link • report

@egk

That's just what I said last week. If metro is going to drop to 30 minute headways they they should be very predictable and on time 30 minutes. Trains should run on a tight schedule constantly and be predictable otherwise people will avoid them.

by James on Dec 25, 2009 12:49 am • link • report

This entire plan sucks. Thirty minute headways would make me rethink going anywhere beyond where I live at night.

That being said, if it has to be done, I don't think eliminating the Arlington Cemetery segment should be done. Others are right in pointing out that it helps get a lot of people from South Arlington/Alexandria to North Arlington, and from my few experiences riding that segment at night, it can be somewhat crowded, possibly more crowded than the orange line at the same time.

Maybe a better way at going about it would be to introduce two or four car trains late at night running every 15-20 minutes. If memory serves me, metro did this a few years ago to save money (not sure how it saved money since I suspect labor costs would rise). It was somewhat succesful because it saved some money, but WMATA stopped the program because riders were skeptical of getting on a two car train for some reason.

by Max D. on Dec 26, 2009 6:40 pm • link • report

Going to 2-4 car trains wouldn't change labor costs, but would cut down a little on maintenance costs. Just from what I've seen, Metro could easily go to 4 car trains on some legs during late evenings and even weekends. Then if there's a special event downtown (Nats/Caps/Wizards/etc), they could bump up to 6 car trains to cover that.

by Froggie on Dec 27, 2009 8:18 am • link • report

Your map makes me wonder if the Silver Line should run to Mt. Vernon Square via the Pentagon, only overlapping Orange in the Ballston corridor. The Blue Line would keep its route, but at longer headways. Granted, this would reduce service on the Franconia-Springfield branch. But I think the reduction of service at just those two stations is a worthy trade-off for eliminating the "Orange Crush" under the Potomac.

by Brian on Dec 30, 2009 1:14 pm • link • report

^Oops, I forgot about the Blue Line branch to Largo. So maybe not reduced headways then.

Still, if you could get to the Ballston corridor on a re-routed Silver via the Pentagon, maybe the Orange Line would see less of a crush on its trains.

by Brian on Dec 30, 2009 1:19 pm • link • report

Brian, that routing for the Silver line is not possible with the current track configuration. There is no switch in that direction at Rosslyn - inbound Orange/Silver line trains have no option to turn.

Likewise, there is no option to 'turn' at the Pentagon, either - southbound Blue line trains must continue south.

Turning the trains at those locations would only be possible by pulling into the station and then reversing direction - and operational considerations pretty much makes that a non-starter.

Regardless, this map is a plan for a (hopefully) short term solution to a budget problem. The Silver Line will not be operational for another 3+ years. Metro will have to address these budget issues, both near term and long term, well before that new service begins.

by Alex B. on Dec 30, 2009 1:25 pm • link • report

^Right, I assumed building the necessary turns. I still think that the Potomac needs a third crossing (and Georgetown to be served by Metro), but that's easily many years in the future. Perhaps, just building two turns for alternate Silver Line routing would be a much cheaper, shorter-term solution to ease Rosslyn congestion.

by Brian on Dec 30, 2009 3:10 pm • link • report

Just ran across this after getting linked to the February storm history post.

They've been doing 30 min headways during the storms; the worst thing about it is changing lines.

Two things
- set schedule, evenly spaced. WMATA.org might be sufficient so that one can actually plan when to show up at Metro.
- PAUSES at transfer stations. Saturday night I went from Foggy Bottom to Waterfront. It took 1:15, as it was a good half hour before the orange line train (with one not long behind) and ANOTHER half hour to wait for Green at L'Enfant. Presumably Green had come and gone a few minutes earlier - had it waited for the connecting line it would have been far less painful. Knowing this possibility has been enough to keep me from planning multiline trips during the rest of the snow.

by val on Feb 10, 2010 4:06 pm • link • report

I believe the Tube in London does the same time thing since off-peak times frequently result in longer waits on shared lines. For the majority of riders in the Metro system, this shared route idea sounds like a workable one.
Peace
D. Jones

by Destinie A. Jones on Feb 11, 2010 8:23 am • link • report

"...trains have been coming at 20 minute intervals. Reducing these headways to 30 minutes..."

"...the line would see an increase in headways from 20 minutes today to 15 minutes..."

"All other segments served by one line would drop from 20 to 30 minutes and from 10 to 15 minutes..."

"...service between L'Enfant Plaza and Pentagon would increase from 20 minutes to 15 minutes..."

/* begin rant */

I got whiplash trying to read this article from the ways quantities kept jumping around in directions opposite what the text was saying. I don't know if WMATA misuses these terms similarly and we're just copying them, but this lack of attention to the meaning of "increase" and "decrease" is the kind of thing that confuses people about numbers until they're saying manifestly impossible things like "300 percent discount" when they mean "three-quarters off".

Unless I'm totally confused, "headway" is the time between trains. Going from 20 to 30 minutes is an *increase* in headway (more room), which is a *decrease* in service. If you have to use the terms "reduce" or "decrease" when you describe larger intervals between trains (perhaps assuming that Homer Simpson would ignore the text and think "Decrease... BAD"), you should say something like "decrease from 3 to 2 trains per hour". Decrease trains... BAD for service; decrease headway... GOOD for service.

"Innumeracy" -- the difficulty or even inability of people to use numbers for ordinary, everyday problems -- happens partly because we get sloppy when we talk about these things. It's exactly the same reason why illiteracy is promoted when people don't care if their speech or writing is clear. In both cases, we gradually lose the ability to communicate effectively. No single article on transit schedule changes, or "omg! u r such a qt!" text message, will do any appreciable harm, but the accumulation of apathetic usage erodes our ability to understand each other.

/* end rant */

by jeffq on Apr 20, 2010 7:20 pm • link • report