I found it odd that the article about ridership didn't mention that bus fares are a lot cheaper than metrorail fares. Not to mention that they're flat (no peak fares! No peak of the peak!), and you can even get a 7-day unlimited pass for the price of one round-trip per weekday.

by Gray on Mar 13, 2012 8:29 am • link • report

Yep, but we have endless debates on this blog that transit pricing doesn't matter. Just make it more expensive.

More true: another article on WMATA that didn't mention the pension system as driving the expense.

by charlie on Mar 13, 2012 8:32 am • link • report

Meanwhile, according to a very recent Adam Tuss tweet, the head of Metro's Union is saying that "you can't keep funding your shortfall on the backs of riders". I find that mildly ironic.

by Froggie on Mar 13, 2012 9:11 am • link • report

The study on reducing bus bunching may be significant. It shows, at least for the routes studied, that you can get less bunching and maintain shorter headways over the whole route if you add more control points where you can add dynamic pauses.
I'm thinking Purple Line here. MTA has an operational analysis that they say shows they cannot single-track even for a few hundered feet in the Bethesda Tunnel, because they need to have long "dwell time" built into the schedule at the Bethesda Station to regulate their train schedules. But I think they are assuming only two "control points", at the end of the line at Bethesda and at New Carrollton.
The study on bus bunching suggests the Purple Line might achieve better regulation if they added more control points (say at major stations like College Park, Langley Park, Silver Spring) and used dynamic pauses. If that will work, then the "dwill time" needed at Bethesda could be greatly reduced, single-track could be used in the tunnel without compromising the 6 minute headway goal, and the Capital Crescent Trail could stay in the tunnel alongside the single-track section.
Far fetched? Maybe. But worth a look when there so much at stake for the CCT in Bethesda.

by Wayne Phyillaier on Mar 13, 2012 9:12 am • link • report

How is the bus bunching study different than Circualtor?

by charlie on Mar 13, 2012 9:16 am • link • report

A measure for metrorail that would provide a clearer picture of what's going on is ridership per train rather than total ridership. Ridership per train is surely on the rise. It's just that the number of trains has fallen significantly with all the track work. For example, this past weekend, headways on the orange line were 24 minutes, half of normal. If you cut the number of trains in half and ridership only decreases marginally, then the demand for transit is much stronger than what's shown by a stat showing total ridership is flat.

On the flip side, I'm sure WMATA will claim a huge success when total ridership spikes upward with the opening of the Silver Line.

by Falls Church on Mar 13, 2012 9:27 am • link • report

"Adam's Morgan"?

by ah on Mar 13, 2012 9:31 am • link • report

Re: "Adam's Morgan's"

"Adams Morgan's"? I would've gone with "The Adams Morgan root canal that is the streetscape improvement project..." just to avoid having to try to make Adams Morgan possessive.

by dc denizen on Mar 13, 2012 9:46 am • link • report

If you cut the number of trains in half and ridership only decreases marginally

Well, I can tell you that weekend ridership of the Red Line certainly hasn't fallen "marginally." Walking's often faster than taking the train when there's work being done on the Red Line.

by andrew on Mar 13, 2012 10:40 am • link • report

@charlie

"Yep, but we have endless debates on this blog that transit pricing doesn't matter. Just make it more expensive."

I'm not sure what you're talking about, but pricing definitely does matter. Are you saying that someone has argued here that transit demand is perfectly inelastic?

by Gray on Mar 13, 2012 10:44 am • link • report

Also, I wrote this on the WashCycle blog, but I have no idea where that idea for the 11th St Bridge came from.

BOTH spans are on-schedule to be demolished by Skanska-Facchina this summer.

The old 11th St Bridges were built with a design that is now known to be inherently unsafe, and the bridge deck of the existing outbound freeway bridge is considerably higher off of the water than the new Local bridge will be.

This concept would require the existing deck to be removed, and a new deck be installed closer to the water (presumably by shortening the piers), and also somewhat wider so that it could can match up with the new local bridge. You'd also have to do something about the abutments at either end, which are both a considerable height off of the ground.

And, all this for what? Connecting the Navy Yard and Washington Gas hazmat site with Poplar Point and Anacostia Park? That's the whole purpose of the new Local bridge! Until Washington Gas cleans up their site or the Navy decide to pack up and call it quits, development around the northern end of the bridge is going to be very minimal.

I can buy the argument that we might as well keep the bridge piers around; they might be useful someday. However, for now, there are many other pie-in-the-sky projects around the river that would cost much less, and have far more merit (ie. a bike/ped connection across the river along the CSX tracks).

The time to propose and plan this project passed about 5 years ago. There was a lengthy public comment period surrounding the reconstruction of the 11th St bridges, and I don't believe that this idea was ever pitched.

by andrew on Mar 13, 2012 10:53 am • link • report

@FallsChurch; that seems an incredibly weak metric. I agree with you the trains are more full, but you have to look at total revenue. Perhaps you could look a rush hour, or off-peak during the week. But trying to say, well, we took half our system offline for two days a week last year, revenue didn't grow, and yes, we need to raise fare?

Same goes for bus; I wonder how much bus revenue is up vs. ridership.

by charlie on Mar 13, 2012 10:55 am • link • report

Transit systems are not for-profit businesses, and I think that the "we need to raise fares" discussion is a different topic.

Framing discussions in terms of revenue or profit is not a great exercise in the first place.

In fact, it's difficult to have any sort of discussion about off-peak Metrorail ridership right now. The service patterns are constantly in flux, and should eventually return to normal.

A better way to frame things would be: "We were forced to decrease the level of service, and haven't seen a drop in ridership yet."

Still, this doesn't bode well for Metro. Remember that the system lost almost an entire week of ridership in 2010 due to the snow. If we got similar numbers in 2011 (which had no weather events that closed the system), ridership effectively declined.

(Oh, and what's with that bus bunching study? Have they tried their algorithm on a route with more than 3 buses? Sure, they improved headways, but the route that they tested it on already had fairly good headways to start... This sounds kind of similar to WMATA's new strategy for the 90/92/93 line, which, while still far from perfect, seems to have been fairly successful so far.)

by andrew on Mar 13, 2012 11:11 am • link • report

@ andrew; looking at the fare structure is more than fair (get it, hah) when WMATA is pushing for an increase. Although, I'd like to see a disclaimer in any article (WMATA costs too much because of pensions. They underfunded it for years. However, the current system is broken and needs to be changed)

by charlie on Mar 13, 2012 11:16 am • link • report

A thought for future streetscape projects. Can the district enact a sales/restaurant tax holiday for the areas impacted by the work? Instead of tossing money at the businesses or offering them a tax break. This would attract more shoppers/diners and be a win win for businesses and residents. Also the 11th street bridge idea looks awesome.

by Johnny on Mar 13, 2012 11:23 am • link • report

This sounds kind of similar to WMATA's new strategy for the 90/92/93 line, which, while still far from perfect, seems to have been fairly successful so far

By what metric has it been successful? I find along U street that the 90/92/96 are constantly bunched. Service has been reduced past 14th street from the 92 buses turning around, and there's no point in walking to 14th to catch a 92 since the 90 and 92 seem to leave in packs for some reason.

Just take a look at the live map on Nextbus any day and you'll see buses right behind one another with big gaps. I thought the headway idea was a good one but WMATA seems unable to keep the buses on headway given that they have multiple overlapping services with multiple start points.

by MLD on Mar 13, 2012 11:29 am • link • report

@Wayne Phyillaier: You really think that the Purple Line should be single tracked, limiting its throughput for decades to come and at best requiring pausing at multiple stations (i.e. lengthening trips), just to maintain the CCT in the tunnel?

by Gray on Mar 13, 2012 1:44 pm • link • report

@Gray:

I agree that Purple Line operational capabilities should not be allowed to be seriously restricted.

But I am proposing that this alternate method be given a serious evaluation to see IF it can maintain the headways and throughput the Purple Line needs, even with one very short single-track section. Why should we be afraid to look at least look at this to see if it will work?

As for the "for decades to come" I would point out that a major reason for dropping an alternate proposal to redevelop the Air Rights Building, so a wider station could be built at the same time the building is rebuilt, is that the time schedule for building the Purple Line by 2020 does not give time to coordinate with the building redevelopment. But the Air Rights Building is already aging and will not continue to survive "for decades to come". When the Air Rights building is redeveloped in the future a more spacious station can be built and the single track section can be removed.

by Wayne Phyillaier on Mar 13, 2012 2:08 pm • link • report

@charlie

The riders per train metric was meant to measure transit demand, not justify fare increases. The relevant metric for fare increases is operating deficit. The deficit has to be plugged with additional monet (fares or subsidy) or decreased costs.

The riders per train metric is meant to factor into an altogether different debate about transit vs. roads.

by Falls Church on Mar 13, 2012 2:16 pm • link • report

@Gray and Wayne,
Why can't the design dwell time be increased from 6 minutes to something like 9 minutes, so that an incomming train would have to be 4 minutes late (instead of 1 minute late) before it has to wait for the outgoing train (and thereby have its dwell reduced to 2-3 minutes which would probably cause it to leave Bethesda late.
Jim

by Jim T on Mar 13, 2012 2:19 pm • link • report

@ Jim T

I don't completely understand the operational simulation, but I think the interaction between incoming and outgoing trains is more complex than that. There will be two tracks at the platform itself, one on each side of the platform. (There is a little more width available at the platform location than under the Air Rights Building.) The Purple Line would be single-track only on the approach to the platform. A second train can enter the station and take the second track position at the platform. The first train does not have to leave to make room for the second train at the platform, but it would have to hold long enough to allow an incoming train to clear the single-track section.

by Wayne Phyillaier on Mar 13, 2012 2:40 pm • link • report

@Jim T: How do you plan to maintain 6 minute headways with 9 minute dwell times?

@Wayne: You do realize that "dynamic pauses" are another way to describe lengthening trip times, right?

by Gray on Mar 13, 2012 3:53 pm • link • report

@Jim T: How do you plan to maintain 6 minute headways with 9 minute dwell times?

9:00 Train on track 1 departs while train on track 2 stays
9:01 Train leaving on track 1 passes the crossover
9:02 Train arriving on track 1 passes the crossover
9:03 Train arrives on track 1.

9:06 Train on track 2 departs while train on track 1 stays
9:07 Train leaving on track 2 passes the crossover
9:08 Train arriving on track 2 passes the crossover
9:06 Train arrives on track 2.

9:12 Train on track 1 departs while train on track 2 stays
9:13 Train leaving on track 1 passes the crossover
9:14 Train arriving on track 1 passes the crossover
9:15 Train arrives on track 1.

and so on

Note that the arriving trains could be a few minutes late and it would just shorten the dwell time.

by Jim Titus on Mar 13, 2012 9:48 pm • link • report

Ah, so you want to run trains in opposite directions on the same track within seconds of each other.

That seems like the simplest and safest solution to a problem that doesn't exist.

by Gray on Mar 14, 2012 8:55 am • link • report

@Jim Titus

Do you work in rail operations? Because I'm pretty sure you can't run trains that close together. You have no buffer room in there so if something goes wrong your entire operation is screwed. One minute a train passes through a switch going one direction and the next minute a train goes through in the opposite direction? Doesn't work.

by MLD on Mar 14, 2012 8:58 am • link • report

I was only trying to suggest that dwell be increased from 6 to 9 minutes and explain why--I had no idea that you all doubted the feasibilities of crossovers at all.

@Gray. Why the sarcass? I've answered what I assumed to be a reasonable question and instead of engaging in substance you switch to ridicule. (Or was your original question rhetorical and I misconstrued it as substantive.) Whether you think there is a problem or not, many people would like to see both a trail and light rail in that tunnel and are looking to see what is feasible.

@MLD. I'm not sure if you read MTA's analysis of how they would do it. I think all I have changed is the dwell time, and rounded things off to the nearest minute.

As you probably realize, the problem of crossovers timed to the minute is something that Metrorail faces at the end of the line. They have run trains with 5-minute headways, and while their crossovers are short, their trains are long, so the time it takes to travese the crossover is about the same.

and their trains are longer

Perhaps you could explain more why you doubt the feasibity of these crossovers

If you really want to get at what is feasibl;e

If you don't like my notation and level of rounding, try explaining in your own words how Metrorail has been able to run trains with a crossover and headways with 5 minutes, recognizing how much longer their trains are

by Jim Titus on Mar 14, 2012 9:41 am • link • report

PS: Disregard the last 6-7 lines which were text I meant to edit out.

by Jim Titus on Mar 14, 2012 9:42 am • link • report

@MLD You have no buffer room in there so if something goes wrong your entire operation is screwed. One minute a train passes through a switch going one direction and the next minute a train goes through in the opposite direction? Doesn't work.

That's why I was suggesting that the dwell time be increased from 6 minutes to 9 minutes. With 9 minutes, if the incoming train is 3 minutes late and that outgoing train is 2 minutes late, the incoming train can wait for the outgoing train, and still come in with a dwell time of about 3 minutes for an on time departure.

by Jim Titus on Mar 14, 2012 9:48 am • link • report