Saturday, April 11, 2015

The most important post of 2003

Josh Barro is not alone in being stuck in the mindset that transit always needs more riders. It seems to be a hard thing to grasp, but sometimes you have enough riders to fill your buses and you have to create more capacity so that you can take more people out of their cars. These people must have a hard time going on weekend trips with tiny purses and briefcases, or buying large sodas when they're really thirsty.

You might think that Barro would be familiar with crowded transit since he lives in Sunnyside, where the 7 train has been crowded and unreliable for the past few years. At least some of our neighbors have figured out that we need more capacity, and organizing a Facebook campaign that has attracted over 800 members in just a few weeks. Some of their anecdotes and photos, on top of my own experience, have convinced me that the NYMTC's capacity estimates are inaccurate, out of date, or otherwise unrepresentative.

So what can we do to increase capacity on the 7 train? Some people say that once the MTA finishes installing the new Communications-Based Train Control signaling in 2017, we will have more trains. At the very least, CBTC will help things run more smoothly. But there are reasons to be skeptical.

As Capt. Subway and Alon Levy have taught us, a train line requires both trunk capacity and terminal capacity to function properly. CBTC may help increase our trunk capacity (but keep reading), but how much use is that if we're still constrained by our terminal capacity?

Well, the MTA actually tested that almost exactly thirteen years ago: they spent the morning of Saturday, April 13, 2002 trying to run thirty trains an hour on the 7 line. I remember when they did it, but didn't hear much about the results. If you're wondering why, here's a report by an independent observer named Stephen Bauman (still posting today) who watched the test from the 111th Street station and compared it with his observations of the normal rush hour on the previous day.

Bauman calculated that the MTA was able to increase the number of trains per hour from 25 to 28. Since they were running ten cars per train instead of the normal eleven, that represented a decrease in capacity. With more train cars and newer ones, they might be able to run 28 eleven-car trains today.

A bigger concern that Bauman conveyed was that the MTA was simply not up to the task, organizationally. As he observed, the dispatcher's clock in Main Street didn't even show seconds, the published timetable is vague and the internal timetable may not be any better, the trains were likely not timed right leaving 111th Street, the conductors did not wait for a signal before closing the doors, and "they ran out of trains around 8:30."

There was one train that sat in the station for six minutes. Bauman writes, "I would definitely catagorize the delay of nearly 6 minutes in getting operating personnel to operate a departing train to be part of the TA's lack of operational ability. There were about 5 supervisors on the Flushing bound platform. There weren't any on the platform where the trains were supposed to leave for Manhattan."

Some of these shortcomings are self-correcting: if the MTA tried this on a weekday the passengers would prevent the trains from leaving early. Others may just be kinks that could be ironed out over time. But overall the outcome is discouraging. We should expect and demand more, but we may not be able to get more any time soon. That means we'll have to look into other improvements, like bus lanes on the bridge and the tunnel, and increased frequency on the Long Island Rail Road.


arcady said...

Internal NYCT timetables have (or used to have) timings with half-minute resolution, but that's useless if the dispatcher's clock doesn't have that resolution. On the London Underground, the busiest lines' timetables (30 tph and up) go down to 15-second resolution. In terms of terminal capacity, Flushing has 3 platforms, so one would hope they'd be able to turn trains pretty quickly, and maybe the new terminal at 34th St will help things. Maybe tail tracks past Flushing would also improve terminal capacity. In London, the Victoria Line does well over 30 tph with a two-track terminal at one end (Brixton).

capt subway said...

Actually I remember that "experiment" well. It was done at the behest of the Car Maintenance Dept, which didn't like maintaining 11 car trains. As they were gradually converting all the R62 & R62A cars to 5 car train sets with full width cabs at each end of the set they couldn't seen to figure out where the 11th car would fit in. (Evidently, for some obscure reason, a 6 car train set was not possible.) The theory was, if you could up the TPH from 27 to 30 you could, at least in part, make up for the missing 11th car. I was in charge of IRT Schedules at the time and supervised the production of the timetables and work program for the test. And, yes, it didn't not work out very well. TSq could only process about 27-28 TPH, under optimal operating conditions - or, as we used to say (and they probably still use the phrase) - with everything "on time and in place".

It is to be hoped that the new terminal at 11th Ave will do a better job at processing the trains than does T Sq. With some lay-up capacity south of the new terminal at least a small number of trains will be able to drop out after the AM peak and be stored there, rather than be run back to Queens, which would help just a bit.

Unfortunately the mindset amongst TA middle and upper management is to (1) come up with new ways to run fewer, not more trains, and (2) to come up with new ways to slow the trains down rather than speed them up - all done, allegedly, in pursuit of chimerical standard of greater safety.

What they could do to improve service right now is to abandon that ridiculous service pattern in the AM leaving MS, wherein the southbound expresses run on the local track to WPt and then are crossed back over to the express track, such move being made to make room for the post AM peak drop outs to terminate on the express track at W Pt, instead of terminating on the local at 111 and then running straight into the yard, as had been done up until about 2008, when it was changed on H H Roberts' watch (a totally clueless bonehead if there ever was one). The southbound expresses should leave MS on the Tk M and stay on Tk M, as it is a much smoother, speedier and reliable operation.

capt subway said...

BTW - viz-a-viz Arcady's observation: NYCTA subway timetables do use half minutes. And the clocks all had second hands by the time of this experiment. In fact, to the best of my knowledge, all terminals, signal towers, the command center, etc, etc, now all work off the same central digital clock, showing hours, minutes, seconds.

Unfortunately terminal operations are still pretty sloppy - with supervision regularly loosing track of crews - as well as crews would simply couldn't be bothered to get on their scheduled trains on time.

arcady said...

That's so frustrating that they keep slowing things down and keep running fewer and fewer trains. Of course the safest and cheapest thing is not to run any trains at all.

Anyway, I think the limiting factor for the Flushing line may well be the Times Square terminal: the problem is that the trains are relatively long and relatively slow due to the curves in the area around the crossovers, and due to needing to be careful approaching the end of the track. The new 34th St station will hopefully have a higher approach speed through the crossover and less need to enter the platform slowly and carefully since there will be plenty of overlap distance past the station through the southern crossover. Once that's done, terminating the locals at 111th is likely to be the best way to get the best use of terminal capacity on the Queens end. And of course, CBTC probably won't do much for terminal capacity one way or the other.

jazumah said...

"as well as crews would simply couldn't be bothered to get on their scheduled trains on time"

When I was a little kid, crews at Utica Avenue waited for their trains. Now, trains wait for their crews there. This should never happen and it drives me crazy.

Why can't five train operators start their workday at Times Square so that a crew is ready to grab a train as soon as it pulls in? I can't imagine it would take a T/O more than 90 seconds to get set up and pull a train out.