Communications-based train control is coming to the 7 train. For years, Transit has talked up this technology improvement, and this weekend, installation begins. Per the press release:
MTA New York City Transit announces that this coming weekend will be the first of five planned service suspensions on the 7 line between Queensboro Plaza and Times Square this fall. There will be no 7 subway service between Times Square and Queensboro Plaza from 11:30 p.m. Fridays through 5 a.m. Mondays during the weekends of October 7-10, October 28-31, November 4-6, November 11-14 and November 18-21, affecting an estimated 280,000 customers each weekend. The E, F, N, Q, S and free shuttle buses will provide alternate service.
This fall, as we continue our maintenance efforts in the Steinway tunnel, we begin installation of a new signal system known as CBTC – Communications Based Train Control. This automated train control system ensures the safe operation of trains using wireless data communication that will allow for more frequent service and the use of countdown clocks in the future. Fiber optic and computer equipment will be installed on the tracks along the entire line. This work requires service changes in October and November and will continue for several years. We realize this will be an inconvenience, but the work is necessary to modernize and improve the reliability of the 7 line.
Eventually, when all is said and done, CBTC will allow the MTA to run more trains on 7 line — a necessity as the route will soon be a mile and one stop longer — than they currently can. “Several years” of service changes to accommodate this week sounds pretty painful though. Is that the cost of progress or indicative of the slow pace at which the MTA works?
32 comments
The MTA should be outfitting, or at least partially outfitting, each line with CBTC whenever they do routine work, even if not implementing the system, so that it can be put into use far faster and sooner than it anticipates.
They should, but they don’t. Part of the problem is the funding. CBTC ain’t cheap. There is also the problem of workers getting in each other’s way. Tunnel space is limited, especially the Steinway Tunnel.
Does anyone know the TPH capacity of the new 34th St terminal? The existing Times Sq terminal is roughly 30 TPH.
It should be the same since it’s supposed to be the same 2-track arrangement with tail tracks.
Last I heard the tail tracks were eliminated to keep the project within budget, cutting capacity. Of course that was years ago. This is just getting started now?
The tail tracks were not eliminated, and construction on the extension started years ago.
Yeah, no way tail tracks were eliminated. If I’m not mistaken, there will be a storage facility farther south towards 23rd Street, which could prove to increase capacity as trains could simply go to the facility if need be.
Well if the tail tracks are in, that’s a huge relief. I had heard they were dropped in a financial spat between the city and the MTA over over-run costs.
With the tail tracks, the capacity should be up to 40tph, if the MTA can ever afford to buy more trains or maintain the ones they have with fewer spares.
Well, how many tph will they get?
The pace is unacceptable. What’s the install time been on other transit networks like WMATA and Paris Metro? Is this normal? At that rate CBTC won’t come online for all lines until the turn of the century.
You mean: What’s the install time been on other transit networks that run 24/7? Because a lot of the work can only be done when no trains are running, so a system that shuts down every night has a huge advantage over ones that rely only on GO’s. And some systems (I don’t know about your specific examples) are willing to close a line around the clock for a month or two at a time, which makes it much easier to get the work done quickly – for better or for worse, NYCT isn’t willing to do that. (For the better, in my opinion, since there’s no suitable alternative for many Flushing line riders, and they can’t be asked to just stay home for two months.)
(I don’t think WMATA has CBTC anywhere – its original signal system isn’t terribly old yet.)
Remember that CBTC is a brand new signal system and is being installed first and foremost to bring the signal system to a state of good repair. So, for instance, on most of the IRT, the signal system was modernized in the 50’s and 60’s, and it’s simply not prudent to install CBTC there just yet. There are many lines that have had new signal systems installed in the past 20 years or so – West End, Sea Beach, Brighton, Concourse, Jamaica, CPW, Jerome, Pelham, White Plains – and they won’t be seeing CBTC for a long time. Once Flushing is done, expect to see much of the IND converted to CBTC, because much of the IND is still using its original 1930’s signal system. The whole process is going to take decades, but there’s no particular hurry on lines with new signal systems.
I believe WMATA does have CBTC but Paris is still installing it.
Line 14 in Paris has CBTC, and a few others being retrofitted. WMATA still uses their original signal system, which employs automatic train stop/supervision/control. It’s a fixed block system, CBTC is moving block. BART uses an earlier version of this type of signal system, with lower TPH limits, but is also working on developing a CBTC system in conjunction with their new fleet coming into service at the end of this decade.
Line 14 is in fact fully automated, with Line 1 being retrofitted (with some GOs, supposedly something that only exists in the special snowflake on the Hudson). The RER A’s central segment has moving-block signaling, allowing 30 tph, but no automation.
I don’t know whether the shared RER B/D tunnel, which has 32 tph at the peak, has moving-block signaling. But it’s not directly comparable to the rest, because it has no stations: the B and D stop at different tracks at both ends of the tunnel. The local equivalent in New York is Penn Station.
But of those, Line 1 is the only one that has had its original signal system entirely replaced.
Building a brand new line with the most modern signal system available isn’t particularly challenging. Upgrading an existing line while keeping it in service is the challenge.
I’m well-aware. I’m just pointing out something about train capacity on the RER. I believe the only place where there was a retrofit was the RER A.
I don’t know about months-long shutdowns of any but the most scantly used lines, but even the MTA shouldn’t be afraid of scheduling lots of overnight shutdowns. People would bitch, but it’s a fair compromise.
The bright side about the 7 is at least it has a lot of potential for getting the work done piecemeal. Between the Times Square and LIC, the E is at least vaguely parallel in service and scope (or the M from Bryant Park), so at least some redundancy is offered in Queens by various transfers to and from the QB lines.
Don’t worry, there are going to be lots of overnight shutdowns.
But the busiest station outside Manhattan is Main Street, which is nowhere near the E. Overnight shutdowns mean lots and lots of buses.
And overnight shutdowns are relatively unproductive – in the typical 5-hour overnight GO, the first hour is devoted to shutting down and securing the track and the last hour is devoted to preparing the track for service, so there are 3 hours of actual work taking place. Since both of those hour windows are safety-oriented (safety of the workers and safety of the morning riders), they’re not optional.
Which is why at times they also do weekend shut downs. 24 hours Saturday and 24 hours Sunday (if you have around the clock shift works) can equal 6 days of straight 8 hour shifts. Luckily they can take advantage of redundancy (the E being nearly parallel to the 7 line from Times Square to LIC, ditto for the R).
Flushing will be a bitch, I guess. But it seems otherwise fairly moderate-ridership segments will be affected. Between, say, LIC and Roosevelt, I doubt it will be as bad as the L.
Serious question: has the MTA looked into shutdown and track maintenance practices in London, Berlin, Paris, Tokyo, and other cities with short overnight shutdown periods? For example, the Yamanote Line runs 20/7, and it most likely uses more than two hours of maintenance per night. On the other hand, it uses catenary rather than third rail.
I can’t speak to the details, but there’s a lot more room for error with GO’s that take out individual track segments one night at a time than with nightly systemwide closures. With GO’s, the track has to be secured to make sure that an errant train operator doesn’t run into the GO area, procedures have to be followed to ensure that workers don’t mistakenly stray outside the GO area, etc. These may sound like silly mistakes, but they’re also potentially deadly, so they’re taken seriously. With nightly shutdowns, the workers only need to know when the last train has passed (and power removed, if necessary) and when the first train will come in the morning.
Depending on how flexible the signals are, they could be set to red ahead of the track.
If the GO area has an interlocking at either end (as is often the case), then, of course, the tower operator shouldn’t be giving trains access to the track that’s out of service. But tower operators are human and sometimes make mistakes, so there need to be procedures in place to ensure that they can’t possibly inadvertently send a train down a track that’s out of service. Specifically, that means physical lever locking devices on the interlocking machine. And what if there’s a switch failure? So there needs to additionally be a physical obstruction on the track to prevent a train from proceeding further.
If the GO area is not bound by interlockings, there is no way to set a signal to red (aside from climbing onto the roadbed and bridging the rails, fooling the signal system into thinking there’s a train in the block). Automatic signals are automatic.
This sort of mistake is, of course, uncommon, but it only takes one to kill somebody.
I agree that overnight G.O.’s are unproductive, and you’re very right on why we only get 3 hours of work. I actually work on the Flushing CBTC contract, and I can tell you that very few night (and I would assume no 10A-3P) G.O.’s will be used for the portions in Manhattan.
All weekend G.O.’s in Manhattan will need to take out both tracks from QBP to TSQ (TA will not flag with adjacent single-track operations). Any work beyond QBP is track-by-track G.O.’s, and generally lead to minimum disruptions in service (mostly no local stops in a particular direction, so riders need to do the double-back).
Finally, as last weekend was our first in the tunnel, I can tell you that (unlike certain other projects I see while as a straphanger) these G.O.’s are utilized to the fullest. There was probably 1000+ MH spent by contractors, and maybe the same by NYCT workers who were also there.
I suppose the biggest thing for me is just the time span it takes to do the work. I’d personally be better with the idea of an extended period of no service on certain late nights. Better that than a piecemeal approach over several years without any noticeable difference until the service is finally online.
I take service disruptions in stride as long as I’m actually notified beforehand via poster or the MTA’s dashboard (unlike last night’s Queensboro Tunnel down to one track delays).
Last I read QB line is next in line after flushing for CBTC. Any clue if that’s still the plan? I haven’t been able to find anything on this.
As far as I know it is still the plan.
CBTC takes for-EVER to install on old transit systems.
Hell, even plain induction-loop-based cab signalling takes forever to install. I think it’s been over 30 years of work for Chicago’s L already. Boston is *still* running the Green Line on waysides and line-of-sight.
At least they stopped calling it the “Steinway Street Tunnel” – somebody must have finally corrected them on that!
Finally? I think that was a one-time error.
I thought there was at least YouTube video and the detailed poster in the trains. At the end of the day not a big deal. I actually found it kind of amusing as it’s nowhere near Steinway Street.
I saw the poster. I figured it was just a mild Freudian slip.