With Byford’s blueprint in hand, will Andrew Cuomo save the subways?
The politics of the moment, however, found a way to intervene, and the dynamics of the Democratic primary reared its head late last week. Shortly before Cynthia Nixon announced a kitchen-sink plan to fix the MTA — endorsing the Byford plan while calling for both congestion pricing and a Bill de Blasio-inspired millionaires’ tax to fund transit — Cuomo decided the plan he commissioned was one worth endorsing. Toward the end of last week, he made the call for a congestion pricing plan to fund the Byford proposal to save the subway. You can (and should) read Emma Fitzsimmons’ coverage in The Times. The question now is whether Cuomo will follow through, but since the Byford plan is a preview of the Transit asks for the next two MTA Five-Year Capital Plans, we won’t know for another year or so if Cuomo is serious.
The politics are the politics are the politics. After a while, having to convince the governor of New York to support the economic lifeblood of the largest city in the state gets exhaustingly tiresome. The governor doesn’t appreciate the transit system, and the person who should be New York City’s biggest champion thinks he’s the mayor of some suburban town of ten thousand drivers. Sometimes, I can’t help but thrown my hands up in disgust at the whole thing, but right now, that’s neither here nor there. The politics will play out in the coming months, and forces will likely align behind the bulk, if not all, of Byford’s plan. Which brings me in a somewhat roundabout way to a question: What exactly is in Byford’s plan? Though I wrote about it last week, I haven’t delved into the details so let’s do that.
In a sense, as I’ve mentioned, the plan is a preview of things to come. Byford accelerated the 40-year plan to replace the bulk of the subway signal system and will instead do it in ten years. He wants modernized interlockings and over 300 stations to be brought to a state of good repair. He wants a new fare payment system, 130 new ADA-compliant stations, over 3600 new subway cars (which I hope will include open gangways) and nearly 5000 new buses (which I hope all use clean-air technology). “We propose doing in 10 years what was
previously scheduled to take more than 40, including major progress in the first 5 years,” Byford said. “This means lines that are currently capacity-constrained will be able to carry more people, more smoothly and reliably.”
Visually, the plan looks like this:
I’m not quite sure what happens with the parts of subway lines that aren’t included in the ten-year upgrade approach. Do these non-modernized segments act as chokepoints that still limit the number of trains subway lines can accommodate? If, for instance, the 1 line between Van Cortlandt Park and 96th St. isn’t modernized while the remainder to South Ferry is, can the MTA run additional trains? And what of, for instance, the F between York St. and Church Ave.? Or the entirety of J train?
The plan isn’t without pain, and the pain is the key issue. The MTA considered and dismissed time-barred full-line shutdowns to accommodate the work and plans to maintain weekday train service. But Byford warns that “continuous night and weekend closures” may last for up to 2.5 years per line with both express and local service shutdown where applicable. What the plan does not detail is how exactly the work will go from taking 40 years to taking 10 or whether costs will fall in line with even the upper bounds of international standards rather than current spending which far exceeds that of any other comparable transit system.
Beyond the signal upgrades and CBTC installation, the rest of the plan does what the MTA should be doing but on an aggressively fast schedule. More stations renovated in shorter time frames. More ADA and other accessibility upgrades. Better management (which may be short for cleaning house). Actually delivering a new fare payment system. Route overhauls “to reduce reliance on critical interlockings.” It’s all what the MTA should have been doing for decades.
You can read through the plan document as a PDF right here. It’s a quick read, and it’s a blueprint for the future. Publishing it was the easy first step though, and the harder part is someone else’s political lift. That, as The Times’ editorial board writes today, is the hard part. “Mr. Byford’s plan asks New Yorkers to make sacrifices. They will have to pay more in taxes and fees and endure night and weekend subway shutdowns as workers fix lines and stations. But most people would be willing to bear that pain for a safer and more reliable transit system,” the editorial notes. “What is less clear is whether New York’s elected leaders can summon the necessary political will to turn this plan into reality. It was heartening to see Mr. Cuomo belatedly embrace Mr. Byford’s plan last week, but he has to back up his words with action. Because Mr. Byford is right: New Yorkers can’t wait 50 years for a modern transit system.”
27 comments
I am somewhat disappointed that Byford’s CBTC plan will only deliver full CBTC for one service (the E) in 5 years, and no additional services after 10 years. And since the E shares all of its track with other services that won’t get the full CBTC treatment, I don’t think we can expect more service even on the E. I suspect, in a situation where a subway service runs on a combination of traditionally signaled and CBTC-signaled lines, the benefit comes from the improved reliability of the signal system, especially throughout nearly all the lines in the Manhattan CBD. That will hopefully mean that trains can run a bit closer together there, keeping headways shorter, especially as services combine to offer more service on individual lines. And also remove any threat of signal failure that exists at present.
The Post complained about the MTA not replacing signals on the J/M/Z in the next 10 years. Those signals were replaced in 1989. The MTA cannot afford to tear out recent signals.
I see the plan this way. The MTA has traditionally replaced signals at a 60-year rate — except in periods such as the 1970s and the last 10 years when retroactive pension increases and debts caused capital reinvestment to stop.
All this plan would do is allow the MTA to catch up with that hole, and get back to the 60 year schedule.
By the year 2030, when this plan is done, additional second generation signals will be coming due for replacement.
Trying to figure out what the term “retroactive pension increase” might refer to.
Are you talking about the normal effect retroactive general wage increases always have on any employee’s pension amount?
Or is it that there have been retroactive modifications to the arithmetic formulas by which each tiers’ pension amounts were calculated?
Retroactive changes to the terms. As in when those promised full pensions after 30 years of work at age 62 are suddenly granted pensions after 25 years of work at age 55, as for NYC teachers, with no money having been set aside to pay for that.
And 1,000 other things like it.
What has been promised might be expensive but it could be responsibly funded. But all the retroactive pension increases are disastrous, even if (as in NYC) the taxpayers do their part. It’s why NYC is in such a bad hole — though other places where taxpayers didn’t fund the pensions that had been promised to begin with are also in such a hole.
If the MTA replaced the signals in 1989 then why does the train run about as fast as an elevated streetcar? I live off the JMZ and a trip that should only take 15 minutes takes about 30. Trains always stop on the track and just sit, and wait. They literally never run at full speed (and yes, I’m aware they’ve been paired back by the MTA due to an accident but even then, it’s pretty glacial). To be clear I’m not challenging the accuracy of what you’ve said, but the experience just doesn’t add up.
Are the sitting and waiting at the point where the M merges in?
That is one of only two places left on the system where trains need to cross each other at grade. Inbound, you may be waiting for an outbound M to cross over the Manhattan bound J/Z track.
Everywhere else has expensive “flying junctions” which are like highway exits. If that interlocking had one, the outbound M would diverge from the outbound J/Z, and a connecting track would pass over or under the inbound J/Z track so the inbound J/Z would never have to wait.
https://www.nycsubway.org/wiki/New_York_City_Subway_Track_Maps
He may also be referring to the incredibly frustrating slowdown coming down off the bridge eastbound into Marcy Avenue. I recognize that the curve needs protecting, but that location is another prime example of the overly conservative spread of grade / station timing all over the city that has slowed the entire system down. Trains have to practically come to a complete stop before proceeding.
Many people I work with will say they cannot bear the pain of higher fares.
In an ideal world, Byford is counting on costs coming down and expertise increasing in signals installation such that the scope of the upgrades in the second capital budget could be expanded, and be so successful that by the time the third capital plan rolls around, the work can be completed. Though I think it will be four capital plans before every last inch of track is CBTC.
Byford’s CBTC plan is supposed to result in higher service levels and decreased maintenance costs. Benchmark metrics for these achievements are missing from his plan. It’s very difficult to evaluate the plan’s technical merits without such information.
Ben wrote:
This is absolutely a valid concern, but the examples you list are not the areas of concern.
When I was working on Queens Boulevard CBTC, the CBTC signal equipment covered the most heavily trafficked rails: the segment between the 63rd connector and Kew Gardens. This is the section that needs the increased capacity, so while the 63rd street tube wouldn’t see an increase in absolute capacity, it’s not going to see that increased traffic anyway. The other key concern is that, with the current CBTC system, trains need to enter CBTC territory outside of the choke point. So the F would actually have CBTC installed up to 21st-Queensbridge (it also needs to enter CBTC territory at a station in order to be able to operate in ATO. If it enters while in motion, then the operator will need to operate the train in a manual CBTC mode until it makes a station stop).
So going back to your examples, you are correct that the 1 line will be governed by capacity north of 96th street. However, the 1 line is nowhere NEAR maxing out that capacity. NYCT has been able to achieve 30 trains per hour on their fixed block system before and presently. The 1 currently operates closer to 20 tph, peak. So the signal system is not constraining capacity right now; NYCT could definitely add service on the 1 without doing anything to the signals.
The F is a little different, because it shares tracks with other lines (most crucially the E in Queens). Whenever CBTC on the Queens Boulevard line goes active, the F line will get a boost in service due to the fact that the choke point in Queens has seen increased capacity. Fun fact: before the opening of the 63rd Street Connector, the F operates 18 trains per hour, while the E operated 12. Those additional F trains became the E runs to/from 179th street during the peak hours. But the non-CBTC segment between Jay street and Church won’t constrict the capacity of the line, because even with CBTC, those segments wouldn’t be able to handle the added F trains.
But the one real choke point that you missed is the Brooklyn IRT. By only extending CBTC to Atlantic, the 4/5 service to Brooklyn can never exceed the capacity of the non-CBTC tracks between Atlantic and the Rogers Junction. At best, they’ll be able to add some additional short turns that use the Bowling Green-South Ferry loop, but that won’t help Brooklyn riders on the IRT.
Still, it looks like they are trying to get all of the trunk lines upgraded, which is laudable. The real question will be: what can they do to control costs?
For Rogers Junction, all that needs to happen is 2 switches are added into the existing configuration so that 2/3 trains service Flatbush with the excess trains Flatbush can’t handle go to New Lots, 4 operates local after Franklin all the way to New Lots, and the 5 operates express to Utica. The 2 new switches should be added in a way so 2/3 trains can merge to the local while a 4 crosses to/from the express tracks. Expanding Flatbush to handle more trains (and to further cover more of Brooklyn) is going to more expensive than adding the switches, but should be a long term goal that MTA looks towards.
What is with this “3 to Flatbush” BS? There’s a reason why the practice was stopped. To add to that, there’s a reason why the 5 sticks with the 2 in Brooklyn.
As of right now, all trains that serve the branch to Flatbush and the local tracks between Franklin and Utica must share a small stretch of track. While this is the case, sure it makes sense to offer some service from Lexington Ave. But if the Rogers Junction is fixed so that expresses can be shifted to the local without needing to merge with locals headed to Flatbush, then it absolutely makes sense to send all Lexington Avenue trains to Utica/New Lots.
Except doing that needlessly reopens one wound and opens several more. Again, there’s a reason why the routes were altered to what they are now.
Doing that fixes a massive shortcoming with the Rogers Junction. Every Lexington train that needs to use Flatbush must share tracks with every 7th Ave train. It’s a massive capacity constraint for the 7th avenue line.
I guarantee you’re grossly overstating the benefits. Aside from ridership patterns (those 5 trains can see serious loads even before reaching Franklin Avenue, and not just because of delays) and the general operations between the two routes (look at where the routes go, what rolling stock they use, and where said stock is stored), you may want to look back to 1983 for why this proposal is so poor.
The benefits are clear, and MTA is well aware of what they are. Operations Planning and Capital Program Management are looking at doing exactly this. If Rogers junction is ever fixed, they will implement the rerouting as described, and anyone who needs Lexington from the Nostrand Avenue Branch will simply make a cross platform transfer at Franklin, hardly a poor proposal at all.
All that amounts to is a small amount of shuffling for no real benefit. Actually rebuilding the junction is the only way to address the deficiencies without putting the simpler, more flexible operations in jeopardy. Reverting the 3 and 5 to pre-1983 service patterns will not just cripple them, but also harm the 2 and 4 (though the 4 would still be in better shape than the others). The problems with this plan would present themselves on a constant basis, though it would be especially apparent on weekends and during rush hours.
No one is advocating returning back to the service pattern pre 1983. That service pattern had the 3 and 4 serving Flatbush, while the 2 and 5 served Utica/New Lots, with some 4 trains also serving Utica.
Additionally, if nothing is done to fix Rogers, then there’s no point in changing the current service pattern either. Some number of expresses would need to switch to the local anyway, and they have to merge with 2 and 3 trains. So reverse branching the service makes sense with the current infrastructure alignment.
There are two ways to fix Rogers. One involves making it a true flying junction. It would require an extensive rebuilding of the junction, including new grades, and what is likely to be an expanded tunnel. The disruption to service would be extensive, and the cost would be relatively astronomical. The second way involves adding a single new set of switches from express to local. It requires some structural work, and still results in some disruption, but it’s far less scope and as as a result, much lower cost. If Rogers ever gets fixed, it’s probably going to be option 2, and if that happens then it means most 3 trains will be shifted to Flatbush to avoid merging with Lex trains.
J, what you’re advocating for is a partial reversion to pre-1983 service patterns. I’m well aware of what they entailed, and it was a mess, especially with the inconsistencies between weekday and weekend service.
I don’t want to leave the junction in its current state, either, but tacking on some switches just for a 5-second victory will do nothing to address the root issue, which is the horrid setup that was established back when it was constructed. Correcting this issue would have significant impact, but awkwardly tacking on a couple of switches and swapping the 3 and 5 is like sticking a dirty bandage on a festering wound.
It’s not a reversion to pre-1983 service patterns at all. It’s making the Nostrand Ave IRT all 7th ave service, and making the Branch to New Lots all Lexington. It avoids having these services merge. It also improves weekend service, as the current weekend schedule has the 2 and 3 each operating at 12 minute headways, while this new configuration would serve the New Lots branch with 8 trains per hour, and the Flatbush branch with 10. In fact, many transit wonks would argue that this is a better setup, even though it’s cheaper and simpler.
J, prior to the 1983 service changes, the 3 and 5 followed the routes you’re proposing, so claiming that it’s not a reversion to the pre-1983 service pattern is ludicrous.
There are several reasons for maintaining the current service pattern, albeit with a rebuilt junction. For the 2 and 5, they can maintain operational flexibility that would be lost if decoupled. For the 3, it’s able to easily access a maintenance facility, if permanently sent to Flatbush Avenue, would require needless branching and dumping at intermediate stations, creating bad crowding, long headways, and congestion. (The only way to remedy this would be to make changes at Lenox Yard, a yard that’s already been downsized significantly since its inception. The easiest way to go about it is to close the existing stations at that end of the line, as one lacks space to turn trains, while the other was built from the Lenox Yard itself, and closing that station would allow for some sort of facility to be built while minimizing impact on the existing storage crunch.) The 4 may not have much of a merging issue, but the existing setup allows daytime trains to return to the Bronx sooner, which is especially important for a route that is notorious for delays and inadequate service that go far beyond one junction in one direction.
Trying to put your proposal into effect would not end well. Recall what I mentioned about the 3 and yards. Prior to the changes, the MTA had to make those complicated moves for all maintenance, as Lenox Yard had lost the maintenance facility. (To make matters worse, Lenox Yard is at ground level right by the Harlem River, which means the yard itself is far more prone to flooding than one that’s inland, elevated, or both, so even if the yard was restored to some maintenance capacity, that would be of constant concern.) In the end, the decision was made to move the 3 to New Lots Avenue precisely to alleviate this issue. While the 4 and 5 didn’t actually have too much of a need to be swapped, they were anyway to improve consistency and maintain flexibility. This proposal calls for radical changes that actually fail to benefit people. As things stand, the 3 is already the worst carrier of passengers as part of the Brooklyn IRT, exacerbated by the greater number of people seeking Lexington Avenue stations and faster rides. The current setup avoids putting excessive strain on Franklin Avenue and Nevins Street by giving people options in a more flexible environment, meaning that by the time 2 and 3 trains reach Franklin Avenue, there is little need to transfer. Your proposal would end up undoing this, and the result is predictably disastrous. To make matters worse, it would lead to a number of service cuts. The overnight 3 service would become nothing more than a memory with no hope of changing that (short of making a confusing 5 service change), and the trains currently in use on the route would become short-run 2 trains kept to rush hours and the related edge periods, and only because of the high ridership. That issue would only be exacerbated by trying to mitigate the effects of storage loss at a yard that’s already lost plenty just for maintenance combined with the inability to use either 135th or 145th Streets to properly turn trains. Weekend 2 frequencies would remain untouched beyond typical adjustments, weekend 5 service would remain as-is, and weekend 4 service would be stuck footing the bill (increase in service at certain stations for an overall loss in service provided and service quality). If we were to discuss weekend 5 service in Brooklyn, it would still be more favorable to maintain the current routes and rebuild the junction to facilitate these routes. Nostrand Avenue’s weekend ridership is too weak to justify more than the existing number of trains (which I begrudgingly admit, while midday periods on weekdays at least have the excuses of somewhat higher ridership and sitting between peak periods), and there’s no real reason to have two Lexington Avenue routes in a situation where the frequency of one would suffice.
TL;DR: Your proposal will fall apart almost overnight, and all in the name of detangling routes.
There’s quite a few bad assumptions there, so I’ll just leave it at this: last time this was looked at by Operations Planning and CPM, the determination was to make the simple fix of adding the switch as I described it. If they ever have enough will to get it done, that’s what will happen.
Fixing Rogers Junction is certainly something that needs to be done, but it won’t fix the issue that I mentioned with signal capacity constraints. If Rogers is fixed and CBTC is only extended to Borough Hall, the express tracks between Borough Hall and Rogers would be the choke point. CBTC needs to be extended to Utica (with the ability for Local trains to enter CBTC Automatic operation at Nostrand.
The political alignment is swinging into place. Byford has international technical credibility. The NYT Editorial Board opinion carries weight among the elite. The Mayor is on Byford’s side. Democrats have just taken over the State Senate, and the majority leader is from Yonkers. Cuomo is known for folding under pressure. I think there’s a definite chance.