Home Capital Program 2015-2019 The long tail of the subway’s history of deferred maintenance

The long tail of the subway’s history of deferred maintenance

by Benjamin Kabak

During its meeting on Wednesday, the MTA Board will vote to approve a $205 million contract with Siemens and Thales Transport for a part of the communications-based train control installation along the Queens Boulevard line. As the MTA works to pre-qualify other companies with relevant expertise, these two companies — the only two pre-qualified to bid on contracts since 2006 — will spend 67 months installing CBTC along the western part of the Queens Boulevard line. This may mean five years of intermittent service changes, but it also may lead to, as the MTA claims, a “‘state of the art’ train control system.” And we all know it’s about time.

As part of the discussion about CBTC, Transit released the video I embedded above. It explains how the stations aren’t the only part of the subway system. In fact, much of the technology is older than any of us (and nearing, or well past, the end of its useful life). It’s supposed to explain how CBTC can help increase frequency of trains along the Queens Boulevard line, but it also underscores exactly how deferred maintenance throughout the decades can come back to hurt a transit agency.

For decades, the New York City subway system was a victim of politics. It initially operated as a quasi-private business overseen by the elected Board of Estimate with fare policy set not by the operators but by politicians who had to answer to constituents. So the five cent fare lingered and lingered and lingered, even when it became obvious that the transit operators couldn’t run the system breaking even, let alone generate revenue to keep up with technology innovation. This was a repetitive cycle until the MTA came into being in 1968, and the nascent agency had its work cut out for itself.

Since the early 1980s, the city, state and feds have poured billions of dollars into restoring some semblance of investment in the transit system, but the backlog was tremendous. Nearly since its founding, the subways hadn’t undergone top-to-bottom overhauls, and the early money was spent on upgrading rolling stock and restoring stations to a usable state. We’ve had massive track replacement and big-ticket capital projects too, but the signal systems have lingered as a relic of history. Now, the MTA is trying to play catch-up, and it’s going to take a long time.

Elsewhere, transit agencies can focus on technological upgrades without having to sacrifice other projects in part because costs are kept under control. Paris, for instance, can fully automate subway lines quickly as they’re working from a base of newer equipment and shut down the system at night. They don’t face the same labor and corruption issues that New York hasn’t been able to combat.

As you watch the MTA’s nine-minute video, think about how the choices we make today reverberate for decades into the future. We need a state-of-the-art subway system now because it’s the system we’re leaving to our children and grand-children. It has to function then at least as well (or perhaps as poorly) as ours does today. And that incremental $200 million investment for a part of the Queens Boulevard line doesn’t even begin to cover it. Now does anyone want to close that $15 billion capital program gap?

You may also like


Stephen Smith July 22, 2015 - 8:31 am

Why can’t we shut down the Queens Boulevard Line at night or on the weekends? It runs beneath an infinitely wide road that we can run substitute buses on, and parallel to an MTA-operated rail line to Midtown that’s underutilized at every time but rush hour.

anon_coward July 22, 2015 - 10:17 am

they already do that on single tracks. they will shut down the local or express tracks in a segment and you don’t need to take the bus, just take the local or express to the next stop and transfer to the opposite direction.

and car ownership goes up as you go east and people going into manhattan on weekends drops as well

SEAN July 22, 2015 - 10:23 am

Simple – ridership & overall density along the line.

al July 22, 2015 - 10:25 am

The IND Queens Blvd runs under Broadway and Northern Blvd too. They’re not nearly as wide, and are more congested. In the case of the Queens Plaza approaches, its very congested. Finally, there are the local Community Boards, who would scream and holler at the loss of overnight and weekend parking spaces. DeBalsio and Cuomo aren’t going to do the heavy lifting to back the MTA up.

Larry Littlefield July 22, 2015 - 10:35 am

Having worked in capital budgeting in the signals and systems part of NYC Transit, and seen an asset database summary that was current as of the time I left, let me add some details.

NYCT engineers hold that signal systems have a 50-year useful life, before deterioration and obsolescence require replacement. In reality, however, they’ll admit that with good maintenance they can go 75 before things really go bad.

For most of the subway’s history, they have been replacing signals at about a once-in-60-year pace, something in between. The subway was built from 1904 to 1956 (Rockaway Line). By the end of this period all the equipment was original, and the original IRT sectors were shot. Money borrowed for the Second Avenue Subway was instead used to rebuild the IRT. And so signal replacement began and continued…

Until the 1970s fiscal crisis, when it stopped. As a result of the huge debts and retroactive pension increases of the 1960s. Huge debts and retroactive pension increases that have been repeated.


Then signal replacement resumed at the prior 60-year rate, which requires one new major signal project every year. But because of that fiscal crisis hiatus, much of the IND — still operating on original signal equipment — is going over 75 years.

Worse, ongoing normal replacement is heading for another shutdown. All it will take is the return of stock prices to normal levels to show how problem NYC and NY State actually are. We are in the same situation we were in in 1972. You know what happened next.

Larry Littlefield July 22, 2015 - 10:40 am

By the way, I wrote about this several years ago. And said that what the MTA needed was not more cutesy subway map revisions, but maps that show how old the infrastructure in each part of the metro area is. A visual asset database someone might actually pay attention to.


As I said at the time:

What I want is a new version of the New York City subway map. But instead of showing stations, dots on the map would show signal interlockings (places where the trains can switch tracks), with the size of the dot equal to the size of the interlocking measured by its number of switches.

The color of those dots, and the colors of subway lines connecting them, would not be based on the Manhattan trunk route as it is on the actual subway route map. Instead, the dots would be based on the age of the interlocking signal system and the lines would be based on the age of the automatic signals, stops and stop cables (you don’t have to know what those are, because the MTA does).

For signal systems 0 to 19 years old, the color could be blue, as in blue chip. For those 20 to 39 years old, the color could be green, as in good to go. For those 40 to 59 years old the color could be yellow, for beware. New York City Transit estimates the useful life of its signal systems to be 50 years, but aside from the 1970s fiscal crisis it has been replacing its signal systems at a rate of a new system every 60 years. Because of that hiatus, some systems are much older. Those from 60 to 75 years old can be presented in red on the map. Some engineers say conventional signal systems can last 75 years before disaster. Those 75 years or older can be presented in (fade to) black.

APH July 22, 2015 - 3:01 pm

If you put your map data online somewhere (Google spreadsheet or wherever), I can probably make it into the map you envision.

Larry Littlefield July 22, 2015 - 3:48 pm

What I have is more than a decade out of date.

They would have to do it. And they should.

Larry Littlefield July 22, 2015 - 10:47 am

One more point: Siemans has repeatedly failed at NYCT signal projects with far higher price tags. But what does that $200 million cover? How many interlockings? How much of the work? Is it just a design project?

If it is for the entire western half of the QB line, from Forest Hills to Queens Boulevard, there may have been a dramatic DECREASE in what is being charged. That would be big news. Maybe it has been decided that we’ve been robbed enough.

It doesn’t mean they will complete the project for that amount, but I wonder what we are getting for $200 million.

James G. July 22, 2015 - 11:01 am

Something I’ve been wondering: Did they build the 7 line extension and 2nd Ave Subway Phase I with both traditional signals and CBTC infrastructure?

Or will they have to shut down the 2nd Ave Q and add CBTC when they modernize the rest of the Broadway line?

Also, has the MTA released any sort of list of the order in which they’ll deploy CBTC after the Queens Blvd Line?

Larry Littlefield July 22, 2015 - 11:35 am

The Flushing extension is CBTC. The Second Avenue Subway is not — it is a three station extension of a non-CBTC line.

The MTA had a whole rollout plan. The Culver line was next, because it is next in order of aging signals, but they seem to have skipped ahead to Queens Boulevard. Basically it would be the IND minus the Concourse Line, where the signals were already replaced.

Then the whole system would have second-generation signals. But on the IRT in Manhattan, the heart of the system, those second generation signals would be pushing 75 years old, so that would be next.

Don’t you think Phase II of the Second Avenue Subway should be finished before they start shutting down the Lexington Avenue Line to replace the signals?

Seth R July 22, 2015 - 9:48 pm

To repeat James’ question, did they build the 7 extension with both? I didn’t think all 7 trainsets had finished being equipped with CBTC. And doesn’t the whole line have to be cutover to CBTC to make it work or can the trains transition out when they get to 8th avenue?

Would love some further reading on this.

webster July 23, 2015 - 8:09 am

They have installed CBTC equipment on the SAS, as well…
Service simply will not take advantage of that until other upgrades take place, but it definitely has the infrastructure in place.



Chet July 22, 2015 - 11:54 am

Well, in any event, it is quite a good video and worth the 9+ minutes to understand how much work the system needs.

Hmmm…Apple announced that they have over $200 billion in cash. Maybe we can just turn the entire MTA over to them…

Larry Littlefield July 22, 2015 - 2:50 pm

What we need is for some of those tech companies, that somehow think they can create systems to allow autonomous operations of vehicle in mixed traffic, to get into the far simpler business of railroad signaling.

Anonymous July 22, 2015 - 4:50 pm

Nah, turn it over to MTR (or a comparable company). Give them a contract to operate, maintain (and if they feel it’s necessary, upgrade and/or build) the system (or one of the divisions minus the self-contained lines). Make the contract lengthy (20 years or so, preferably more, so there will be an incentive to spend money on upgrades). Include well-defined penalties for bad performance (frequency on trunk lines should be the metric) and limit annual fare increases to inflation only (unless a case can be proven for further raising of fares).

Oh wait. This sounds too familiar. But then again, what would the subway system be like today if the Dual Contracts were written to consider this thing called ‘inflation’ and the BMT/IRT hadn’t been bought out (e.g. in 1954 the contract was extended for another 25 years, a possibility which the contract states)

Larry Littlefield July 22, 2015 - 9:23 pm

If I were going to privatize something, it would be the LIRR. Though the Kelois thing in Boston isn’t going so well yet.

Spendmor Wastemor July 22, 2015 - 10:02 pm

What’s going wrong with Kelois – Boston?
MBTA is a total hole, so I can imagine that whoever takes over their operations has an unfixable, corruptified mess.

Larry Littlefield July 23, 2015 - 7:38 am

The system had been privitized to a new company run by former MBTA managers, which was so bad they lost the contract to Kelois (too bad that can’t happen with non-private managers).

The system is a wreck, due to underinvestment and poor management. So Kelois was in a hole.

Then the worst winter in Boston history struck. Basically, service on many lines was shut down for weeks on end, a disaster for those who rely on transit. What happened to Boston last winter is a microcosm of what could, and perhaps will, happen to NYC.

They get a bit of a pass on that. But disastrous service continued into the spring and summer, because Kelois failed to hire enough engineers to replace those who were retiring. My contact in Boston puts that on them. The company, which is French, fired replaced their local manager.

So we’ll have to see what happens.

As I said, the system is a wreck due to deferred maintenance. One reason the snow eliminated service is that it trashed the DC traction motors on the trains, taking them out of service. All the other transit systems have gone to AC motors to avoid this.

And even when they do buy new locomotives, cars, etc. they get ripped off and they don’t work.

Anonymous July 23, 2015 - 10:11 am

It takes quite a bit of time to fix long-time deferred maintenance. NYC shows this. And privatizing something isn’t the instant fix for everything, but the private sector has a lot less patience for the unions than politicians (because the CEO doesn’t need expensive political campaigns partially financed by the unions). If the contracts are written properly, they also have an incentive to run on time (vertical integration is also very important on a railway).

As for the trains: it takes a while for new trains to arrive. Permanently staffing stations is a good interim solution (as London’s new Overground lines show) as well as improving stations, but it will do nothing to fix unreliable trains. Again, in London the new Overground lines run fairly old trains which were completely run down by the previous operator. New trains have been ordered but will only arrive somewhere in 2018. Yet again as an interim solution the current old trains are being cleaned and repaired (though the electrical equipment may keep them unreliable for the time being), but at least the trains won’t look bad.

And think of the other benefit, taking something away from politician’s control means it can’t be abused for one’s political purposes. Imagine a subway (heck, even an entire MTA) kept as far away from Cuomo’s control as possible.

Larry Littlefield July 23, 2015 - 10:58 am

“If the contracts are written properly, they also have an incentive to run on time.”

The old contract had only token fines. The service stunk. The new contract has very large fines. Kelois is paying up the wazoo.


A total of $7.5 million in fines.

APH July 22, 2015 - 3:05 pm

Would be good for the governor to watch this 🙂

Alex July 22, 2015 - 5:14 pm

He’d just shrug and ask why more New Yorkers don’t drive to work.

Brooklynite July 22, 2015 - 10:38 pm

The video makes CBTC out to be the ultimate solution to all of MTA’s problems. What we should be asking ourselves is why Moscow, Paris, and even 1954 New York (!!) run/ran significantly higher service levels than we can today.

Alon Levy July 22, 2015 - 11:37 pm

Paris’s busiest lines have automated trains (Line 1, Line 14, soon to be Line 4) or CBTC (RER A).

But I’ll grant you Moscow.

Nick Ober July 23, 2015 - 9:49 am

Has anyone found a video on YouTube of a terminal operation in Moscow? It would be fascinating to see how they turn trains so quickly.

tacony July 23, 2015 - 10:56 am

Other than the height of rush hour on a couple lines, the service levels on the NYC subway are mostly constrained by the MTA’s own load guidelines and service standards. CBTC won’t change the worst-among-its-peers weekend and evening frequencies. The MTA’s concern seems to be merely to reduce dangerous platform overcrowding. The IRT had slightly better headways on the day the subway opened than we have today.

Brooklynite July 22, 2015 - 10:40 pm

On the topic of MTA’s perennial budget issues, if they didn’t have capital costs 10x the rest of the world’s the SAS would probably be funded for all four phases right now, as would Triboro RX. Because there is nobody actually accountable for the MTA capital budget (governor blames MTA, MTA blames governor) there is no incentive for improvement.

Abc July 23, 2015 - 9:37 am

So we expect the same agency to run one of the few 24/7 systems, organize one of the largest systems, execute efficiently to have trains arrive on time, maintain vaccum tube 30s and 50s era technology, and build some of the most challenging civil engineering projects of our day.

We give them a shoe-string budget to do all of this and then we yell, stamp our feet, and cut the budget even further if they fail in any one of these dimensions. No other public agency is held to this same standard in order to get appropriate funding.

david vartanoff July 23, 2015 - 3:27 pm

While undoubtedly the actual lineside signalling equipment needs upgrading, one should note that in the 50s/60s CTA ran more frequent and faster service with similar hardware than today with cab signalling/safety enforcement automated supervision. New York ran more frequent service during rush hour in the 50s than today. Until the slow speed enforcement equipment is either modified or scrapped the transit will not be rapid.
As to which multinational conglomerate should be picked to fleece the TA with poorly performing hardware, a hard choice at best.

Larry Littlefield July 23, 2015 - 9:06 pm

I think we should privatize the LIRR because the union/management mafia is ripping us off.

And bring the signal replacement program in house because contractors, consultants and private unions are ripping us off.

Or perhaps we’ll just be ripped off no matter what.

BDawe March 17, 2016 - 5:10 pm

An institution which lacks the institutional strength to manage it’s own operations is one that’s also going to lack the institutional strength to write good contracts.

johndmuller July 28, 2015 - 12:39 am

I wonder sometimes about whether CBTC is really the answer to everybody’s problems like it is sometimes billed. As I understand it information will flow to and from the trains and this will lead to … what, exactly?

One thing it could lead to is enforcing all the speed and separation rules, perhaps to the letter. Those high throughput operations in Russia might just be cutting it close to the edge and relying on the quick reflexes or daring of the operators rather than whatever control system they might have. I the Boston streetcar subway, a good control system would dramatically slow down the lines because the operators would be unable to tailgate or dock two trains in the same station.

One hopes that the spacing and stopping distance rules (which effectively set the maximum speeds) are not counter-productive if accurately observed. I suspect that maximum carrying capacity is achieved at something less than minimum trip time, but one could hope to be wrong about that.


Leave a Comment