The MTA’s map of potential service during the L train closure highlights how increased subway service, buses and ferries will help mitigate the effects of the 18-month shutdown. (Click to enlarge)
No matter how you spin it, the looming L train shutdown is not good news for the city. It’s not something Gov. Andrew Cuomo is going to announce with a press release and a staged media event at the Transit Museum, as he has done in recent months with news of new-look subway cars and renovated stations. Rather, the decision to move forward with an 18-month total shutdown of the L train between Manhattan and Brooklyn, set to begin in January of 2019, is one of the MTA released to The Times on Monday morning and the world at large during its Board committee meetings yesterday. It has sent shockwaves down the BMT Canarsie line and has spurred on more bickering between city officials and Cuomo’s MTA.
Following seven months of public hearings and numerous surveys of riders, Transit President Ronnie Hakim announced the details behind the shutdown on Monday. With nearly 80 percent of L train riders supporting a shorter shutdown over partial service that would be woefully inadequate for three years, the agency will complete close the L train’s Manhattan stations and the tunnel under the East River. Trains will continue to run between Bedford Ave. and Rockaway Parkway at eight-minute headways, providing riders with connections to the 3, A, C, J, Z, M and G trains. The MTA will have to rebuild two tubes that are over 7100 feet long and replace and harden all of the infrastructure within those tubes. A shutdown, Hakim said, is “the least risky way to do a project of this nature and the amount of work that needs to be done.”
As part of the work, the MTA will add more entrances and elevators to the Bedford Ave. station and an accessible entrance to the 1st Ave. stop at Avenue A, thus opening up more of Alphabet City to the L train. It’s small carrot for what will be a year and a half of transit pain, but these are necessary improvements that should bolster commuters for all. The agency has unfortunately dismissed the idea of adding tail tracks at 8th Ave. I’ve written in the past as something that should be on the table. Here’s the MTA’s rationale:
The existing L subway line terminal at 8 Av, which allows for a maximum of 28 trains per hour, is currently not, nor is it projected to be, the capacity constraint on L subway line frequency. Long term ridership forecasts do not show the need for frequency of service beyond what the terminal can already accommodate. Extending the tail tracks, which are tracks just beyond the end lines that can be used for storing and turning around trains, would allow trains to enter the station at higher speeds, but the large cost of constructing such a project would not justify the relatively minor gains in passenger travel time.
As the hammer came down on Monday, the MTA announced some contingencies plans but largely punted on the rest. There are, after all, 29 months for planning before the L shuts down. What we know though is that the MTA will run full-length G trains while increasing service on the G, M, J and Z trains. The agency will provide free transfers between the G at Broadway and the J/M/Z at Lorimer St. and between the L and 3 at Junius/Livonia. As for other mitigation efforts — a bus bridge or increased ferry service — the MTA stated, “A range of additional bus and ferry services are being developed along different portions of the corridor. We plan to work closely with the City and State to develop routes and determine service levels needed to accommodate projected ridership.”
As MTA plans go, this one is well-thought-out and comprehensive with the agency indicating a need and willingness to work with the city. It earned praise from transit advocates and local politicians both for its transparency and community engagement efforts, but that didn’t stop the city from trying to get in a jab at the Governor’s transit agency. In a statement to The Times, Deputy Mayor Anthony Shorris threw some unnecessary barbs while ignoring months of outreach:
The L train carries more than 300,000 riders per day and is a vital transit artery for neighborhoods on both sides of the East River. While we recognize the need for the MTA to perform these important repairs and upgrades, we are deeply concerned that it would announce an 18-month shutdown of this critical service without a clear plan or a commitment of resources for mitigating the impact of this closure on hundreds of thousands of New Yorkers. Well before this shutdown occurs, New Yorkers deserve clarity from the MTA on how it intends to minimize inconvenience and keep people moving throughout the duration of the construction.
Mayor Bill de Blasio added his two cents later on in the day which included some unnecessary defense as part of his ongoing schoolyard feud with the governor:
First of all, I’ll remind everyone the MTA is run by the State of New York. The amount of time that they have projected — the 18 months — is a very big concern for me and for the City government. We’re going to have some very serious conversations about the MTA, about whether it has to take that long and how it’s going to be handled. I want to make sure there’s a lot of redundancy in place. By the time it happens, one — small but important factors — we’ll have the citywide ferry service in place, so that’ll be helpful, but we’re going to need a lot more than that, obviously. So I want to press the MTA to show us that 300,000 riders really will have good and consistent alternatives. And we’re certainly going to look at what we have to do in terms of the bridge as part of that. We’ll have an answer on that after those discussions with the MTA.
The reality is that it’s incumbent upon the mayor and his Department of Transportation to lead here. The MTA has signaled a very clear intent to work with the city to develop contingency plans, and even though a ferry from Williamsburg is likely one part of that plan, it’s clear that the MTA hopes to run constant buses over the Williamsburg Bridge, a move that requires support and authorization from the city. Additionally, numerous local officials have lined up in support of the 14th St. Peopleway, and for the MTA to realize this call for a transit-only street focusing on buses, bikes and pedestrians, DOT again would have to act. That’s firmly on the Mayor’s shoulders no matter who ultimately runs the MTA. It’s all politics to him though even when the commutes of hundreds of thousands are at stake.
So we have a plan, and we have a timeline. Now, we need the contingencies. It shouldn’t be hard for various agencies to come together on a plan, and ultimately, the city can see what dedicated transit space can do. The L train shutdown will be ugly and painful and stressful for numerous communities of various backgrounds, but maybe the city and state can make the most of a crisis. Cuomo might not want to own it, and de Blasio may want to politicize it today. But when all is said and done, New York City could just be better off for it by the time mid-2020 rolls around.
97 comments
I think that we should all start jumping the turnstiles. The MTA doesn’t deserve our money if they can’t provide good service.
18 months of closure will kill Williamsburg… it can’t be done any faster? It’s not like they’re building a whole new tunnel.
Damn the politicians and the unions… no one cares about us subway riders.
Might as well move to a city that cares about the people that live in it.
Let’s pick one day this year where we all just jump… if we’re all doing it than we can’t be stopped.
Who is with me?
Young agile underemployed male anarchists are with you.
Trollers gonna troll!
Williamsburg will be just fine. It was still there before the hipsters came, and now it will be even more hipster-y for them to bitch about all the trains they had to take to go see a Disco Biscuits show on the LES.
“it can’t be done any faster?”
No, it can’t. Are you an engineer? Then you have no basis to assess the amount of time it will take to do these things.
And there will be plenty of other ways to get to Williamsburg.
Respectfully, get bent.
To be fair, my time in engineering school suggests that it can easily be done faster. There’s no real reason that removing a bunch of concrete and cables and installing new ones should take 18 months. To convince yourself, watch this video and then imagine how long it would take MTA to do something similar:
http://en.rocketnews24.com/201.....-of-hours/
That looks to be in the open air and not in a tunnel thats less than 20 ft wide so in no way shape or form is it any sort of equivalent
When they remove the track and duct banks they are goign to have to be careful not to damage the tunnel walls. THere isn’t alot of space to work in and it all has to be done safely.
Fed up of the arm chair engineers who think things are easy.
I specifically said “imagine how long it would take MTA to do something similar” not “here is a similar project.” In your opinion, how bad does it have to get before members of the public, some with engineering degrees, are ‘allowed’ to ask questions?
“The existing L subway line terminal at 8 Av, which allows for a maximum of 28 trains per hour, is currently not, nor is it projected to be, the capacity constraint on L subway line frequency. Long term ridership forecasts do not show the need for frequency of service beyond what the terminal can already accommodate.”
Given that with tail tracks and with a second, interim terminal on the other end, and given that the TWU prevented the MTA from going with OPTO, the 40 tph for CBTC was thus not much of an investment, was it?
There is another argument for the tail tracks. Redundancy. Even if in normal times 28 tph would be enough to keep the riders merely crush loaded, what about the next Williamsburg Bridge shutdown?
And if that terminal is not the constraint, what is?
>And if that terminal is not the constraint, what is?
Substations. The current amount of power cannot handle any more trains. They will be building more as part of this shutdown, though, to increase capacity.
The BOT and TA gave 32 tph (536 ft trains) as the service level capacity in their 1949 and 1954 reports. They may have included short turning trains at 6th Ave with a wye that has since been removed. I find it difficult to believe that anything greater than 24 tph is possible at 8th Ave. I’ve made timing studies. Since that time, the MTA has removed faster acting electro-pneumatic switches with slower all electric ones.
There’s a problem with the TA’s service loading guidelines. They are based on floor space per passenger. The problem is that loading times are greatly increased on a per passenger basis at such loading levels. Only an estimated 80% of the TA’s service loading guidelines can be accommodated while simultaneously maintaining scheduled dwell times.
The TA currently operates 20 tph through the 14th St tunnels at 100% of service guidelines. That translates to 25 tph with operating at 80% of the service guidelines. That’s more than the 8th Ave terminal can handle, according to my measurements.
If the intermediate station service level capacity is 40 tph, it makes no sense to limit the line’s performance to the restriction imposed by the terminal. Customers expecting a seat during rush hour (or any other hour)? Oh the horror.
One of the major constraints on dwell time is the issue with doors closing. If the doors closed firmly but at moderate speed, were not holdable, and did not have door recycle (in other words like the doors on Eastern European metro trains), dwells would be reduced significantly. The doors would need to be open-able by about an inch without losing indication for people who get objects or clothes stuck. (Our own SMEE cars have this feature, so it’s not unprecedented).
Right now, people hold the doors because they know they can, and because they know it works.* Both of those should be fixed.
*Another contributing factor is the frequency we operate; if the L ran 40tph people wouldn’t mind waiting 90 seconds for the next train, as opposed to now where it’s hopefully 180 seconds away.
Your timing studies, last I asked you about them, were made prior to CBTC. Trains enter the terminal much faster now than before CBTC. Your timing studies are irrelevant.
(Your timing studies, as I recall, also fail to acknowledge that trains can enter and exit the station simultaneously as long as both trains are making straight moves.)
Somehow I suspect that the entry speedup with CBTC has a much greater impact, yet you opt to ignore it.
Estimated by whom? Where are you coming up with this?
I agree that dwell times are a problem. To a large degree they’re a problem because a significant minority of the ridership insists on standing in the doorways for stop after stop after stop. (The R211 order is getting wider doors – hopefully addressing this problem rather than simply encouraging even more people to block them.)
Your measurements are outdated and your calculations are wrong.
But that’s irrelevant in this case, since the power system can only accommodate 20 tph today and there’s funding to improve it only up to 22 tph.
The line’s performance is already constrained by power availability.
And I haven’t seen any evidence that the intermediate station service level capacity is 40 tph – but two-track lines are often constrained by their terminals.
Are you proposing that NYCT aim to schedule service for seated loads during rush hours? That would be simply infeasible on many lines, even with unlimited power and the most advanced signaling available. And even where it is feasible, it would require a massive increase in cars (and enlarged maintenance shops) and a massive increase in operating personnel. Who did you expect to pay for your proposal?
The L is a relatively short line, and the heavy crowding is concentrated in its inner section – while fairly intense, nobody has to endure it for terribly long. If you’re going to advocate for more seating capacity, wouldn’t you pick a longer line that carries standees for greater distances?
“Your timing studies, last I asked you about them, were made prior to CBTC. Trains enter the terminal much faster now than before CBTC. Your timing studies are irrelevant.”
It’s my understanding that trains are under manual control when entering and leaving the terminal stations. This would mean that any ATO improvements are not present at the terminals. However, I should get off my rocking chair and take newer measurements.
“(Your timing studies, as I recall, also fail to acknowledge that trains can enter and exit the station simultaneously as long as both trains are making straight moves.)”
Uniform time displacement between departing trains is essential for 90 second headway operation. Gaining a few extra terminal tph by using the uneven departure times that straight in-out slot provides is negated by intermediate station requirements. It’s a question of providing what the line requires and not what the terminal can provide.
“But that’s irrelevant in this case, since the power system can only accommodate 20 tph today and there’s funding to improve it only up to 22 tph.”
The power system is currently providing power for 20 tph in two directions simultaneously. That’s equivalent to operating 40 tph in a single direction. In terms of overall power requirements for the entire line, the schedule that I propose requires 18 train sets. That’s a decrease from the 24 now required for the am rush.
“And I haven’t seen any evidence that the intermediate station service level capacity is 40 tph – but two-track lines are often constrained by their terminals.”
The derivation for intermediate service level capacity is given in: “Urban Rail Transit, Its Economics and Technology,” Lang and Soberman, MIT Press, (c)1964, pp. 116-124.
The Moscow Metro currently operates some lines slightly in excess of 40 tph. The Third Ave El used to operate at 42 tph. What do a bunch of foreigners know or what did our grandfathers know?
Only an estimated 80% of the TA’s service loading guidelines can be accommodated while simultaneously maintaining scheduled dwell times.
Estimated by whom? Where are you coming up with this?”
I’ve been researching the entry/exit times for buses because of the current interest in newer fare collection systems. The research notes that at least 0.5 seconds should be added to the various times, if standees are present. Here’s a link to an article that quotes an 85% figure, on page 33.
http://www.nctr.usf.edu/wp-con.....Dueker.pdf
I’ve seen other articles that cite 80% and apply the results to both buses and trains with level platform loading.
“re you proposing that NYCT aim to schedule service for seated loads during rush hours? “
That remark was with tongue in cheek. Let’s look at the serious side.
Rush hour demand has significantly decreased in the past 30 years. The growth in ridership has been exclusively during the non-rush hours. The MTA has reduced rush hour service on practically all lines. Had they maintained their previous rush hour service levels, most rush hour passengers would get a seat.
The L is a relatively short line, and the heavy crowding is concentrated in its inner section – while fairly intense, nobody has to endure it for terribly long.
Loading patterns on the L vary by time of day. Bedford Ave is the busiest Brooklyn station but not during the am rush hour. You need to look at hourly turnstile counts to get a more accurate picture.
The Moscow Metro currently operates some lines slightly in excess of 40 tph. The Third Ave El used to operate at 42 tph. What do a bunch of foreigners know or what did our grandfathers know?
Just to clarify this point. Moscow runs just under 40tph. The Third Avenue El track that ran 42tph didn’t have signals. Trains ran on line-of-sight.
I’m not sure why you’re bringing up ATO. The faster terminal approach applies even when trains are manually operated in CBTC.
How did we get to 90 seconds? The claimed capacity (that you challenged) was 28 tph – that’s over 120 seconds.
Since I don’t know what constrains the power system, I have no way of knowing if your claim is correct. (I suspect you dno’t, either.)
Power requirements != train requirements.
You’re deriving the capacity of an early-2000’s signal system from a 1964 textbook?
And neither the Moscow Metro nor the Third Ave El has/had the signal system currently in place on the Canarsie line. And I suspect they operate(d) under somewhat different safety standards.
Pardon? 85% of what? 80% of what?
There is no fare collection on board subway trains, so fare collection has no impact on subway dwell times. The extent to which crowding affects dwell times depends highly on car layout. (Why do you think the R211’s will have wider doors than older cars?)
Really? Rush hour demand has significantly decreased from 1986 to 2016? How significantly?
Primarily in the non-rush hours, sure. Not exclusively.
I’ve seen no evidence that the ambitious service levels that you so often cite actually operated in any regular or reliable fashion. And, of course, you are aware of the modifications to operating rules and signal design in response to improved safety standards.
I never claimed that Bedford Avenue was the busiest station during the am rush hour. I simply said that the heavy crowding on the L is concentrated in its inner section and that the entire line is short to begin with.
“The existing L subway line terminal at 8 Av, which allows for a maximum of 28 trains per hour, is currently not, nor is it projected to be, the capacity constraint on L subway line frequency. Long term ridership forecasts do not show the need for frequency of service beyond what the terminal can already accommodate.”
Given that the L already handles 20 TPH in the peak, and the trains are overcrowded and the area continues to grow, I find it hard to believe that 28 TPH will be enough indefinitely.
As Larry asked, WHAT IS THE CAPACITY CONSTRAINT IF THE 8TH AVENUE TERMINAL ISN’T IT?!?!?!?!
At the moment, the (L) is held back by insufficient power; the substations on the line can’t support any more trains. The MTA is looking to build three new substations along the line to rectify this. Also can you not type in all caps, it’s really not necessary.
Then aren’t they being disingenuous? If the actual constraint will become a non-issue, then the issue of the terminal being the constraint is still valid! And thus the option of building tail tracks should not be dismissed so easily.
I typed in all caps for a reason. I was shouting out of exasperation. I was asking a question that never should have needed to be asked, for various reasons. You’re supposed to use caps to show that you are shouting. Otherwise how would anyone know you are shouting?
The current capacity of the L is 20tph. Adding the new substations will increase that to 22tph. The capacity of the 8th Av terminal is reported to be 28tph.
…International best practices allow 40tph on a two track line.
Ok. Riverduckexpress seemed to be implying that the three new substations would make it such that the available power was no longer a capacity constraint.
Nonsense. Very few systems worldwide are capable of 40 tph operation in practice.
In London, most lines run at 30-32 tph and they are gradually upgrading them to 36 tph.
In Moscow, 40 tph is typical and they are planning for 44 tph on new lines. I wonder if they have lower safety standards than us.
Here is an interesting worldwide summary, a decade out of date though:
http://districtdavesforum.co.u.....ro-systems
The highest-frequency lines in the summary (“VAL”) are rubber-tired, BTW.
“Very few systems worldwide are capable of 40 tph operation in practice.”
Most systems are capable of 40 tph, including NYC. They just don’t operate at that level.
It’s fairly simple to measure intermediate station capability in NYC. All it takes is a stopwatch and patience during off-peak hours. Measure the time it takes for an entering train to stop from the station entrance. Next measure the time from when a train departs until the signal just before or at the station entrance turns green. Add a nominal 30 seconds for dwell time and the sum will be the minimum headway for the system.
Take several measurements at each station. There is great variability among train operators. There was one operator who refused to enter Grand Central despite a green signal. He managed to screw up the morning rush hour several days in a row. If one operator can operate at the rated service braking rate of 3 mph/sec, there should be no reason why all can’t.
The greatest documented intermediate service level I’ve read about was on the Brooklyn Bridge shortly after it was converted from cable operation. It was 66 tph in each direction. The limitation was the weight restriction on the Bridge. They had a trick. Each station, Sands St and Park Row, had two tracks in each direction. Trains would alternate between the two tracks and thus not be affected by dwell time.
The need to operate at that service level was short lived. The IRT’s 1908 extension into Brooklyn effectively killed demand for the BRT’s service levels.
As mentioned, the BOT noted 42 tph in the opposite direction for the Third Ave El in their 1949 report.
I’ve seen reports that service levels on the Brooklyn Bridge were 90 tph in each direction during the cable days. The cable provided for a uniformity of speed control that CBTC/ATO has yet to match. Our great-grandfathers weren’t dummies.
Interesting that the proposed service map doesn’t show any bus that actually travels from Bedford to 14th Street. A trip from Bedford to 1st Ave goes from the current 5 minute trip under the river to… half an hour? M15 headways will have to be improved. This will impact more than what’s shown on that map.
Agreed. Hopefully this will happen in their continued talks between the MTA and the DOT. In a dream scenario, make both couples of outer lanes bus-only and have a dedicated line that goes directly from Bedford to the L stops on 14th. Make the right turn onto Clinton St, and then Clinton St and Ave B bus-only from 7-7 M-F (also Ave A and Essex St bus-only for the return trip). You could call it the WB14 bus route for shits and giggles. I know this probably isn’t realistic but if everyone came together and was level-headed, it would be the best alternative.
The intra-Manhattan and interboro ridership flows should be separated, if possible. Few people are going between 1st Av and Brooklyn, and it’s faster by road to turn from the bridge to Bowery which becomes 3rd Av. From there, IMO, the bus should turn down 14th to either Union Sq or all the way to 8th Av.
There will always be complainers about something ( certain people are never happy). That said, two issues that have been discussed over and over again, have been the free transfer at Livonia between the L and 3, and at Broadway between the G and the L and J. I have done the G Transfer many times with the unlimited Metro Card and it is very useful. Maybe it takes five years from today, but in conjunction with F Express, transit should be much better in Brooklyn. That is a good thing, no matter what the complainers think.
It will be interesting to see how much additional J/M/Z service the MTA can crib from borrowing railcar sets from the L line, since it’s still going to need a certain amount to get people to the Broadway Junction, Myrtle-Wyckoff and Metropolitan-Lorimer transfer points (swapping out more cars with the C train to put additional 8-car sets on the Eastern Division is also an option, and the R-179 arrivals by 2019 should help the railcar situation … as long as the MTA doesn’t decide to start retiring some of the R-32/R-42 sets the moment the new cars come in.)
The M’s apparently limited to a max of five more TPHs due to the constrains of sharing track with the F on Sixth Avenue and with the E on 53rd, while the J/Z combo with the M across the Williamsburgh is going to be less than those combos because the bridge is speed-restricted.
Yes, the plan is to retain some, if not all, of the R32 fleet for the duration of this service change in order to have enough equipment on hand to operate full length G trains and additional M service.
Although… where are the R179s?
Bombardier is the hold up. The MTA had a release a few months back of how they had to extend the life of R32s due to the late delivery of the R179s by Bombardier. The test train is not even here yet, and the order was supposed to start coming in this year!!
Is it really necessary for buses to go over the bridge? The main win there is getting additional people to the F train at Delancey, but those are already pretty packed at rush hour, right?
That is indeed an issue. Ideally the buses would go to Union Square, or at least to a place NOT served by the M.
No. http://newyorkyimby.com/wp-con.....owding.jpg
It should be no surprise the shutdown everything option was chosen. The MTA skewed its presentation to favor this option. It also minimized the inconvenience caused by the only two options it would consider. It also completely ignored the cost to the public in its calculations to come up with a “cost effective” solution.
It’s irresponsible to consider an option without simultaneously including the transit alternatives for any service disruptions. They need to be considered together, as a package.
The MTA would have us believe there would be minimal impact to L-Train service in Brooklyn. That’s not true. Service levels are reduced to 8 minute headways (7.5 tph) from 20 tph (3 minute headways). They will be providing only 37.5% of peak service, which they claim to be “near normal service” for Brooklyn only L-Train passengers. The one-tube option offered 12-15 minute headways (4-5 tph) or 20% of peak service levels in Manhattan. The MTA characterized 20% service as unreliable but 37.5% service as “near normal.”
Approximately 400,000 people use the L-Train on an average day. If each of them had his/her journey lengthened by 30 minutes, it would come to 25 million hours per year. At $15/hr minimum wage, that’s a $750 million yearly project cost that the MTA has pushed onto the riding public. If the service interruption lasts 18 months, that’s $1.125 billion the public will absorb. The MTA’s reconstruction cost is estimated at only $800 million. Excluding losses from businesses located along the L-Train, the project cost is $1.925 billion. The MTA’s tab is only 42% of the total and far less if business losses are included.
Excluding collateral damage from the cost calculations biases the option choice to those that maximize such damage. One option would have been to build a third tube. This would have eliminated all collateral damage. It was never seriously considered. It might have been the least costly option, had all the costs been considered.
Indeed it’s difficult to divine any intention on the MTA’s part to consider what the public will do during the shutdown. Their decision process has been to first decide on a reconstruction option and consider its consequences later. That’s planning at its worst.
It can’t wait 25 years. The amount of time it would take to get clearances and build it. Not that it would get past a Major Investment Study.
Building a third tube is one of many options not considered by the MTA that would reduce the impact on subway users. It’s the zero impact option. Its environmental impact is minimal because construction is entirely underground.
Their stated reason why weekend and night time closure wasn’t pursued is the presence of silica dust. They considered only reconstruction methods that worked for small tunnel segments.
There are many industrial processes where far more dangerous airborne particles result. They don’t wait days for the particles to settle out before proceeding. They use positive ventilation systems that direct the air through wet collectors or similar particle removal devices. The setup is expensive but stopping production is even more expensive. The setup doesn’t scale down, nor is necessary, for the small concrete removal jobs the MTA has done to date. However, the concrete process that the MTA has proposed doesn’t scale up. Their solution has been to stiff the users.
Another alternative for concrete removal is intense, concentrated heat. It turns concrete into glass. The glass doesn’t present a silica dust problem. We’re talking lasers or plasma torches for the heat source. It’s expensive machinery that the MTA’s approved contractor list does not use. Laser and plasma torches also don’t scale down for the small jobs the MTA has done to date.
The MTA isn’t quite using hand tools like picks and hammers. They have graduated to jack hammers. They have not considered anything more modern. They did not have to because they considered only their immediate cost and not the project’s total cost.
you can’t wave your hands and say there will be little environmental impact. There have to be studies. And comment periods. And lawsuits. And more studies. Then the years it would take to build. assuming the studies don’t come up with “it costs too much money”
Trump said that, if elected, he was going to immediately build a third tube on each and every one of the East River crossings – and he was going to make Latvia pay for them.
Yes, construction is entirely underground. Except for the massive TBM launch box that will need to be built. Even if we skipped the process of an environmental study, I don’t think there is anyone who thinks this thing could be designed, constructed and commissioned in the next 3 years. And then what would we do with it once it’s complete?
Does anyone make a silica dust collection system large enough to handle the scope of the project? What would be the cost implications of setting up such a system to remove it 48 hours later, every weekend?
Plasma torches are certainly used on MTA projects; on steel elevated rehab projects if rivets cannot be forced out with a pneumatic hammer, they are supposed to be removed with a plasma torch. In this case, I don’t see how a plasma torch would help remove the ductbanks in question, and I could see them being excluded for the same reason that use of a Brokk is also excluded from this work.
You also made a time calculation about time lost by those who have short rides already. It’s good for you to take note of the following though:
1. Many people who ride through this tunnel don’t live directly above the L line, and so for them a walk/bus ride to a nearby train line will produce a smaller addition to their journey than 30 minutes.
2. If we followed the weekend logic you presented, weekend riders would face the same loss of time at a similar overall total amount.
I know you have your personal vendetta against the MTA, but sometimes all of the options are not great. Residents of Williamsburg will, probably to your surprise, survive this.
“Plasma torches are certainly used on MTA projects; on steel elevated rehab projects if rivets cannot be forced out with a pneumatic hammer, they are supposed to be removed with a plasma torch. In this case, I don’t see how a plasma torch would help remove the ductbanks in question, and I could see them being excluded for the same reason that use of a Brokk is also excluded from this work.”
I spent a decade of my professional career designing control systems for high power plasma torches. There are not that many manufacturers and there were even fewer customers. I knew almost all of them; it is a niche industry.
The MTA was definitely not a customer, nor were any construction contractors. Most customers were associated with jet engine manufacture or maintenance. The plasma torch’s high heat in a small area (hotter than the surface of the sun within the size of the last joint on your pinky) was used principally to melt exotic materials to be applied to metal parts.
Another application was to decommission nuclear plants. The concrete was turned into glass under the intense heat. The glass was more stable for storage for the millennia it would take the radiation to subside.
My guess is that were one of these torches applied to the steel of an elevated structure, it would cause the steel to catch on fire.
Plasma cutters most certainly have been used in the construction industry and on TA projects.
None of your comment addresses the issues that all of your proposals raise.
It does seem strange that they could only manage 12-15 minute frequencies on a single track. The distance between the stations adjacent to the East River is only 1.6miles. With new cross-overs at the stations, it should be possible to cross that in 4 minutes — and allow 10 minute frequencies.
Plus, with a local/express pattern before and after the tunnel, one could maybe have two trains pile together before going through the tunnel, so they could go together through the single track section, effectively allowing two trains every 10-12 minutes per direction. The CBTC signalling system should allow service pattern shenanigans like that.
Plus, the three-minute rush only lasts a bit more than an hour. Have they considered using a single track in the rush-hour direction only during the rush hour, essentially piling a bunch of trains downtown during rush hour? It may have to involve building those tail tracks after all, and shutting down the stations in Manhattans far side for storage.
The MTA is not considering any innovative solutions, when considering single track operation. They have even ruled out operating trains between Atlantic Avenue and Rockaway Parkway over the Williamsburg Bridge. That’s a service they used to and can still operate.
The 12 to 15 minute headway figure is based on the worst possible scenario. A single train in each direction. Running time between Union Sq and Lorimer (where the switches are located) is 7:30 each direction. That comes to 15 minute headways.
Many people have suggested an alternative scenario, where many trains operate in one direction before the direction is changed. If the trains are operated more frequently than the current 3 minute headway, increase the average throughput.
My own variation is to operate 6 trains in each direction at a time. On the Manhattan side the trains would terminate at 8th Ave, 6th Ave and Union Sq. These are the 3 stations west of the Union Square switch; each has two tracks. Hence the choice of 6 trains before reversing direction. These trains would operate at 40 tph (90 second headways), the theoretical limit for intermediate stations. The 6 trains would take 6*1.5 or 7.5 minutes to clear a particular point. Add in the 7.5 minutes to traverse the single track section and that’s 6 trains every 15 minutes. Two such 15 minute slots per hour and it averages out to 12 tph or 5 minute headways. That’s 60% of current peak operation.
There are two variations on the Canarsie side. Two trains would terminate at the two-track Rockaway Parkway terminal. The other 4 trains would terminate at the yard adjacent to the Rockaway Park Station. In one variation, the MTA would build two temporary wooden platforms in the yard for the trains. N.B. elevated platforms used to be wooden. In the other variation, the trains would terminate at E 105th St and proceed to the same yard minus the platforms.
As I noted, the MTA did not want a single track option. They skewed their options to force the total shutdown option. They made the single track option as unpalatable as possible.
Your proposal requires a kind of precise schedule keeping and efficiency that I doubt could be achieved here. Although with CBTC in place, it might come close.
I doubt it, given the insane levels of crowding his proposal would see.
At current headways, peak hour trains are filled to 97% of guideline levels (as of 2014), on average. So at Stephen’s proposed 60% of full peak operation, trains would be filled to 162% of guideline levels if everybody tried to board. And since service would, by necessity, be unevenly spaced, some trains would be significantly more crowded than that average.
Operating at theoretical limits assumes an absolutely perfect operation. If anybody so much as holds a door for a few seconds, the operation falls apart. Do you think that somebody maybe might hold a door when encountering this degree of crowding?
(For comparison, the Lexington Avenue express is at 104% of guideline. These trains, under Stephen’s proposal, would be over 50% more crowded than rush hour Lexington Avenue expresses.)
NYCT’s one-track proposal suppressed demand on the tightly constrained segment by splitting the line between Bedford and Lorimer. Without that suppression, I doubt an 80% or even 40% capacity reduction would have been entertained.
At some point it no longer makes sense to encourage people to use a service that can’t possibly accommodate anywhere near all of them and can’t possibly be reliable for those who are accommodated and instead focus on bringing them to multiple alternatives.
These proposals would result in less quality of service. If the frequency that the users perceives goes from every 3 minutes to every 12 minutes, many people may find alternative routes based on the lesser frequency. Ridership is directly related to frequency, so assuming the same number of people would try to take it is quite an oversimplification.
There are also direct ways to reduce ridership and only having people that really need the service (besides the reduction of frequency), like shutting down some station that offers parallel service, or charging different fares.
Operating trains at 90 second headways requires more precision than current practices – operating them at random.
CBTC provides the ability to operate trains at track capacity in either direction over the single track section. That’s CBTC’s only contribution.
Two other modules that come with CBTC are important. Train supervision is what’s responsible for keeping trains on schedule, as opposed to the CBTC component which is responsible for preventing trains from colliding with one another. The Automatic Train Supervision (ATS) module specified by the MTA for the Canarsie line mimics the manual train supervision system of the manual days. This means that the new Canarsie system could never perform any better than what it replaced. 90 second headways requires that schedules be written and operated to the nearest second. This is what Paris and Moscow do with manual train supervision that operates at or near 90 second headways.
The other component is automatic train operation (ATO). Uniformity in train operation is important to prevent trains from getting ahead or behind schedule. ATO eliminates operator variability. Eliminating the conductors on the Canarsie line during the emergency would eliminate another important source of schedule deviation. I don’t mind having them on the train. Just keep them away from the door controls.
Some form of platform entry control would be desirable. This would serve the same function that the platform entry gates used to perform in Paris. This could probably be linked to the turnstiles. Passengers would not be permitted to go through a turnstile, if a certain number of passengers had passed through them since the last train’s arrival. An equally effective algorithm would be to lock the turnstiles at a fixed time before the train arrived at the station and release them immediately after the train left. That was the algorithm used by the old portillons automatiques on the Paris Metro.
Yes, the MTA did very much skew their presentation to encourage the two-track closure, which I would favor if it weren’t for the fatal flaws of neighborhood disruption and the simple lack of alternate capacity. I very much like your out-of-the-box thinking and proposal. One thing I would suggest to speed service even more is to have trains go nonstop from Lorimer (*) to their respective Manhattan terminals, to avoid dwell time issues at the intermediate stops. However, one issue I can imagine cropping up is crowd handling – two jampacked trains arriving within minutes of each other, entirely unloading, then reloading and leaving in a timely fashion would be difficult to organize the people movements for.
*There isn’t enough room between Lorimer and the crossover to line up six trains while they wait for their slot; as such, trains heading to 6th Av and Union Square might have to bypass Lorimer as well as Bedford, 1st, and 3rd. Would that be an issue? I’m not sure.
To add on to my previous comment, the reason why 6th Av and Union Sq trains would have to skip Lorimer is to make the six slotted trains as close together as possible. If trains load/unload at Lorimer, that introduces a dwell time constraint, which will cap service at 40tph in the ideal case. If 6th Av and Union Sq trains didn’t have to stop at Lorimer, the six trains could be lined up between the crossover and Graham while they wait for their slot, and when the last train cleared they would proceed into the single track section one after the other. Same thing in the other direction – for the last train to clear the switch as quickly as possible most of the trains ahead will have to skip Lorimer.
Lining up trains to speed them through the single track section is not the way to provide 90 second headways. The 90 second headways have to be maintained throughout the entire line. Operational uniformity is required to achieve this. Let’s assume a train departs Canarsie and its follower leaves exactly 90 seconds later. If the follower maintains the exact speed profile as its leader, then the follower will always be exactly 90 seconds behind the leader at all points along the route. The physical distances between trains will vary.
…and then you will end up with Six trains, operating at 90 second headways, and then a gap in service of 7.5 minutes. How do you think dwell times will work on that lead train? Never mind the fact that destination stations in your proposal are actually only served by 2 trains in a 15 minute period. This proposal works in a theoretical world, certainly not in the real one.
I dunno about 6 trains, but a bit fewer trains could be using skip-stop patterns before and after the tunnel so that they bunch up inside the tunnel without bunching up before and after.
They could use A,B,C patterns, and have three trains go through the tunnel at the same time.
But again, why not create more crossovers in the tunnel to make the single-track stretch shorter.
And why not consider a short time where the single track becomes uni-directional during rush hour.
Aside from the dwell time issues caused by having wildly uneven headways, the time between trains heading through the tunnel is ‘wasted,’ since the constraint on throughput is the westernmost train of the bunch. However, the westernmost can’t start its eastbound journey until the easternmost train does, so it’s best to have the easternmost train get to its Manhattan terminal as soon as possible (in other words, as close behind its leader as possible). Therefore, I believe that stacking trains is the optimal solution here.
In practice, not all of these 400,000 riders are going to seek alternate routes. Many of them are going to simply not make the trip.
People have 2 1/2 years warning, that is plenty of time to plan to not be able to cross the river at 14th St come 2019. People living in Williamsburg or Bushwick who work in Manhattan will move or find a new job. People living in Williamsburg or Bushwick who are dating someone in Manhattan will move, have their SO move, or break up.
Don’t be surprised if ridership numbers start noticeably dropping months before the shutdown begins as the more proactive of the affected people alter their routines so as to not need that part of the L train. And don’t be surprised if, after the tubes reopen, it takes a couple years before ridership numbers recover to pre-shutdown levels.
The MTA has had a vendetta against Williamsburg North for some time. The neighborhood’s resurgence prevented the MTA from abandoning the Canarsie Line altogether. The MTA has not forgiven them. Not providing adequate alternative service during the tunnel reconstruction is the MTA’s revenge.
Ok, let’s see if I understand this correctly…
Over 25 years ago, there may or may not have been a proposal to eliminate the L train. Since that time, senior leadership of the MTA has experienced turnover multiple times and the neighborhood has gone through a complete rebirth (well AFTER the alleged abandonment scheme). A storm flooded the tunnel and damaged critical equipment. The MTA now wants to, not only rebuild the damaged equipment, but also upgrade Power Distrubution and access at two stations, and to do so, they have proposed shutting down the line to accommodate this work and this is all part of some elaborate revenge scheme?
I think I understand a lot more about your comments now.
I suppose I should indicate a tongue-in-cheek observation.
Senior management plays a game of musical chairs among the different transit agencies around the North America. One year they are in Chicago, the next in NYC. They play this “out and up” game so that they enjoy a couple of tenures at each agency. Thus, despite rapid turnover, today’s managers are familiar with what transpired decades ago. They were present; it was many steps ago on their career ladder.
When they finally wear out their welcome at all the transit agencies, they retire with a pension. They then become consultants for one of the consulting firms they brought in to turn their agencies around.
Senior management wants to spend money on expensive projects. It pads their resumes for their next career step. Whether or not the expensive project is necessary or even useful is immaterial. Their salary is based on the dollar value of the projects they manage.
It’s operating trains that they find tedious.
My cynicism is based on more than half a century of observing the NYCT follies.
Even if you’re making a valid point about the revolving door of transit management in the US (and world at large), the idea that this 18-month L train shutdown is somehow a vendetta against Williamsburg is laughably insane. Maybe there was a plan around 40 years ago to shrink the city, but there’s no indication in any of the MTA’s current records that shuttering the entire L train was on the table. Meanwhile, nearly everyone involved with the MTA from the late 1970s is dead or close to it. You should drop this thread; it’s not helping you make your otherwise interesting and valid points.
Okay, so now that argument is that middle level management really, really wanted to shut down Canarsie, but failed, and so now that they’re top dogs, they are taking their vengeance against a new group of northern Williamsburgers….
BTW,
The people who run the trains?The people who build the trains
Your comment about “them” not wanting to run trains is absurd and wrong.
Should say:
The people who run the trains != The people who build the trains
No. The MTA characterized a particular single-track operation as unreliable because it would have been unreliable. With a few people holding doors on crowded trains at the busy 1st Avenue station, the 12 tph service becomes unachievable (leading to even more severe crowding). A single-track operation like this has a very low tolerance for irregularity.
In contrast, service between Canarsie and Bedford can operate quite reliably on an 8-minute headway. Since L trains won’t be operating into Manhattan, they won’t be carrying Manhattan-bound riders all at once into Manhattan – instead, they’ll be distributing Manhattan-bound riders to other lines at four transfer points in Brooklyn. A current Canarsie-to-Manhattan rider who takes the L the whole way will instead get off at Livonia for the 3 or Broadway Junction for the A; perhaps his space will be filled by somebody else at Wilson, who then gets off at Myrtle for the M; then her space gets filled by someone at, say, Grand, who then gets off at Lorimer for the G.
The average L rider won’t have his/her journey lengthened by anywhere close to 30 minutes – especially not when you include the 175,000 L riders who don’t ride the L across the East River. Even including only the 225,000 riders who cross borough lines, I doubt the average impact would be as much as 15 minutes.
So the conclusions you draw are irrelevant.
There’s a fairly easy way to determine whether your analysis or mine is closer to the actual situation. This isn’t the Normandy invasion, we can run full scale tests ahead of time and see what happens. We can have multiple chances to get it right.
Take a weekday to a week and don’t run Canarsie trains between Bedford and 8th Aves and see what happens. Give riders ample warning, implement the many service changes on other lines and bus lines that are supposed to compensate for the loss of service. Let’s see what happens.
If all goes well – great. If not, it’s back to the old drawing board, as we in engineering used to say.
I made a dry run suggestion at the first community presentation in the armory in Williamsburg. The MTA dismissed it out of hand.
Nice try.
But the plan relies on multiple mitigations that are not in place yet. There’s a proposed ferry line serving a dock that doesn’t exist, connecting to multiple bus lines, running with buses that likely don’t yet exist and will likely use bus lanes that don’t yet exist. At least one station (Court Square; see the recent New York Times article) will be physically modified with additional staircases before the shutdown begins. The increase in G service is, I assume, contingent on subway cars that don’t yet exist. So, no, the proposed service plan can’t be implemented today, even if there were a desire to do so.
Even if it were possible, it wouldn’t be very useful. Most riders will have multiple feasible options, and it will take them quite a bit of trial and error to settle on the ones that make the most sense. A one-week trial wouldn’t be very interesting because riders would still be in the middle of the trial-and-error period when it ended.
And, of course, you’re proposing to severely disrupt a lot of people’s travel – completely unnecessarily.
Meanwhile, you didn’t address any of my points.
“free transfers between the G at Broadway and the J/M/Z at Lorimer St. ”
this should start now and continue forever
Yes, and it should be set in concrete with escalators or elevators.
I wonder if the MTA might just run all F trains express during the shutdown to make those south of Church happy, since there will be more G trains, under a “share the pain” policy. And then just keep it that way.
I doubt it. Park Slope will howl over their lost one-seat ride to Manhattan, just as they did the first time around. Some Fs will probably run express though – we can’t expect the IND Culver to handle 30tph on a single track right?
According to the MTA, more than 7000 feet of tunnel was cleaned and re-lined in the Montague tubes. From this article, we need to do 7100 feet of re-lining in two tunnels, just like Montague.
That repair, which did not involve station upgrades, took 13 months. How is this repair different? I note also that the Montague repair took a month less than originally scheduled, and was under budget by quite a bit. What’s the best that the MTA figures 15 months for the repair, reports 18 months, as so is the recipient of grateful thanks when they finish in their original schedule?
What’s the bet? Stupid autocorrect.
Typically modern infrastructure construction/reconstruction project planning includes schedule contingency as a best practice. Large projects generally budget 20% of the total project timeline to schedule contingency. In this case the stated project duration of 18 months would include the unallocated contingency of 3 months (20%). Therefore the actual project duration could possibly, if all site conditions are found to be exactly as anticipated in the preliminary engineering documents, be 15 months. 15 months would also preclude any human error, procurement problems, and acts of God/any other unforeseen event that could impact work. This is seldom the case.
Let’s hope contingency was budgeted more intelligently than was done for East Side Access. It took until the 4th enterprise project plan re-assessment to include an adequate schedule contingency of 3 years, which is being drawn down regularly.
What’s the possibility of direct Canarsie-6th Ave service? Can the Willy B handle the traffic?
Most of the people complaining are Williamsburg folks who are used to their 10 minute commute to Manhattan. Canarsie people have other options, so I don’t see the need for a new service.
Less ‘New Service’ more ‘return of old service’; but also allows direct service to Manhattan from Canarsie without a transfer, especially if people are going to destinations near 6th Ave, or Nassau Street.
This is only tangentially related, but has there been any talk of reopening the Grand Street entrance at Metropolitan as part of this plan? That would shave off several minutes of walking each day for people coming from the south, on a line that will certainly see an uptick in ridership during the shutdown. I caught a glimpse of this closed mezzanine back in March and it seems to be as clean and in good repair as most that are still in service. (There’s a post on this blog with pictures of this Mezzanine from a few years ago. It still looks the same today.)
At least some people in MTA believe that reopening closed entrances will trigger ADA. I’m not a lawyer so take my words with a grain of salt, but I honestly don’t think ADA will be triggered now if it wasn’t during all the ground-up station rebuilds over the last few years.
Realign the J/M/Z stations to have one top of the G, with a free in-system transfer.
Now.
Thankfully, there’s already an in-system transfer from the G to the 7.
I was thinking the same. I would reduce station dwell time and train run time by condensing the amount of stations in Willamsburg (not by alot, but still significant enough) and allows an easy way to create a free transfer.
I disagree with the characterization of Shorris’s comments as unnecessary. NYC Transit inevitably, and understandably, wants to conduct major rehabilitation projects at minimum cost, particularly with respect to supplemental service. Why run shuttle buses, that may reduce delays faced by affected commuters when you can just direct them to backtrack on the line and transfer or take an existing bus line (and burn a transfer)? The Riders Alliance is pushing for more supplemental service. I think it is perfectly for the deputy mayor of an MTA constituent government to make a similar request, especially since there is only one shuttle bus proposed.
You all should check out the proposal by ReThinkNYC. They seem to feel this will fix all of the problems, for cheap…
The Fulton Street line manages to fill up 26tph worth of trains from people boarding east of Hoyt. Cramming them (and some L riders!) onto 18tph so that you can have 8tph of E service switch over from the Crosstown line isn’t viable. Before anyone asks, yes, capacity of Cranberry tunnel is rather low.
The issue isn’t the tunnel. It is the switches before and after the tunnels that limit the capacity. This proposal is probably the same switches that are at Hoyt-Schermerhorn Sts right now. If the switches could be lengthened, you could increase the frequency to 30tph, therefore cramming Fulton and the Rockaways onto 22tph with the E using 8tph. Bu that would mean changing switches at Canal and at Hoyt-Schermerhorn.
I have always been surprised this connection wasn’t built, but agreed it doesn’t solve this issue because there is no track capacity between Hoyt and Canal, due to switching…which you can’t fix by adding more switches.
Plus, if you really need the E, what’s wrong with the no-build option of just using the existing connection at Court Square, which is faster to Midtown for everyone east of probably Clinton-Washington. And if you need downtown, you’re already accommodated by existing A/C transfer at Hoyt.
They still should update those switches though, trains must go painfully slow and posted capacity aside, there are frequently AM backups between Lafayette and Hoyt waiting to change tracks,
Agree, well-timed cross-platform transfers between the A C and G basically eliminate the need for a physical connection.
Well-timed? At 26 tph in the morning rush, there should be an A or C train coming along every 2-2.5 minutes. What sort of timing did you have in mind?
Build a floating bridge. Run the shuttle buses over it.
I’m serious. I-90 in Seattle is carried by a pair of floating bridges. (https://en.wikipedia.org/wiki/Interstate_90_floating_bridges) I’m aware that those are over a lake and this is a tidal strait with stronger currents, but there’s no way that is an insurmountable obstacle.
and close the entire East River to any boat traffic? It is still a working river.
The floating bridge can be high enough to clear most boats. Any extra-height special moves would probably have to go around LI, but there are not many overly-tall boats that use the East River AIUI.
Except for the multi year environmental review process and the lawsuits at every step. They should be able to begin construction in 2030!
They need to build the south. 4th Williamsburg is exploding at the seams
They need to start the expansion of the system for the the IND Second System that was proposed decades ago. It would have provided the necessary capacity to shut the L a long time ago instead of waiting until 2019.
Electric bike. Beats the hell out of that bridge climb.
Well…why close it on 14th street?
Should at the bare minimum still operate as a 14th Street Shuttle…no?
Lots of people use the L as a shuttle train, so it seems silly not to operate a shuttle between 1st and 8th ave…
Um, because there’s no way to service the trains if they can’t get to Brooklyn. There’s no yard or shops in Manhattan.
(I should say: There’s no yard or shops in Manhattan that the 14th Street line can access.)