Amtrak Gateway Tunnels

[east point]

One tunnel bore fails and is shut down. Now we have complete dependence on the other bore. Right now Amtrak is alternately closing one bore each weekend for necessary maintenance. So that then means each bore is shut down every two weeks. However Amtrak is able to stretch that to every 3 - 4 weeks during holidays and high traffic times.
Now with just one bore operational what happens when a necessary shut down to maintain the unreliable status happens? Now you have a shutdown of all traffic including Amtrak Thru NYP. Then what happens if the repairs are not completed by the Monday morning rush?

[rr503]

Before closing one tube 24/7, you'd clear the maintenance backlog to prevent dual closures. This isn't unheard of; in NYC even, NYCT has done/does it.

The unreliability issue needs to be discounted to some extent. It isn't like having Gateway was going to make it so that disruptions became service neutral. This absolutely increases risk, but I'm unsure that this risk warrants massive, unreformed investment.

[Defiant]

Current throughput is 24tph, which is a train every 2.5 mins. Giving a (possibly overly) generous 10 minutes of buffer time between fleets, and dedicating 35 of the remaining 50 mins to peak direction traffic while the other gets 15 gives you 14 tph of peak capacity. That's a cut of 10 peak direction tph. This sort of fleeting is done every weekend, and is a common single tracking technique elsewhere.

What to do about the lost capacity, though?

Currently, no NEC/NJCL trains serve Hoboken whereas a good percentage of their traffic is Hoboken-bound. Building out the other half of the Watefront Connection would probably allow you to sap 5-6tph from those lines' combined PSNY throughput, meaning you're left with 5tph. A combination of Hoboken reroutes (coupled with more PATH service and lower NJT fares), Amtrak short turns, and peak service spreading (ie more trains earlier and later) would be able to preserve capacity. You could also throw in a second XBL, as remember, the current XBL carries more people than the train tunnels.

I am sorry, are you a daily NJT commuter to either Hoboken or Manhattan? Have you seen what happened in Hoboken two years ago when the emergency work on NYP forced some NYP bound trains to be diverted to Hoboken? The crazy overcrowding, packed PATH trains. These conditions are just not very safe long term, especially in winter. Hoboken was not build for this volume of people. Even if you built six more platform, there is just not enough space for everyone to safely get to the ferry or PATH. Hoboken to 33rd street PATH itself often experiences problems and does not have a lot of spare capacity. I have heard that the switch leading to Hobken needs to be replaced. Overall, pretending that there is no problem in diverting a huge number of daily commuters to Hoboken for 2-4 years is really disingenuous and cynical...

I, too, think that we need new tunnels. But to advocate for them as an immediate, inflexible need (thus allowing you to bypass some very, very necessary cost and operations reform) is bad for the region.

I disagree. Having horrible commuting disruptions, unsafe overcrowding and eventual job loss will be really bad for the region. Cost reforms can take decades and may not produce much. Reforming how the whole country pays for the transportation projects is not easy and may not happen. Holding the region hostage to those reforms that no one committed to is irresponsible. The tunnels need to be build no matter what though of course we don't want to waste public money. But I agree, the costs of building anything in this country are insane...

[rr503]

Not a daily commuter but a frequent user. I'm fully cognizant of the issues you note.

PATH is complicated. The platforms are packed, yes, but there are ways to distribute loads -- extending 33rd service to Newark, running more WTC trains (historically, that branch handled >>30tph), hell running more 33rd trains (IINM you can eeke out another 3-4tph before you run up against terminal capacity limits). Then you can add high capacity ferries and buses to deal with the rest. On the NJT platform side, simply managing terminal operations better (making sure departing crews are ready to go, or shortening crew turn times) should help speed things up, and I'm sure there are some pedestrian flow improvements to be had. Remember, too, that we're not actually talking about that many trains...

Is this perfect? No. But spending 14 billion on a project that would cost 3-4 billion elsewhere is frankly ridiculous. Cost reform efforts are only as effective as the people pushing them; if the political strum und drang was directed in that direction rather than to buildatanycost-isms, we could probably have Gateway and a whole bunch of other eminently necessary improvements. Holding Gateway hostage for the sake of reform may sound cruel, but think of all the things that would be enabled by a rethought regional cost structure. I think it's worth it.

Addendum:

There's also a more cynical argument to be made here: with lower local costs, there's less reliance on (fickle) federal funding to get things done, which means more certainty, fewer moving parts, etc.

[east point]

The quickest way to tur trains would be for another crew ready to board outbound cab car or loco as soon as the trains stops inbound. That way the initial PTC initiation would be complete as soon as passengers were all off and any out bound loaded. However the low platforms would need changing to high level platforms. Actually as now set up with inbound crew taking outbound same equipment ( most times) turns are going to take longer for the PTC set up. It takes time to clear train and engineer to walk to outbound affront against detraining passengers.

Also MJT is short crews and will be for 3 - 5 years from what one person told us?

[Hawaiitiki]

PATH is complicated. The platforms are packed, yes, but there are ways to distribute loads -- extending 33rd service to Newark, running more WTC trains (historically, that branch handled >>30tph), hell running more 33rd trains (IINM you can eeke out another 3-4tph before you run up against terminal capacity limits).

You might be right from a purely scientific standpoint, but how could you ever have any faith in the Port Authority to make the extreme changes that you suggested. They still don't have cell phone or WIFI in most of their underground stations, they suckered commuters into believing things would get better after PTC was installed, they were years late on opening WTC, years late on opening (and still not done) opening the new Harrison Station, and for some reason they think their most important PATH-related capital project right now is an extension to EWR?! And for the past 5 years, the Jets have been more likely to win the Super Bowl than the PATH following any sort of consistent weekend schedule.

And by your optimistic standpoint surrounding the PATH and the Port Authority's ability to do anything above the bare minimum, I can tell you're not a regular PATH commuter. (I did 3-straight years/5 days a week (2015-2018) of watching the PATH system detiorate around me.) And don't even get me started on the continuing, seemingly unregulated, construction of massive apartment complexes in Jersey City, Harrison, and now Newark, with complete disregard for the addtional strain on the PATH.

Sir, I simply wish I had your optimism.

[rr503]

But herein lies the problem, no? We have become so used to operational intransigence that we have inducted it into our planning processes, electing to build where we could operate.

Take NYCT. The subway’s signal systems once operated service levels in excess of 30tph on most lines, and did so while the peak was a lot more crowded than today. And yet we’ve convinced ourselves that, despite the fact that the average core-bound track carries only 18tph, we need to spend 40 billion on CBTC for capacity’s sake. I don’t claim to know as much about PATH so I can’t point to specific issues, but I’m confident that there are ways to increase its capacity by simply making the PANYNJ do its job right.

[JoeG]

It isn't only incompetence on PATH's part. A few years ago I met a PATH official and I asked him why the HOB-WTC run took 11 minutes, while when I was a kid in the fifties it took 9 minutes. (That was for the slightly longer run to Hudson Terminal.) He was impressed I remembered and said that the FRA had insisted on the slower times. He also said that the FRA had wanted PATH to do a full, start of day brake test each time the train changed ends at terminals. This would have made it impossible to keep current headways. Fortunately FRA relented on this.
I also noticed that changing engines, and changing to and from electric power, takes much longer than it did back in the day. (This observation is for Amtrak and NJT.) An engineer who used to post on the NJT board once explained that various safety rules made the change take much longer. These changes used to be done in 5 minutes or even less; now they seem to take a half hour. And some things are no longer done. For instance, the Lackawanna would run its electric MU trains from Hoboken to Summit. Then part of the train would run to Gladstone and the rest to Dover. The return trains would be re-assembled at Summit. This seemed to me like a great system. Perhaps it is no longer done because of time-consuming FRA rules. I don't know enough to know for sure.

[Hawaiitiki]

For instance, the Lackawanna would run its electric MU trains from Hoboken to Summit. Then part of the train would run to Gladstone and the rest to Dover. The return trains would be re-assembled at Summit. This seemed to me like a great system. Perhaps it is no longer done because of time-consuming FRA rules. I don't know enough to know for sure.

Combination of some now archaic FRA safety checks and the more complicated coupling that goes along with more complicated equipment these days.

That being said, that process takes place on regularily scheduled trains in Europe allday/everyday, on both rapid-transit and suburban services...with more modern equipment than in the US...in the 5 minute range you speak about. Sooo...yea. Have to throw my hands up sometimes.

[John_Perkowski]

New York and New Jersey can build a Hudson River Tunnel on their own

Click the link to an article from the Poughkeepsie Journal.

[ConstanceR46]

I don't think a rail tunnel on our busiest amtrak line into one of the most important cities is exactly local.

Not to be a pessimist, but how long until a tunnel gives and we have to half the trains... or even worse, a train's in a tunnel when it gives.

[NaugyRR]

...Not to be a pessimist, but how long until a tunnel gives and we have to half the trains... or even worse, a train's in a tunnel when it gives.

Or when the government shuts down and flights get grounded because ATC's aren't showing up for work, forcing people to switch to trains. What happens then when doctors and senators and other people vital to the country can't get through, because the tunnels finally kicked the can?

Not to get political, but last I checked our infrastructure is crumbling to the point where our country is going to enter a literal standstill, yet we just NEED a wall on our southern border for hundreds upon hundreds of billions of dollars?

[EuroStar]

Catastrophic failure of the tunnels such as a collapse is unlikely for two reasons. First, the tunnels are inspected frequently, so the likelihood that signs of imminent collapse get caught is high. Second, civil engineering structures are overbuilt by an order of magnitude, especially the old ones, so they do not just collapse. When they do collapse it is spectacular, but most of the time they do not. They just slowly deteriorate. Slow deterioration is a reliability killer: chipping concrete falls on tracks here, a wall anchor pulls out there dropping a pipe or cable into the way of the trains, a corroded bracket breaks in two dropping the catenary, leaking water freezes and creates icicles that break a pantograph or two, hundred years old paper insulation disintegrates with age and shorts the cables and so on. All of these cause disruption after disruption, stuck trains, broken pantographs, maybe a derailment or two, loss of power and require more and more time for emergency repairs. Eventually the safety risk grows enough that the safety gods require reduction in the speed of the trains to mitigate the consequences of accidents. At that point they cannot run the same number of trains at the lower speed and some service gets cut. And this goes one until a sign of major deterioration is observed and a tunnel gets closed permanently due to safety concerns. Such a sign is not likely to appear for another 25-40 years, but the service will deteriorate from current level much earlier than that.

[Tadman]

Does anybody have a statistic that shows the passenger count of tunnel passengers from outside the NEC area? Outside the NJ/NY area? Would there be any non-NEC passengers if the Florida/Carolina trains were terminated at DC or PHL?

Look at it this way: Given seven long distance trains into NYC from the tunnels, and a train capacity of 400 each, that's 2800 passengers, or less than the capacity of two NJT trains.

On the other hand, wikipedia posits daily ridership of 113,000 for NJT/NEC and 165,000 for Midtown Direct.

Wikipedia indicates 3.5m and 8.5m annual ridership of Acela and Regional. Although all are not going to North Jersey or New York, most are NEC regional riders.

The numbers would indicate that despite the presence of national trains, the tunnels are vastly used by North Jersey residents and New York residents. Neither areas have voters or politicians remotely friendly to President Trump, and I'm not surprised by his decision.

Of course this problem could be solved by making a trade (IE a wall for a tunnel) but nobody seems interested in doing so.

[EuroStar]

But herein lies the problem, no? We have become so used to operational intransigence that we have inducted it into our planning processes, electing to build where we could operate.

Take NYCT. The subway’s signal systems once operated service levels in excess of 30tph on most lines, and did so while the peak was a lot more crowded than today. And yet we’ve convinced ourselves that, despite the fact that the average core-bound track carries only 18tph, we need to spend 40 billion on CBTC for capacity’s sake. I don’t claim to know as much about PATH so I can’t point to specific issues, but I’m confident that there are ways to increase its capacity by simply making the PANYNJ do its job right.

Unfortunately this is not going to change no matter how much you or I do not like it. Our current operational intransigence is for the most part due to "safety" concerns that have been forced on the system by decades of litigation, plaintiff sympathetic juries and politicians who think they know better than the industry experts. In many places speed limits are lower than what they could be because of "safety". We do not allow splitting/merging of consists because of safety even though it used to be done all the time and is still done safely in Europe. We require PTC in the name of safety even though less complex systems will suffice in most situations. We require super heavy equipment in the name of safety, even though the Europeans and the Japanese get along fine with much lighter cars and engines. With all the piling of pointless or duplicative feature in the name of safety, there is no way we are going to get back to the old days in terms of throughput on the Subway, PATH or elsewhere.