Railroad Forums 

  • PATCO to Glassboro?

  • Discussion relating to Southeastern Pennsylvania Transportation Authority (Philadelphia Metro Area). Official web site can be found here: www.septa.com. Also including discussion related to the PATCO Speedline rapid transit operated by Delaware River Port Authority. Official web site can be found here: http://www.ridepatco.org/.
Discussion relating to Southeastern Pennsylvania Transportation Authority (Philadelphia Metro Area). Official web site can be found here: www.septa.com. Also including discussion related to the PATCO Speedline rapid transit operated by Delaware River Port Authority. Official web site can be found here: http://www.ridepatco.org/.

Moderator: AlexC

 #1504595  by EDM5970
 
It shouldn't be too difficult to build an electric-DMU that could run on diesel east of Camden, then switch to DC third rail to get into Center City, via the PATCO line. This would eliminate the danger of third rail outside of the tunnel, and offer a one seat ride into the city. If the car floor and platform heights were the same, such a car could also run on the existing RiverLINE and into the city. But of course the RiverLINE cars have a different floor height.
 #1504611  by WashingtonPark
 
Wouldn't be possible during the rush hour when needed the most. The track is now used to capacity 7 to 9 AM and 4 to 6 PM. There's no room for more trains then. I know because I was involved in testing this. This is why the Riverline never connected with PATCO. You don't even want to begin to get into the politics of the Teamsters and NJT's union trying to work things out, let alone PATCO and NJT and whoever they farm the operations out to trying to work out an agreement. The plan right now is to not even give you a one seat ride from Glassboro to Trenton if you want it, although that could be done operationally and if this thing ever goes through hopefully they would change their mind on that.
 #1505159  by jamesinclair
 
WashingtonPark wrote:Wouldn't be possible during the rush hour when needed the most. The track is now used to capacity 7 to 9 AM and 4 to 6 PM. There's no room for more trains then.
Does Patco run a train every 80 seconds?

Nope. There's plenty of room using global best practices.
 #1505166  by ExCon90
 
I don't see the possibility of 80-second turnarounds at 16th & Locust without extensive (underground) expansion of terminal capacity. (Now if they'd extended it west to form Tracks 11 and 12 at 30th St. Lower Level, which I believe was talked about at one time, or if they could turn it north under 19th St. and the Parkway to someplace where there's room for a proper terminal, but someone would have to figure out where the money would come from.)
 #1505167  by WashingtonPark
 
jamesinclair wrote:
WashingtonPark wrote:Wouldn't be possible during the rush hour when needed the most. The track is now used to capacity 7 to 9 AM and 4 to 6 PM. There's no room for more trains then.
Does Patco run a train every 80 seconds?

Nope. There's plenty of room using global best practices.
They obviously should have had you do all the testing and you could have pointed out where they were wrong on every finding.
 #1505250  by jamesinclair
 
WashingtonPark wrote:
jamesinclair wrote:
WashingtonPark wrote:Wouldn't be possible during the rush hour when needed the most. The track is now used to capacity 7 to 9 AM and 4 to 6 PM. There's no room for more trains then.
Does Patco run a train every 80 seconds?

Nope. There's plenty of room using global best practices.
They obviously should have had you do all the testing and you could have pointed out where they were wrong on every finding.
No, I am not qualified. But the folks running the systems in Paris, London, and Mexico City that do hit 80 second headways on similar lines are, and should be consulted.
 #1505274  by WashingtonPark
 
I'd be interested in knowing how you switch equipment from the Glassboro line into the Lindenwold line on 80 second headways as would the consultants who determined there wasn't enough capacity to do this during the rush hour after almost a year of testing. Maybe the European rail lines have come up with brand new methods they didn't have when they were consulted the first time.
 #1505336  by jamesinclair
 
WashingtonPark wrote:I'd be interested in knowing how you switch equipment from the Glassboro line into the Lindenwold line on 80 second headways as would the consultants who determined there wasn't enough capacity to do this during the rush hour after almost a year of testing. Maybe the European rail lines have come up with brand new methods they didn't have when they were consulted the first time.
Impossible to say without knowing what was studied, what were the assumptions, and what were the constraints. Was it an at-grade cross-over, or a flyover/under? Current signal system and older vehicles or future signal system with new vehicles?

It makes no sense to argue that it is technically impossible to offer 80 second headays on a line with two branches when it is does every day around the world.
 #1505343  by WashingtonPark
 
jamesinclair wrote:
WashingtonPark wrote:I'd be interested in knowing how you switch equipment from the Glassboro line into the Lindenwold line on 80 second headways as would the consultants who determined there wasn't enough capacity to do this during the rush hour after almost a year of testing. Maybe the European rail lines have come up with brand new methods they didn't have when they were consulted the first time.
Impossible to say without knowing what was studied, what were the assumptions, and what were the constraints. Was it an at-grade cross-over, or a flyover/under? Current signal system and older vehicles or future signal system with new vehicles?

It makes no sense to argue that it is technically impossible to offer 80 second headays on a line with two branches when it is does every day around the world.
I know what was studied, what were the assumptions, and what were the constraints. It does make no sense to argue since absolutely, 100%, without a single chance is the Glassboro line going across the Ben Franklin Bridge for a one seat ride to Camden. There's only a 50% chance it will even be built in the foreseeable future. Let's keep this topic discussion to things that are actually on the drawing board to take place.
 #1505356  by ExCon90
 
jamesinclair wrote:It makes no sense to argue that it is technically impossible to offer 80 second headays on a line with two branches when it is does every day around the world.
There has to be track capacity at the end points to support that frequency. You can't "flip" a train at the terminal in 80 seconds; you need enough trackage to accommodate all the trains that are arriving while a previous train is being turned. PATCO has 2 stub tracks and an island platform (and tail tracks?) at 16th & Locust, and that won't support a very tight headway. Of course, when the station was designed it was intended to be a through station on a downtown distributor loop, with no reversals contemplated.
 #1505472  by rr503
 
I'm doubtful of this 80 second headway statistic. The shortest sustained intervals I'm aware of are the 90-100 second headways (=36tph) achieved on some European systems, like London's Victoria line. Now, PATCO currently runs 12tph, so the difference between those aspirational stats is little from a current->potential perspective. All of this said, running that frequency requires either a high density fixed block or moving block signal system, excellent terminals/terminal operations, short dwell times, and well-operated merges. PATCO could certainly have those, but it is currently very much lacking in the realm of terminals (the crossover for the 15th/Locust terminal is east of 12th/Locust, which I imagine massively reduces capacity over what it would be if the x was placed between 12 and 15), and I would not be at all surprised if its signals were designed for relatively low capacity operation -- capacity intensive fixed blocks are $$$$ after all. Investments in those sorts of things would, IMO, be amazing for SJ/Philly transport, but they would need to happen all the same.
 #1505494  by rr503
 
Okay, did some googling. A study from the '90s seems to indicate that PATCO is capable of handling 24tph over its entire route with present signalling, so you could double current service levels at little cost. This is really surprising to me. The terminal config (see: https://youtu.be/AKPVeymCJUQ?t=1335" onclick="window.open(this.href);return false; ) does not seem at all conducive to those sorts of service levels. Glad to be wrong here, though.

https://www.dvrpc.org/Reports/91023.pdf#page=16" onclick="window.open(this.href);return false;
 #1505578  by WashingtonPark
 
rr503 wrote:Okay, did some googling. A study from the '90s seems to indicate that PATCO is capable of handling 24tph over its entire route with present signalling, so you could double current service levels at little cost. This is really surprising to me. The terminal config (see: https://youtu.be/AKPVeymCJUQ?t=1335" onclick="window.open(this.href);return false; ) does not seem at all conducive to those sorts of service levels. Glad to be wrong here, though.

https://www.dvrpc.org/Reports/91023.pdf#page=16" onclick="window.open(this.href);return false;
Unfortunately, subsequent real time testing by actually trying to run that many trains proved many of the assumptions wrong as this was tried during late night-early morning hours and was tying up the owl trains so bad that testing had to be stopped. You're not wrong. Your original post was pretty much spot on. It could be done with trains just running end to end and making no stops. Once you add the station stops in operating more than 16tph causes major delays. Another problem, as ExCon90 pointed out, is the terminal situation. You can do it that fast by running trains right into the pocket and right out without stopping, but that doesn't do the patrons waiting at 16th street any good. This why rush hour local trains don't operate closer than 4 minutes apart, otherwise you'd be getting constant reduced cab codes.
 #1505587  by rr503
 
WashingtonPark wrote:
rr503 wrote:Okay, did some googling. A study from the '90s seems to indicate that PATCO is capable of handling 24tph over its entire route with present signalling, so you could double current service levels at little cost. This is really surprising to me. The terminal config (see: https://youtu.be/AKPVeymCJUQ?t=1335" onclick="window.open(this.href);return false; ) does not seem at all conducive to those sorts of service levels. Glad to be wrong here, though.

https://www.dvrpc.org/Reports/91023.pdf#page=16" onclick="window.open(this.href);return false;
Unfortunately, subsequent real time testing by actually trying to run that many trains proved many of the assumptions wrong as this was tried during late night-early morning hours and was tying up the owl trains so bad that testing had to be stopped. You're not wrong. Your original post was pretty much spot on. It could be done with trains just running end to end and making no stops. Once you add the station stops in operating more than 16tph causes major delays. Another problem, as ExCon90 pointed out, is the terminal situation. You can do it that fast by running trains right into the pocket and right out without stopping, but that doesn't do the patrons waiting at 16th street any good. This why rush hour local trains don't operate closer than 4 minutes apart, otherwise you'd be getting constant reduced cab codes.
Dunno about this. Some more research found a track map on NYCSubway which seems to show relay capability at 15th. Squinting at a bunch of RFW videos corroborated this; there's a crossover just beyond the station and there seems to be, in at least one of the videos, a train parked beyond the crossover.

Map:

https://s3.amazonaws.com/nycsubway.org/ ... -track.pdf" onclick="window.open(this.href);return false;

Video (watch in HD to see properly):

https://youtu.be/2b1LLCXN7X8?t=1760" onclick="window.open(this.href);return false;

...which suggests that there is at least some potential to operate a (vastly more efficient) relay terminal rather than today's insane two-station long pocket setup.

The DVRPC report, FWIW, seems to have been done after field tests ("Their calculations and actual field tests show that 24 six-car trains could be operated during the evening peak hour, the period in which 20% of the daily ridership must be carried in the peak direction, according to PATCO experience") which is again suggestive of there being more capability with some relatively simple operational changes.
 #1505612  by WashingtonPark
 
rr503 wrote:
WashingtonPark wrote:
rr503 wrote:Okay, did some googling. A study from the '90s seems to indicate that PATCO is capable of handling 24tph over its entire route with present signalling, so you could double current service levels at little cost. This is really surprising to me. The terminal config (see: https://youtu.be/AKPVeymCJUQ?t=1335" onclick="window.open(this.href);return false; ) does not seem at all conducive to those sorts of service levels. Glad to be wrong here, though.

https://www.dvrpc.org/Reports/91023.pdf#page=16" onclick="window.open(this.href);return false;
Unfortunately, subsequent real time testing by actually trying to run that many trains proved many of the assumptions wrong as this was tried during late night-early morning hours and was tying up the owl trains so bad that testing had to be stopped. You're not wrong. Your original post was pretty much spot on. It could be done with trains just running end to end and making no stops. Once you add the station stops in operating more than 16tph causes major delays. Another problem, as ExCon90 pointed out, is the terminal situation. You can do it that fast by running trains right into the pocket and right out without stopping, but that doesn't do the patrons waiting at 16th street any good. This why rush hour local trains don't operate closer than 4 minutes apart, otherwise you'd be getting constant reduced cab codes.
Dunno about this. Some more research found a track map on NYCSubway which seems to show relay capability at 15th. Squinting at a bunch of RFW videos corroborated this; there's a crossover just beyond the station and there seems to be, in at least one of the videos, a train parked beyond the crossover.

Map:

https://s3.amazonaws.com/nycsubway.org/ ... -track.pdf" onclick="window.open(this.href);return false;

Video (watch in HD to see properly):

https://youtu.be/2b1LLCXN7X8?t=1760" onclick="window.open(this.href);return false;

...which suggests that there is at least some potential to operate a (vastly more efficient) relay terminal rather than today's insane two-station long pocket setup.

The DVRPC report, FWIW, seems to have been done after field tests ("Their calculations and actual field tests show that 24 six-car trains could be operated during the evening peak hour, the period in which 20% of the daily ridership must be carried in the peak direction, according to PATCO experience") which is again suggestive of there being more capability with some relatively simple operational changes.
Field tests were done after that nearly 30 year old report. I know, I was involved in them. 24tph works fine without any station stops. Add in picking up and discharging passengers and things back up terribly. You could run a lot more than 24 if you didn't care how long it took you to traverse the line and you had unlimited car equipment and turn capacity. If you want an efficient operation 24 won't work. There is a crossover west of 15th. There is barely enough room in the pocket to turn a 6 car train and you must creep up to EOT to avoid hitting the bumper block, which is an automatic work suspension. None of this really matters for Glassboro anyway. There are no plans to run those trains to Philadelphia over PATCO property. PATCO and TNJ have determined it isn't feasible. It might look good on paper but it doesn't work in the real world without a huge amount of money being spent. The Federal government hasn't even committed their share to providing service that is basically a restartup of glorified RDC service on a right of way that is already there.
  • 1
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9