16-Car Trains on the Northeast Corridor
The dominant length of high-speed rail platforms in China, Japan, South Korea, and Europe is 400 meters, which usually corresponds to 16-car trains. The Northeast Corridor unfortunately does not run such long trains; intercity trains on it today are usually eight cars long, and the under construction Avelia Liberty sets are 8.5 cars long. Demand even today is high enough that trains fill even with very high fares, and so providing more service through both higher frequency and longer trains should be a priority. This post goes over what needs to happen to lengthen the trains to the global norm for high-speed rail. More trains need to be bought, but also the platforms need to be lengthened at many stations, with varying levels of difficulty.
The station list to consider is as follows:
- Boston South Station
- Providence
- New London-HSR
- New Haven
- Stamford
- New York Penn Station
- Newark Penn Station
- Trenton
- Philadelphia 30th Street
- Wilmington
- Baltimore Penn Station
- BWI
- Washington Union Station
Some of these are local-only stations – the fastest express trains should not be stopping at New London or BWI, and whether any train stops at Stamford or Trenton is a matter of timetabling (the headline timetable we use includes Stamford on all trains but I am not wedded to it). In order, allowing 16-car trains at these stations involves the following changes.
Boston
South Station’s longest platforms today are those between tracks 8 and 9 and between tracks 10 and 11, both 12 cars long. To their immediate south is the interlocking, so lengthening would be difficult.
Moreover, the best platforms for Northeast Corridor trains to use at South Station are to the west. The best way to organize South Station is as four parallel stations, from west to east (in increasing track number order) the Worcester Line, the Northeast Corridor and branches, the Fairmount Line, and the Old Colony Lines, with peak traffic of respectively 8, 12 or 16, 4 or 8, and 6 trains per hour. This gives the Northeast Corridor tracks 4-7 or possibly 4-9; 4-7 means the Franklin Line has to pair with the Fairmount Line to take advantage of having more tracks, and may be required anyway since pairing the Franklin Line with the Northeast Corridor (Southwest Corridor within the city of Boston) would constrain the triple-track corridor too much, with 12 peak commuter trains and 4 peak intercity trains an hour.
The platform between tracks 6 and 7 is 11 cars long, but to its south is a gap in the tracks as the interlocking leads tracks 6 and 7 in different directions, and thus it can be lengthened to 16 cars within its footprint. The platform between tracks 4 and 5 is harder to lengthen, but this is still doable if the track that tracks 5 and 6 merge into south of the station is moved in conjunction with a project to lengthen the other platform.
Of note, the other Boston station, Back Bay, is rather constrained, with nearly the entire platforms under an overbuild, complicating any rebuild.
Providence
Providence has 12-car platforms. The southern edge is under an overbuild with rapid convergence between the tracks and cannot reasonably be extended. But the northern edge is in the open air, and lengthening is possible. The northern edge would be on rather tight curves, which is not acceptable under most standards, but in such a constrained environment, waivers are unavoidable, as is the case throughout urban Germany.
New London
This is a new station and can be built to the required length from the start.
New Haven
The current station platforms are only 10 cars long, but there is space to expand them in both directions. The platform area is in effect a railyard, a good example of the American tradition in which the train station is not where the trains are (as in Europe) but rather next to where the trains are.
A rebuild is needed anyway, for two reasons. First, it is desirable to build a bypass roughly following I-95 to straighten the route beginning immediately north of the station, even cutting off State Street in order to go straight to East Haven rather than curve to the north as on the current route. And second, the current usage of the station is that Amtrak uses tracks 1-4 (numbered west to east as in Boston) and Metro-North uses tracks 8-14, which forces Amtrak and Metro-North trains to cross each other at grade from their slow-fast-fast-slow pattern on the running line to the fast-fast-slow-slow pattern at the station. In the future, the station should be used in such a way that intercity trains either divert north to Hartford or Springfield or go immediately east on a flying junction to the high-speed bypass toward Rhode Island, without opposed-direction flat junctions; the flying junction is folded into the cost of the bypass and dominates the cost of rebuilding the platforms, as the space immediately north and south of the platforms is largely empty.
Stamford
Stamford has 12-car platforms. Going beyond that is hard, to the point that a more detailed alternatives analysis must include the option of not having intercity trains stop there at all, and instead running 12-car express commuter trains, lengthening major intermediate stops like South Norwalk (currently 10 cars long) and Bridgeport (currently 8) instead.
To keep the mainline option of stopping at Stamford, a platform rebuild is needed, in two ways. First, the station today has five tracks, a both literally and figuratively odd number, not useful for any timetable, with the middle track, numbered 1 (from north to south the numbers are 5, 3, 1, 2, 4), not served by a platform. And second, the platform between tracks 3 and 5 can at best be lengthened to 14 cars, while that between tracks 2 and 4 cannot be lengthened without moving tracks on viaducts. This means that some mechanism to rebuild the station should be considered, to create four tracks with more space between them so that 16-car platforms are viable; this should be bundled with a flying junction farther east to grade-separate the New Canaan Branch from the mainline.
A quick-and-dirty option, potentially viable here but almost nowhere else, is selective door opening, at the cost of longer dwell times. Normally selective door opening should not be used – it confuses passengers, for one. However, here it may be an option, as intercity traffic here is unlikely to be high; traffic today is 323,791 in financial 2023, the lowest of any station under consideration in this post unless one counts New London. The only reason to stop here in the first place is commuter ridership, in which case mechanisms such as restricting unreserved seats to the central 12 cars can be used.
New York
Penn Station has multiple platforms already long enough for 16- and even 17-car trains, including the one we pencil for all high-speed intercity trains in the proposal, platform 6 between tracks 11 and 12, as well as the two adjacent platforms, 5 and 7. (Note that unlike at New Haven and Boston, platform numbers at Penn increase south to north, that is right to left from the perspective of a Boston-bound traveler.)
Thank the god of railways, since platform expansion requires a multi-billion dollar project to remove the Madison Square Garden overbuild in the most optimistic case; in a more pessimistic case, it would also require removing the Moynihan Station overbuild.
Newark
Newark Penn Station’s platforms are in a grand structure about 14.5 cars long. Thankfully, they extend a bit south of it, producing about 16 cars’ worth of platform on the west (southbound) side, between tracks 3 and 4; as in New York, track numbers increase east to west. On the east side, PATH interposes between the two tracks, which have a cross-platform transfer from northbound New Jersey Transit trains to PATH. The platform structures and their extensions do have enough length to allow 16-car trains – indeed they go as long as 18 – but the southern ends are currently disused and would require some rehabilitation.
Trenton
Trenton has a 12.5 car long southbound platform and an 11.5 car long northbound platform. There is practically no room for an expansion if no tracks are moved. If tracks are moved, then some space can be created, but only enough for about 14 cars, not 16.
However, traffic is low, the second lowest among stations under consideration next to Stamford. The suite of Stamford solutions is thus most appropriate here: selective door opening with only the middle 12 cars (naturally the same as at Stamford) open to commuters, or just not stopping at this station at all. The only reason we’re even considering stopping here is timetabling-related: trains should be running every 10 minutes around New York but every 15 between Baltimore and Washington, or else significant expansion of quad-tracking on the Penn Line is required, and so a local stop should be added as a buffer, which can be Trenton or BWI, and BWI has twice the current Amtrak traffic of Trenton.
Philadelphia
30th Street Station has 14-car platforms. Selective door opening is basically impossible given the high expected traffic at this station, and instead platform expansion is required. There is an overbuild, but the tracks stay straight and only begin curving after a few tens of meters, which gives room for extension; from the north end to the overbuild to where the tracks begin curving toward one another to the south is 15.5 cars, and there is room north of the overbuild between the tracks.
Whatever reconstruction project is needed is helped by the low traffic at these platforms. SEPTA uses the upper level of the station, with tracks oriented east-west. The north-south lower level is only used by Amtrak, which could be easily reduced to three platform tracks (two Northeast Corridor, one Keystone) if need be, out of 11 today. Thus, staging construction can be done easily and intrusively, with no care taken to preserve track access during the work, as half the station platforms can be closed off at once.
Wilmington
Wilmington is frustrating, in that there is platform space for 16 cars rather easily, but it’s on inconsistent sides of the tracks. Track numbers increase south to north; track 1 has a side platform, there’s an island platform between tracks 2 and 3, and then track 3 also has a side platform on the other side, extending well to the east of the island platform. The island platform and the track 1 platform are about 12.5 cars long, and the track 3 side platform is 13.5 cars long. Thus, an extension, selective door opening, or a station rebuild is required.
The island platform can be extended about one car in each direction, so it cannot be the solution without selective door opening. Both side platforms can be extended somewhat to the west: the track 1 platform can be extended to 16 cars, but it would need to be elevated in the narrow space between the track viaduct and the station parking garage; the track 3 platform can be extended in both directions, avoiding a new elevated extension over North King Street.
If for some reason an extension of the track 1 platform is not possible, then selective door opening can be used, but not as reliably as at lower-traffic Stamford or Trenton, and overall I would not recommend this solution. A station rebuild then becomes necessary: the station has three tracks but doesn’t need more than two if SEPTA and Amtrak can be timetabled right, and then the removal of either track 1 or track 2 would create space for a longer platform.
Baltimore
Baltimore Penn has seven tracks, numbered from south to north 1, 3, 4, 5, 6, 7, F. Their platforms are 10 to 13 cars long. Northbound trains are more or less forced to use the platform between tracks 1 and 3, since the way the route tapers to a three-, then four-track line to the east forces all eastbound trains to use mainline track 1; this platform is rather narrow at its east end but has space to the west for a 16-car extension. Westbound trains can use either the platform between tracks 4 and 5 or that between tracks 6 and 7, with tracks 4 and 6 preferred over 7 as they reach the express westbound track (track 5 stub-ends). Both platforms can be extended, with the platform between tracks 6 and 7 requiring a one-car extension to the east where a ramp down to track level for track workers exists whereas that between tracks 4 and 5 has ample unused space to its west.
BWI
The two side platforms at BWI are just under 13 cars long. However, nowhere else on the corridor is an extension easier: the station is located in an undeveloped wooded area, with space cleared on both sides of the track so that tree cutting is likely unnecessary west of the tracks and certainly unnecessary east of them.
The station itself needs a rebuild anyway, due to already existing plans to widen it from three to four tracks. This is required to enable intercity trains to overtake commuter trains anyway, unless delicate timetabling on triple track is used or another part of the Penn Line is set up as a four-track overtake. The plans are rather advanced, but platform extensions can be pursued as an add-on, without disturbing them due to the easy nature of the right-of-way.
Washington
Washington is set up as two separate stations, a high-platform terminal to the west and a low-platform through-station to the east on a lower level. Track numbers increase west to east, the western part taking 7-20 (though only 9-20 are high and wired) and the eastern part 23-30. None of the western platforms is long enough, but multiple options still exist:
- The platform between tracks 9 and 10 has room for an extension.
- The platforms between tracks 15 and 16 and between tracks 16 and 17 look like they already have extensions, if not open for passengers.
- The platforms between track 17 and track 18 and between tracks 19 and 20 are only 12 cars long, but tracks could be cannibalized in the open air to make a long enough platform, especially since the reason track numbers 21 and 22 are skipped is that there used to be tracks there and now there’s empty space.
- The platform between tracks 25 and 26 is long enough, and could be raised to have level boarding.
The existing platforms that can be extended easily are sufficient in number, but probably not in location – it’s ideal for the platforms to be close together, to simplify the interlocking as trains have to be scheduled to enter and leave the station without opposite-direction conflicts. If it’s doable even with a split between platforms separated by multiple tracks then it’s ideal, but otherwise, the extra work on tracks 17-20 may be necessary, converting a part of the station that presently has six tracks and four platforms into likely four tracks and two platforms.
Conclusion
All of this looks doable. The hardest station, Stamford, is skippable if selective door opening is unviable after all and a rebuild is too expensive. Among the other stations, light rebuilds are needed at Boston, Wilmington, and maybe Washington; New Haven needs a more serious rebuild as part of the bypass, but the station platforms are a routine extension where there is already room between the tracks. The most untouchable station, New York, already has multiple platforms of the required length at the required location within the station.
Pedestrian Observations 
16 car trains could be run as 2×8 car sets, with only one set’s doors opening at the stations with shorter platforms. This could simplify communication with travellers, but interlocking on one side of the platform might need to be reorganised/rebuilt because more space would be required on that one side.
The problem with doing it this way is that Boston would only get the northern set and Washington would only get the southern set if the platforms are as they are today, so at least one of the two needs work.
Absolutely, I was referring to the smaller intermediate stations. Big stations like Philly Boston Baltimore Washington New York are non negotiable, all doors must open.
I think it’s preferable for both halves of a 2×8 set to include doors that open at a given reduced-length station.
Have you taken split train services in Europe? the two sections of the trains will have different train numbers (will be considered different services) and the train will be listed as non stop at the stations where doors do not open. I think this would actually cause less confusion than having a train with only select door opening. Since Alon believes that there should not be universal seat assignments passengers only have to understand which train service goes to their destination and are guaranteed that doors open at the stop that they need to go.
Does your South Station analysis take into account the new tower that shortened the platforms?
I believe so; everything in the open air has been sanity-checked on Google Earth, including the southern ends of the South Station platforms. The only places where I’m relying on written information about platform lengths rather than direct measurement are New York, Newark, and Philadelphia.
The Congressional Limiteds, the luxury express between New York and Washington, on the Pennsylvania Railroad, were18 cars long. The Senator, the luxury express between Boston and Washington was 14 cars long. So there was space, at major stations, in the past.
On a tangentially related topic, why are current and the planned future trains 8 cars long? 11 car Northeast Regional trains definitely existed in the past.
If 11 cars would fit why not run an 11 car train every 20 minutes?
Almost all platforms are capable of taking 12-car sets now. Surely you do this and get 50% capacity for a fraction of the cost rather than going to 16-car sets just cause it’s what other places do? Is there a nailed on need for 100% additional capacity if frequency increases too (going by your operations before concrete principle).
HS2 in the UK made the classic mistake of adopting 400m trainsets made up of 2 by 200m. Great when you’re building new infrastructure but the moment you start looking at interfacing with existing stations you either end up;
A. Avoiding station rebuilds by using a single 200m train set, to the detriment that existing rolling stock is often already longer than this at 10-12 car so your capacity actually takes a hit.
B. Blowing up the existing railway in places to get a 400m train set for locations where there’s almost certainly not demand for it. I wouldn’t underestimate the costs of this!
At the end of the day in the UK there will definitely be a desire to have been more flexible in train length. I guess my lesson learnt in this context, be absolutely sure you want and need 400m trains and all the knock on consequences before you dive in.
I don’t think lengthening to 16 car is necessary initially at least. The thing is once you’re locked into 12 car sets you can’t upgrade to 16 car that easily. Will you run 2×6 cars initially or just buy 12 car sets? Safeguarding the room required for platform extension probably should be a higher priority than actually extending the platforms at this stage.
They lengthened half the Pendelinos from 9 to 11 cars in the Uk. It was fine
A lot of the cases where HS2 was planned to interface with existing stations, it was only one existing station as a terminus (e.g. Liverpool, Manchester, Blackpool, Hull) and many of these stations have enough tracks that they could run 2×8 car trains and just split the train at the terminus station across two tracks. Certainly in both Liverpool and Manchester, they could locate those two tracks either side of an island platform so passengers can go to that one platform and then board either side, with the train reassembling on the way out of the station. It might be preferable in practice to run two paths out of the mainline station and combine a little later, so as not to go shunting in the throat of a busy station, so they’d leave a minute or two apart and then combine at Stockport (for Mancester) or Runcorn (for Liverpool), allowing for one driver to just travel back and forth Manchester-Stockport or Liverpool-Runcorn.
The plan for trains to Scotland was for a 2×8 train to leave London (or Birmingham), to run as a single train into Scotland and then split, with half going to Glasgow and half to Edinburgh. For trains from London, they run non-stop north of Birmingham; for trains from Birmingham, they would either stop in intermediate stations with SDO or they would split earlier with one half running non-stop and the other half stopping in all intermediate stations.
The plan for London-Newcastle (which has been completely cancelled) was 1×8, but with the intention to eventually run 2×8 with one half running non-stop and the back half stopping in York and Northallerton (there’s no particular shortage of paths on the ECML north of York). They’d probably have to do the shunting movement in York station, so the “non-stop” train would be stationary in York, but it wouldn’t necessarily be aligned with a platform.
I think you’ve misunderstood some of the HS2 proposals. The platforms at Runcorn and Stockport aren’t long enough to split and join 2 x 8-car trains. If you tried to do this in their current arrangement you’d be sat foul of signals and locking up the station throat and they’re very hard to alter. The issue is the same for the Scotland split whether you pick Carlisle or elsewhere. Hence you need to do lots of painful infrastructure work to make this happen and probably have to trade off other functionality to make it happen.
Splitting and joining in the Picc / Lime Street throat would be madness, they’re far too busy and it’s unlikely it would be acceptable as a safe operational practice. For the reason where there were splits in the NW I think this was always slated to happen at Crewe which was to be rebuilt to have 400m platforms.
HS2 was not properly planned to interface with existing infrastructure which is why rolling it back to Phase 1 only is so impactful as no purpose built terminal station at Piccadilly will leave you with a capacity reduction or some painful infrastructure changes.
Slightly confused where you got the idea that HS2 was going to run using SDO and run trains to Hull or Blackpool, I don’t think these were ever part of the plan?
“platform numbers at Penn increase north to south, that is right to left from the perspective of a Boston-bound traveler“
*-1 one of these?
…south to north. Fuck.
Curiosity: Is there any restriction on train length, besides platform size? Freight lines in Europe have sidetracks designed to allow for 750m trains. Could you connect platforms to one such sidetrack and make a very long passenger train, or there’s some safety regulation against that? Between 400m and 750m, is the cost of running the train proportional to the length, or you could actually have economies of scale?
The US routinely runs 5-6km long freight trains. Belgium ran a 1.7km passenger train once as stunt. (https://en.wikipedia.org/wiki/Longest_trains)
However there are practical problems with running that long in passenger service. The length of the platform is beyond walking distance, so you need assigned seating and passengers need to arrive at the correct end of the platform (if you park or your bus arrives on the north end but your seat is on the south end that is a problem), and when you arrive the place you are going (likely a transfer bus) better be on that end of the train. This becomes impossible to schedule with a full train (why are you running 16 car trains if you are not full?) where people come from different locations in their start city and many different destinations in the final city – someone will need to get on the train on the south end, and get off on the north end (or have a long walk on a platform – implying a missed connection).
The US likes locomotive hauled trains. Alon has written extensively on why locomotive hauled trains are a bad idea, so search out their previous posts on the subject if you don’t understand why longer trains are worse when locomotive hauled.
In general I’d say if 8 car trains every 5 minutes isn’t enough you should be building more tracks to different locations. The NEC has a few possible routes hitting different cities with different trip times (North you can go up the coast or inland. South there really isn’t any obvious second route). It would be better to build all those and run shorter trains because of the additional destinations served. Running long distance trains every 5 minutes also means that local trains can no longer use these same tracks, which means building them dedicated tracks. At reasonable construction costs and timelines the above could be done and be profitable long term, but this is the NEC so we can’t assume reasonable costs or timelines. Thus Alon is advocating for 16 car trains here, it is probably the best option that we can get in my lifetime even if not my preferred option.
8-car trains every five minutes (presumably in each direction) is not practical with the New York tunnel bottleneck and the need to share the tunnels with NJ Transit.
Also, NY Penn Station operation would have to change to true thru-running, instead of current long stops there.
If the US (and NYC) had reasonable construction costs we would build more tunnels and thus not have to share. Tunneling under the river will always be expensive, but it shouldn’t be nearly as expensive as it is. There is more than enough demand in NYC to pay for more tunnels at reasonable prices and the better service more tunnels enable would boost ridership as well. However instead of reasonable prices we have every reason to believe that unreasonable prices are all we can get and thus more tunneling isn’t affordable and won’t pay for itself if we do it anyway.
Alon has written extensively on why thru-running is a good idea anyway. There should not be long stops at any station in the middle of a route (even at the end you should do the long stop in a maintenance yard in the middle of nowhere if you need one, not in the station where there is better use for that land than storing a train.
Unless the tunnel is free there will be sharing. Because even low low cost tunnels aren’t free.
I think you overestimate the difficulty of partial door openings, now that Amtrak has 100% reserved seating. You just program the computer not to sell seats in cars whose doors don’t open at the origin or destination.
This doesn’t allow 16-car trains on the state-run railroads, but that wouldn’t be possible anyway because it would require long platforms at local stations.
Reserved seating is not assigned seating. NEC Amtrak trains (except Acela and Regional Business class) don’t have assigned seats, you are merely guaranteed a seat on the train, not a particular one.
So you reserve someone a seat on a particular section of the train. First, and second/third classes often exist on the same train. The computer knows where you want to go and so it reserves you a seat on the blue cars if you need it, and the orange cars otherwise. Or some such scheme to section off the train.
I’m surpised that although you’ve mentioned grade separating the connection to the New Cannan Branch, you haven’t mentioned the need to grade separate the connection between the stamford railyard and the southbound local track, to allow local trains to change direction without crossing the express tracks. As these trains will be running at 12tph, I don’t think its a good idea to make them cross the express tracks at grade.
Good idea, but sounds like too much construction to be achieved promptly. Could 12 car trains be used much more quickly from a trainset and platform perspective?
When would Amtrak have enough cars for 12 or 16 car trains?
Passenger access to the far end of a train is an issue, especially at terminals and where there is no platform cover at the far end. (As I recall, most large, Tokaido shinkansen stations have escalators at around car 6 and 10.
How necessary is the stop at Back Bay?
Why not 14-car car trains since we’ll be limited to Avilia Liberty locomotives for 25-30 years and it would be less expensive to upgrade Washington, Baltimore, Wilmington, New Haven, Providence and Boston? It should also be easier to timetable Amtrak trains with commuter trains.
Amtrak doesn’t have a stash of trains they aren’t telling anybody about. If they want to extend trains they have to go out and buy more train. If they want to increase frequency they have to buy more train. It can be something different.
Take two deep breaths. Then have a few sips of whatever beverage you sip and take two deep breaths. Then chant Boston is not Tokyo, New Haven is not Nagoya, Stamford is not the center of the Universe.
Vermont seems to be aiming to serve most of the state via Albany but lets keep this simple. there are 15 million people in New England. They get Japanese like urges to travel by high speed rail that’s 30 million trips a year. One round trip per person. But lets keep it even simpler. At 36.5 million trips a year, a bit over one round trip per person, it’s 100,000 a day. 50 towards Boston 50 away. Some of them will be heading to Albany.
To keep it simple, 5,000 of them heading to Albany per day, is 45,000 going through New Haven. Again, to keep it simple, 15 hour service day is 3,000. Even when the bridge to Long Island is completed in 2247 they need 16 car platforms in a few places. The trains that stop all the places can be shorter. Or instead of three 1,000 passenger trains, four 750 passenger trains. They could be clever and coy and have half hourly super duper almost no stops between DC and Boston 1,000 passenger and four 500 passenger that make all the stops. Six trains an hour between Jamaica and Farmingdale means there would be capacity for half hourly LIRR express on the Main Line and half hourly to South Shore. They can haggle over where the ginormous park-n-rides are in 2231.
New England isn’t Tokyo. It’s not Yokohama either. Take two deep breaths.
New England isn’t tokyo, isn’t yokohama, but Yokohama and Tokyo have multiple times the number of passenger/commuter tracks entering the city compared to New York. Although the travel demand is much smaller, almost all of the traffic is concentrated on one single corridor that requires interlining between intercity and commuter trips. This does not happen in Japan.
Also traffic demand is not spread out like that across 15 hour days. For the initial system 16 car platform probably wouldn’t be needed, but the option to extend them at a later date probably will be desirable. The space should be safeguarded even if the capacity is not needed initially. 8 car sets should be fine initially, then doubling up of select peak hour trains is an option.
The LIRR spent an astounding amount of money to free up capacity in the East River Tunnels. Capacity isn’t a problem for a very long time. Amtrak proposed having their alternative to ARC open this year. It should be any decade now.
The new generation of Acela are 9 cars long with one of them being the cafe car. 8 cars of seats. Now. With options to add cars.
I don’t know or care how long the Regionals are or what kind of short term plans Amtrak has for them. Something because the Amfleets are getting very old.
None of that changes that there are only 15 million people in all of New England. And they, in some alternate universe where they have Japanese levels of demand, sometime far far in the future, need 50,000 seats in each direction, per day. Works out to an average of 3,000. They can decide if it’s three 1,000 passenger trains or four 750 passenger trains or two 1,000 passenger trains skimming off the traffic to and between New Haven, Providence and Boston and a 500 passenger train that makes all the stops along the Shore Line and 500 passenger train that makes all the stops on the Inland Route. Or they take their personal helicopter to the SST port in Willimantic..
The Avelia Liberties are 8.5 cars long, including power cars; they are set up as 2 power cars and 9 coaches, but each car is shorter than the typical American railcar or global high-speed rail car, in the style of the TGV from which the Avelia is derived.
Assuming the seat count on Wikipedia is accurate nearly 27 percent more seats. And there were supposed to be more trips because they will have more trains. Nobody needs 16 car platforms until the 12 car trains are full. Which isn’t going to happen for a long long time.
“need 50,000 seats in each direction, per day. Works out to an average of 3,000.”
But that’s not how transportation demand works in real life. The New York Subway runs 24 hours a day but the trains are packed at 8 am yet empty at 3 am. If you averaged out demand this way the logical thing to do would be to close multiple tunnels in midtown Manhattan, which would be stupid. The 9 PM and 6 AM trains will not be as full as midday trains, the trains on a Tuesday in March will not be as full as on an average Friday or the Wednesday before Thanksgiving. You can’t base your capacity on averaging yearly ridership equally over days and hours.
Also, if you have 100k riders per day you need to provide about 125k seats per day (give or take). Because of the issue above (travel demand is not even across days or time or day) you cannot fill seats evenly at all times. If 1200 people want to travel at the time you have a single 1000 seat train scheduled, then you will lose 200 passengers to another mode of travel. There is also the issue that travel demand is not equal across space either, a train needs seats of the people who get on in Providence as well as start the journey in Boston, which means empty seats in Boston. There are of course days or times when trains will be 100% full, but the geometry of trains is such that it is very easy to provide enough seats that they are 80% full on average, so that last minute travelers can almost always get a ticket, which builds your ridership base.
What part of 15 hour service day got you confused? the 15? or the day? perhaps it was hour.
Or the parts about how 3,000 New Englanders would suddenly get the urge to use railroad that even the Japanese would find excessive?
And I know how railroads work. If New Englanders decided to take the train at levels even the Japanese would find high they could run a 5th train every hour on the same track Might even have to quad track it between Providence and Boston. Someday far far in the future that will never come. Because there are automobiles and aeroplanes.
Long distance ridership is pretty flat to be fair.
Also you can have standing on busy trains
As a measure of both cost and capacity, would it suffice to simply modify existing trainsets (and platforms where necessary) to be a full-walkthrough 12-12.5 car consist in the interim, as well as lengthing existing NER trainsets as well? With respect to the utter ineptitude of the current administration, and its open hostility to rail and capital infrastructure projects, would this alternative have any credence in the current political environment while also being a semi-sensible alternative?
No it’s not. It’s almost exactly halfway between New York and D.C. the super duper New England express that only stops in New Haven and Providence before disgorging a crowd in Boston, is blasting along at 350 kph across the swamps next to I-495, the New England “all stops” and the NY-DC “all stops”, the 10 or 12 cars of it, can toddle through the tight curves on either side of Wilmington.
You can’t on one hand attempt to compare it to the Tokaido Shinkansen and then ignore that there are three different stopping patterns along the Tokaido. Might be three or four trains. But railfans’ brains melt if they have to consider the Inland Route and the Shore Line. Or PRR route versus the Reading between NY and Philadelphia. Or that there can be a New England train that doesn’t go through Midtown but goes through Brooklyn and Wall Street.
Adirondacker is correct here. A full HSR service for the NEC would see the express service stopping in DC-Balt-Phila-NY-New Haven-(Prov)-Bos. That’s where all of the ridership is and how you get 2 hr DC-NY or 4 hr DC-Bos. There is no need to re-create the Acela stopping pattern for a fundamentally different service (the Acela stopping pattern would work for an intermediate level service between express and regional/all stop). The limited/regional services stopping in Wilmington and other places should be able to get by fine on 12 car trains. Isn’t something like 75% of the departures on the Tokaido Shinkansen the Nozomi express service?
Newark and Wilmington are in slow zones; the amount of time saved from skipping them is too small to bother with. New Haven and Providence, same thing.
There is no easy bypass of Newark. There is for Wilmington. If there is an ….intercity train… surfing along the bypass of Harve De Grace Maryland every ten minutes, in each direction, it’s okay if the one that is shoveling people to New England doesn’t make a lot of stops. Every half hour with a 2:30 trip to Boston, from Wilmington, is better than anything other than a private helicopter.
People in Newark understand the concept of trains, if the Shovel-’em-to-NewEngland train goes past the platforms at a gracious speed it saves a few minutes. I know this is brain meltingly complicated to clueless railfans: if you encourage someone who wants to go from Washington DC to Boston to use the express train they aren’t on the local train freeing up a seat for someone who doesn’t want to go to Boston. Perhaps even not-go to Stamford, Center of the Universe.
“the amount of time saved from skipping them is too small to bother with”
You have sometimes identified the savings from certain improvements (wider curves, higher speed switches, etc.) by the number of seconds saved. Even at slower speeds skipping a stop saves a few minutes in deceleration, time at platform, and acceleration. Cutting out six stations at 2.5 min per station (a reasonable penalty at 200 kph) would be 15 minutes saved, how is that too small to bother with?
Also, if one is planning for 16-car trains on the NEC, isn’t one also planning for HSR upgrades such that these stations are no longer in slow zones and thus the savings are greater?
“New Haven and Providence same thing”
New Haven should not be skipped by the Express/Nozomi service. It is the junction between the Hartford/Springfield branch and the Providence branch and thus vital for connections. As it is right in the middle of the dense part of Connecticut along I-95 and I-91) it is also a good spot for people to switch from regional trains so that people originating at a smaller stops like Old Saybrook or Bridgeport can get on the express to Washington or Boston.
Providence is a maybe. It is not quite athird of the way Bos-NY (New Haven is a third) so theoretically provides a balanced location for two stops for connections on the northern half of the NEC. But the area around it is not as dense as New Haven so its not essential. But if the time savings are truly that low due to slow zones then fine keep it, its still a roughly 1M person MSA. The issue isn’t New Haven/Providence on your stop list it is Stamford, Trenton, Wilmington, etc.
It’s also difficult to believe you could really get Japanese style one minute stops, much more plausible that the minimum is 2-3 minute stops. So at 200km/h that would be the better part of 5 minutes of penalty?
Quite. Onux, like Adirondacker, is not doing his arguments any favours. Maybe 2.5 min for city Metro but for HSR? Since you railfans put the Japanese on a pedestal:
Now, 8 min seems a bit long and I wonder if the author hasn’t subtracted the time the train would take to traverse the same distance without stopping? In any case the difference is nothing like 2.5 min. His point about it being “often much longer” is valid whenever you start running express and stopping services on the same tracks. In similar arguments about why big-bad SNCF doesn’t have the Paris-Marseilles TGV stop at Lyon, I recall Max Wyss pointing out “the stopping penalty for Lyon Part-Dieu is more in the 20 minutes range“.
It’s eight minutes on the Shinkansen only because Kodama trains get overtaken by Nozomi trains at local stations. The actual capability of N700s is to decelerate and accelerate between 0 and a top speed of 300 km/h at a time cost of just 139 seconds plus the dwell time.
The numbers on the TGV are much higher, for three reasons:
1. The TGVs are not set up for rapid boarding and alighting so the dwell time is always longer.
2. They have power cars so the acceleration is much lower.
3. The LGVs are set up with the urban stations all off the mainline – the stop penalty at Lyon Saint-Exupéry is not that high, but then at Part-Dieu trains would have to get off the LGV and spend a lot of time on the low-speed approaches.
So, strictly interpreting what you wrote it will be 139 seconds for both decel and accel, thus 278s already 4’38” plus say 1m for dwell time = 5’38” (5.6 min). More than double what Onux was claiming. (Surely, on this blog, we can avoid all this handwaving speculation and find the actual average. Christopher Hood, an acknowledged rail expert has published 8 minutes. Is he wrong?)
Do Shinkansen really only dwell for 60 seconds? Matthew Hutton suggests TGVs stop for 5 minutes though since most run terminal-terminal I’m not sure what he means; I’m pretty sure last time I took it, the TGV from Bordeaux didn’t stop for 5 minutes in Tours (the only stop) on its way to Paris. (OTOH, did it really matter?) Further, the whole issue is kind of irrelevant when comparing the different contexts: Shinkansen stopping trains serving the Tōkaidō corridor, one of the densest developed regions on earth.
As to EMU v LMU, of course it depends on motive power and weight.
TGV Duplex: Power-to-Weight = 23 kW/t;
Shinkansen N700 latest series: Power-to-Weight = 23.88 kW/t
Identical, near as. So they have near-identical acceleration capabilities. Probably for the reason that it approaches what humans or safety authorities consider desirable, or comfortable and/or safe. However the TGV-M is supposed to achieve 20% better energy efficiency but I don’t know if PtW is changing. Moot point.
I think this falls into railfan obsession over something that either isn’t quite as important as implied or isn’t really true; those duplexes have foyers that make preparation for alighting easy. And then the networks have very different characteristics: Japan with almost its entire population along that one corridor while France has its large cities at the farther edges of the hexagon with only light population in between. Forcing everyone, the aged and disabled, with luggage or babies in prams etc, to alight and board in 50 seconds is something you do only if you really have to (eg. lots of stops, huge population density).
Well yes, that’s the point and the reason why Paris-Marseilles trains don’t stop in Lyon (if they stopped at a Lyon suburban station, eg the airport, that would incur an even bigger penalty on those Lyon travellers). And why those bypasses of smaller towns between Bordeaux and Tours reduced the Bordeaux-Paris travel time from 3h to 2h. And why Montpellier is likewise being bypassed by trains heading to Perpignan and Barcelona. It wouldn’t matter if TGVs could achieve N700 shortest possible stop penalty (and I’m unconvinced by anything here that they couldn’t if they needed to) the time penalty for some of these stops, especially Lyon, would be very high, not 2.5 minutes! Choices and compromises had to be made. It seems to work for both Marseilles and Lyon. In the case of Paris to Toulouse apparently all trains will pass thru and stop at St Jean (ie. central Bordeaux) which will incur a fair time penalty but that is because it is a rare situation of a lot of traffic between those two provincial cities (Toulouse & Bordeaux) as well as connections to Spain.
If you look at e.g todays Paris-Barcelona services then there is a 5 minute stop at Montpellier and 10 minutes at Perpignan.
In Britain on the west coast there is a 4 minute stop at Preston and 2 minutes elsewhere.
No, 139 s for the sum of acceleration and deceleration.
The TGV Duplex indeed has about the same power-to-weight ratio as the N700, but that only matters at medium speed. At low speed, the initial acceleration is limited by adhesion, and there, it matters what proportion of the weight is supported by driven axles (14/16 for Shinkansen, 4/13 for the TGV).
You’re saying a train can go from zero to 300kph in 69 seconds? I’m willing to be convinced EMU could accelerate faster but these figures seem a bit fantastical. And it is not what Hood said, and which was in the citation I gave earlier: “it takes just under three minutes for a shinkansen to get from 0 to its maximum speed”. Here is Wiki:
And that is 270 not 300kph. (Acceleration decreases with higher speed so it is important to specify the top speed being used. I don’t think this is true of maglev which can continue with the same accel to top speed.) As I understand it, and suggested earlier, this is close to what is considered comfortable for passengers. That’s why the decel rate is similar and why maglev does similar even though it could easily exceed wheeled trains (in acceleration).
Yes, but it is also affected by the weight on the traction wheels which will be higher for locomotive axels. But ok, the Shinkansen probably has the highest acceleration. The best of Chinese HSR begins at 0.46 m/s2 but drops off a lot above 150kph. I can’t find TGV data. Apparently the Siemens HSRs have terrible acceleration. On long routes with few stops the acceleration rate won’t be critical. It only becomes a significant factor with lots of stops and shorter distances between stops.
It’s certainly tough to believe France could do better than 2 minute stops at smaller places and 5 minute stops at larger places. We have more baggage in Europe than Japan.
And the TGVs are bilevel, and there’s never level boarding in Europe even on the single-deckers, and there’s one door pair per car rather than two, and the doors are often at a vestibule separated from the rest of the car.
On Amtrak, there’s level boarding on nearly all of the NEC, and two door pairs per car, but the doors are narrow and the vestibule separation from the rest of the car is annoying. That said, level boarding solves a lot of problems and a Velaro Novo with two door pairs per car can do 1-minute dwells just fine.
Britain has nearly level boarding and still has 2 minute stops at express stops.
Wilmington will be cheap to bypass. If the train is slowing down for the really tight curves on both sides of Wilmington actually stopping isn’t much of a penalty because it would be going really slow to get through the tight curves. On either side of the station. That are very curvy, some might say “tight”. That trains have to go through slowly. Someday far far in the future when there are so many trains it’s time to consider realigning service patterns the trains that are stopping in Wilmington aren’t slowed down much by the almost 90 degree turn, on either side of the station, because they will be stopping. The ones that aren’t stopping in Wilmington can be on a relatively cheap bypass. One that would be much cheaper than tearing down wide swaths of downtown Wilmington.
I don’t know precisely how much time slowing to gracious speed through Newark would save. Slowing down to …. 30….. takes less time than stopping. Not opening the doors saves time. And getting back up speed when the train is already at 30 takes less time. People in the 23rd Century, when a bypass of Philadelphia is being considered, can consider bypassing Newark. I’m imagining something out along the NJTurnpike that frees up capacity through Trenton for the trains coming in from the Midwest. We’ll all be dead so it’s just idle speculation
To paraphrase the Senator, 139 seconds here, 139 seconds there, it adds up. But that’s okay when you are running more than one train, some of them can be faster than the other ones.
Yeah, Wilmington can be bypassed, relatively straightforwardly; the time gain would be 3:20 relative to making the stop (and skipping on the current alignment is about half of that). At this point, it boils down to precise timetabling on the Penn Line, where intercities need to be spaced evenly every 15 minutes south of BWI in order to avoid needing overtakes of electrified future commuter trains. The faster trains should be distinguished from the slower trains by a single stop, padded to be a 5-minute penalty, and the options are Wilmington, BWI, or Trenton. Two stops together are more than 5 minutes and I don’t see the point in padding the express trains instead. We’re looking at BWI as the main one, purely because Wilmington requires some nontrivial if low by Connecticut standards capital spending and BWI is basically free to extend to four tracks with island platforms.
Um, um, um. both commuter trains can be on the local tracks. It can very easily be four tracked in the places it isn’t already four tracks wide. Imagining that solution doesn’t allow for extensive railfan navel gazing and lamentation about scheduling headaches.
@Matthew Hutton
2 minutes may be a good dwell time for express stops, but the stops I was proposing eliminating are secondary stops like New London not should-actually-be-express-stops like Philadelphia. The passenger volume is much lower at those stops so dwell times can be shorter than the busier stations.
@Michael
So, strictly interpreting what you wrote it will be 139 seconds for both decel and accel, thus 278s already 4’38” plus say 1m for dwell time = 5’38” (5.6 min)
The big problem in your analysis is that you are taking the total time for decel and accel and counting that as the penalty. The penalty is only half the total decel & accel time because the train is moving forward the entirety of decel and accel, while a not stopping train takes time to cover those distances, it does not teleport from the start of decel to the end of accel with no time taken. So Alon is not saying the train decels and then accels in 69 sec each, he is saying they do so in 139 sec each, leading to a penalty of 139 sec plus dwell. Thus based on Alon’s figures the penalty would be 139+60 = 199 sec or 3.33 min.
Also, I was assuming 200kph trains in my 2.5 minute estimate, not 300kph. I’m not sure much if any of the NEC will see 300kph speeds.
There is some fuzzy math in your various sources. If Hood says it takes less than 3 min to top speed and only 50 sec at stop then for an 8 minute stop penalty the deceleration time would be ~11+ min, which is too long (see above on time taken to make the stop vs the penalty compared to a moving train) Theoretically in modern trains acceleration and deceleration should be equal and limited by the force experienced by passengers not the capabilities of equipment. I found multiple sources which state that maximum in service deceleration is either 1.1m/s/s or 2.75mi/h/s which equate to 4 or 4.4 km/h/s, which would get from 300 to 0 in 70-75 sec. This seems too fast, I’m guessing trains ease into the maximum braking force for passenger comfort and perhaps ease out of it too for final positioning at the station, but Alon’s 139 sec figure seems reasonable.
Also, if the N700 has an acceleration rate of 2.6km/h/s then reaching 270km/h would be 270/2.6 or 104 second or 1.75 minutes not three minutes. Again, there may be reasons why the trains are intentionally accelerated slower than this capability.
Note that your Hood quote says “shinkansen on many lines are limited to an acceleration of about 1.6km/h/s” (emphasis added). Without knowing what the “many lines” are referring too I’m not sure your assertion that because Hood is an expert we have to accept his 8 minute figure is correct. It could be the 8 minutes penalty is on mini-Shinkansen lines with slower accel/decel due to legacy track geometry, or certain stations on other lines, etc. Note that even at 1.6km/h/s accel time to 300 kph would be just 187 sec or a little over three minutes, better than the “3 minutes to 270 kph” from the wikipedia source.
Actually I think that is shaky maths. What I said is that one needs to take the accel + decel + dwell time and subtract from it the time that a non-stopping train would take to traverse the same distance (from when the stopping train begins decel to when it achieves the top speed or close to it, presumably 270 instead of 300 because that last bit of speed takes disproportionately longer). That is the time penalty in stopping. I don’t think your statement above makes any sense. There is no reason to suppose 50% like you have. And I think Alon did mean that.
If you have used 200kph that makes a bit more sense but it is barely HSR. I disagree with the view that it defines HSR; it is done purely because British intercity trains achieved that–125mph–more than 60y ago.
Do you know what the acceleration rate of TGVs is? say TGV-Duplex or TGV-M/Avelia Horizon?
I would definitely count Wilmington (or Oxenholme Lake District ) as an express stop
Great big long stretches of it once they replace the wires that bounce around too much above 135 mph. That are already straight. Especially the new right of way for the bypass of the squiggles to every former fishing village in Connecticut.
@Michael
“Actually I think that is shaky maths.”
No, the decel + accel taking twice the time it would take if you didn’t stop is exactly how it works. Graph speed on the y-axis and time on the x-axis – the distance you cover during that time is the area underneath the speed line, for constant speed this is a rectangle (speed x time = distance). If your de/acceleration is constant than the area underneath your speed line is a triangle with half of the area of a rectangle with the same amount of time. In other words while decelerating at a constant rate you cover half as much distance by the time you come to a stop as you would if you maintained your starting speed, and the same for accelerating back up to speed. That means a train stopping and starting takes twice as long to cover the distance you get back up to starting speed as a train not stopping. That means the time penalty is half of the total time spent de/accelerating, exactly as Alon said.
1. Alstom says the Avelia Horizon has an initial acceleration rate of 0.4 m/s^2 but I don’t believe it; for a train with this proportion of its mass on driven axles the maximum is likely lower.
2. Acceleration and deceleration rates aren’t constant. Beyond low speeds, they are limited by train power, based on the formula acceleration = (power/weight ratio)/speed. At high speed, air resistance becomes significant as well.
3. 139 s is for a top speed of 300 km/h. The Velaro Novo performs significantly worse, though still better than what I think a TGV does with its lower initial acceleration.
@Michael on the Shinkansen the fastest non stop services from Okayama to Shin Kobe take 31 minutes, the ones that only stop in Himeji take 35/36 minutes and the first Kodama of the day which doesn’t appear to be overtaken stops 3 times and takes 45 minutes. I would say from that that shows in service the Shinkansen is managing 4.5 minutes of delay per stop.
Eurostar service from London to Brussels is 1h57 non stop and 2h4 with a stop in Brussels.
So with 10 stops the Shinkansen would have 45 minutes of delay and the Eurostar would have 70 so we are arguing over 25 minutes of delay with 10 stops or ~50% extra time.
Also some of the extra time is unavoidable because of more leisure travel and Asian luggage services. Also the Europeans need extra time to catch up from small delays that the Asians do not.
Stop in Lille not Brussels.
My fantasy is that the moment Gateway opens they take at least 9 through 14 out service and double deck them. So that someday, probably in the 23rd Century, they have the capacity for the high speed trains.
Excellent post. Two quibbles:
It seems that 12-car lengths are doable with minimal platform length extensions, which is probably enough for the medium term.
Unless Amtrak decides to open up Acela for non-business class bookings, it’s unlikely the trains will be full. The Acela II trainsets will be 9 car lengths, which is already a 55% increase in capacity if frequencies don’t change, and there are provisions to extend them further to 12 car lengths. Then capacity can be doubled again (~4x total) if Amtrak eventually goes to 30 minute headways between DC and NYP, but one could argue that at current price points, the slots should go to additional Regional trains.
For the current Amfleet services, the Airo trainsets should be 12 cars instead of 8, which represents a 50% increase in capacity (100% for the 6-car Keystone trains). These should hit their capacity sooner than Acela, but a feasible solution should be to designate Airo for only the services that run outside electrified routes (i.e. Virginia, Hartford Line), and use Acela II trainsets for all service between DC and BOS. Acela will naturally have more capacity because of shorter travel times due to fewer stops and higher speeds.
Without looking at timings, the Airo trains can handle all the existing NER train stops. The Acela II trains should drop Metropark to avoid the switches there, but I do think the non-superexpress Acelas should be stopping at Trenton and Stamford to accommodate transfers to local services. Adjust prices so that ridership is maximized on the Acelas.
That would make sense so they won’t do it.
The Congressionals were 18 cars long. There were others.
Squint at Newark, really hard, there is a lot of future proofing. Rumor is that it’s track A because that can be converted into an island platform for track B. I fantasize that the Wall Street trains are there. Move PATH to the upper level and use the current PATH tracks for the Grand Central Shuttle. Squint at Secaucus there are many many ways it can be reconfigured. I’m not sure where the Carnarsie/L trains should go.
Aren’t the moveable bridges just outside of Newark Penn not going to disrupt your plans there? Or are you assuming they will be removed/rebuilt? As for reconfiguration at Secaucus once Gateway is built reconfiguration East of Secaucus gets way more complicated. How are you planning to get the Grand Central Shuttle from Newark to GCT…
If there are lot more trains going to Penn Station Newark it would need more bridge. More track in Harrison and points east too. By the time there are new tunnels under the Hudson the existing bridges will be ancient. A new one that connects Track A and the new Track B, on the lower level and PATH to the new island platform on the level above the mainline tracks. Which gives enough capacity to tear down the existing southern Dock Bridge while it gets rebuilt. Which gives then enough capacity to tear down the equally ancient northern Dock Bridge.
In a tunnel? Clueless railfans don’t grasp the scale and have the urge to have all the trains stop at all the platforms. There are enough people who want the East Side of Manhattan for there to be dedicated tunnels for the suburban market. . . . Like there are now for the LIRR. . . . To avoid building deep caverns to satisfy clueless railfannery, extend East Side Access train….in a tunnel… to New Jersey. Non-stop. Because New Jerseyans who want to go to the East Side don’t need to stop on the West Side.
To keep the arithmetic simple 18 an hour, during peak hours, is bit over every three minutes at Secaucus, every ten minutes at Broad Street Newark and every five minutes at Penn Station Newark. It would likely need new bridges for Broad Street too. Off hours they can do the much vaunted timed transfers to make it almost as painless as it is during peak.
That shuttle to GCT through a tunnel is not gonna work, getting into the south end of GCT from Park Ave is immensely challenging and hence costly…… extending the 7 line or similar to NJ would have been a more feasible option from a pure engineering perspective…..
@Tunnelvision
Adirondacker is not proposing extending a GCT shuttle from the main portion of GCT (upper and lower levels built by the NY Central in the early 1900s). Instead he is proposing extending the Long Island Railroad tracks from the station built below GCT for East Side Access. That underground station is VERY deep and well below any conflicts with the subway or other infrastructure underneath Park Ave.
The tunnels for East Side Access trains extend well south of 42nd Street. And are very deep. Beyond that, since they wouldn’t be stopping on the West Side, could be almost anywhere.
Yeah but no. We looked at extending ESA to connect with what was then ARC years ago, it is not as straightforward as it seems. City Water Tunnel #1 is a potential conflict, depending on the alignment that you take, assuming the idea is to connect Penn to GCT. You’d also have to rebuild the ventilation plant at 37th St and remove the buried TBM between 36th and 37th St. Possible yes, but highly unlikely.
what’s the alternative? something much more expensive or something much more stupid like extending the Flushing line?
City Water Tunnel #1 is a potential conflict, depending on the alignment that you take, assuming the idea is to connect Penn to GCT.
Adirondacker specifically said a connection to GCT that would not stop on the West Side (since existing services already exist to get from Newark to Penn).
You’d also have to rebuild the ventilation plant at 37th St and remove the buried TBM between 36th and 37th St.
If you are discussing extending the ESA tunnels to connect to ARC/Gateway or to a brand new route to Newark, then things like rebuilding a ventilation plan or dismantling a buried TBM are irrelevant issues, not things that make the plan very unlikely. At the scale of cost to do this kind of work the added cost of taking apart a TBM or surface rebuilds are not deal breakers.
The reason Access to the Region’s Core was not going to continue to Grand Central when it opened in 2017 was that Water Tunnel 3 was not in full service. The city, very rationally, asked that there not be any digging near the tunnel, Water Tunnel 1, that has been in continuous service since 1917. Without ever being inspected much less repaired. Water Tunnel 3 is supposed to wrap up the decades long project soon. Water Tunnel 1 can then be taken out of service for inspection and repair. And while it’s out of service, digging near it could be done. Which is an interesting tale because the project was canceled. Railfans turned that into the dastardly city isn’t letting anybody dig anything anywhere forever.
Well, that’s a tour de force and absolutely the kind of programmatic thinking that’s needed.
I want to make sure I understand you right about the Trenton/BWI stagger that reduces frequency from 10 minutes at New York to 15 south of Baltimore. I think you’re saying the service patterns could be as follows:
:00, A, Boston to Washington, “Express” (i.e. no stop in BWI)
:10, B, Boston to Washington, “Local” (stops in BWI)
:20, C, Boston to short turn (does not continue past Baltimore)
:30, A
:40, B
:50, C
Resulting in A, B, A, B on 15 minute spacing south of BWI.
You probably wouldn’t really designate pattern A and B trains too separately, except in a more minimal format for BWI-bound passengers.
Realistically, no train is going to terminate in Baltimore. However, 6 Acelas per hour to DC isn’t feasible without an expensive quad-tracking in Maryland, hence 4 tph south of Philadelphia.
In an optimal plan with 16-car trains and rebuilt stations / platforms, you’d have 3 types of service on a 30-minute takt:
Super express would be:
DC
maybe Baltimore
Philadelphia
New York Penn
maybe New Haven
Boston South
The other stops are too small for a super express.
I was thinking of a standard Hikari superexpress that stops in every state. But if you want a Nozomi superexpress, then the only stop between DC and NYP should be Philly, which Amtrak did operate pre-COVID. Ridership on the future New England HSR would be more like the Sanyo than the Tokaido Shinkansen, so realistically the superexpresses would still be stopping at New Haven and either Providence or Hartford / Springfield / Worcester.
Baltimore is the 6th/7th/8th (depending on year) busiest station in the US, with ridership closer to Boston than Boston is to Phila. or DC. It absolutely gets an express/super-express/Nozomi stop.
Same for New Haven, which is 8th-10th busiest, although more for its position as the merge point for the Harford/Springfield branch than absolute ridership (that it’s ridership is as high as it is is no doubt a result of this rail geography).
Hikari aren’t superexpress. They make more stops than the Nozomi. The super express would make less stops than the express. Which is the Nozomi.
There aren’t enough stations in New England to have four different stopping patterns. Boston is not Tokyo, Providence isn’t Yokohama, etc. And there aren’t enough people in New England to have a dozen trains an hour to have four different stopping patterns.
Decide what you want New England to be. Some place with so few trains people have to take a shuttle to Hartford or someplace where people are smart enough to book the train that gives them a one seat ride from Baltimore to Springfield.
“And there aren’t enough people in New England to have a dozen trains an hour to have four different stopping patterns.”
You can have three stopping patterns with 4 tph: 2 tph express/Nozomi, 1 tph limited/Hikari (approx. current Acela pattern), 1 tph regional/Kodama (all stops).
If some of the trains are going via Springfield and some of the trains are going via Providence that’s two. The faster route is going to be along the Shore Line because that will be cheaper to upgrade to very high speeds. You don’t have enough train or stations to have three different services along the Shore Line. And you don’t have enough trains because Providence isn’t Yokohama and Boston isn’t Tokyo.
And you don’t have enough trains because Providence isn’t Yokohama and Boston isn’t Tokyo.
Metro New York (19.5M) is the same size as metro Osaka (19.3M), metro Boston (4.9M) is close to metro Fukuoka (5.5M). If you use CSAs then NY (21.3M) and Boston (8.3M) clearly pull ahead (I’m not sure which is more comparable to Japanese Major Metro Areas). Osaka-Fukuoka is 554km, NY-Bos is 372km. The Sanyo Shinkansen has the Nozomi, Hikari, and Kodama, plus the Mizuho (like Nozmi but starts in Osaka and continues to Kyushu) and the Sakura (like Hikari but again Osaka to beyond Fukuoka) for five different stopping patterns at 7-8 tph.
I think if we build HSR that connects Boston to NY in 2 hrs or so there will plenty of riders to justify multiple services to Boston.
If you don’t think you will have the passengers you can start with 4-5 car trains anyway.
All of New England is 15 million people. You don’t want me to do the cruel cruel arithmetic with numbers not rounded up to make it easier. New Englanders want to go other places that don’t involve passing through New Haven and that makes the cruel arithmetic even crueler.
The people on long standing room only trains won’t be pleased that clueless railfans want to throw away capacity.
Okayama, Hiroshima and Fukuoka (the metro areas south of Osaka along the Sanyo Shinkansen) is 9.8M people; the metro areas north of NY along the NEC (NH, Harford, Springfield, Worcester, Norwich/N London, Providence, Boston) are 9.9M. Cruel arithmetic says the population is equal, and since the distance is shorter the population is denser than along the Sanyo line, so similar service is possible.
Japanese in southern Honshu want to go places that don’t involve passing through Kobe, and yet they still ride the Sanyo for the places that do involve it.
“Throw away capcacity”?!?! Which is it, New England isn’t large enough for multiple stopping patterns and more of them want to visit Albany and Buffalo than NYC or Philadelphia, or the trains will be standing room only?
It’s hard to see things when Stamford is so dazzling.
Metro Detroit is 700 miles away from Boston. People who live west of Brookline might find taking the train to Detroit intriguing. I’ll see your Baltimore, and raise you Detroit. Depending on how many buses there are to places like Ithaca and Binghamton there are between 4 and 6 million in “Upstate” New York which includes most of Vermont. I’ll see your Philadelphia and raise you Upstate. Apparently Alon hasn’t bothered to look at the satellite images of the Trenton Transit Center. There are two island platforms separated by four tracks. The shovel-train-loads-of-people-to-Boston train could be passing the Springfield-Harrisburg train that is actually stopped in Trenton. I’ll see your Trenton and raise you Syracuse. How much of Southern Ontario and Quebec gets included? Alon seems to think crossing the border greatly reduces demand. I see the Quebec license plates on I-87 and get the feeling that he thinks Montreal is London. The only place to go from Montreal is Toronto or places in the States. Same for Toronto. Cleveland-Boston! Since the tracks and the trains will be there anyway people in Cleveland will be getting to New England through Utica. Utica is a bit west of where it gets realllly realllllly really flat. So flat the railroad tracks are really realllly really straight. And 350 kph trains will be really cheap to run. I’ve run out of Mid Atlantic and Maryland metropolitan districts. I’ll see your Harrisburg and raise you Cleveland. 3,000 an hour, across Long Island Sound, was rounding up twice very very generously.
The distance is shorter so there are less stations to skip. What part of that didn’t cross your addled brain? The Nozomi skips not-Providence, not-Back-Bay , which of the remaining 3 get skipped to differentiate between Hikari and Kodama?
I don’t know how they warp the space time continuum on your planet to get from Old Saybrook to Cornwells Height ( Pennsylvania ) without passing through Queens where there are long standing room only trains or New Jersey where there are different long standing room only trains. Which one of those do you want to cancel so there can be short trains from New England passing through Metuchen? (New Jersey, pronounced Meh touch en ) Those standing room only trains. I don’t know how popular New Haven Line Penn Station Access trains will be. I suppose they could cancel some of those instead of LIRR. Go ahead, tell me which long standing room only train you want to cancel so there can be more short stubby little trains for New England
It’s a pity when clueless railfans only think of one thing at a time. You really have to sharpen your crayons. I’m going to suggest the box of 24 instead of the single one you apparently have.
Metro Detroit is 700 miles away from Boston. People who live west of Brookline might find taking the train to Detroit intriguing.
No they won’t. 700 miles away means a 5 hr trip even if you build extremely fast (320-350 kph) track all of the way. At that amount of time people will fly. Build track in the 250-300 kph range and it will take even longer and people will fly even more. Building 700 mi of super-high speed track to Detroit would be stupid, given how much fewer people live along those 700 mi versus along the 456 mi Bos-Wash corridor. Also, you only get 700 mi to Detroit from Boston by cutting across Canada. Stay south of Lake Erie to serve Cleveland and its 800 mi, so add another hour or more to your time and watch your ridership shrink even more.
I’ll see your Baltimore, and raise you Detroit.
Baltimore is 400 mi from Boston. People will ride a high speed train that distance because it is faster than driving, they will not do so for 700-800 mi. Between Balt and Bos are 29.6M people (only on the current NEC route), between Bos and Detroit are 7.4M. Even including the cities at the ends you are looking at 37.3M versus 16.6M. you do realize that more people who are closer generate more ridership than fewer people farther away, right?
Depending on how many buses there are to places like Ithaca and Binghamton there are between 4 and 6 million in “Upstate” New York which includes most of Vermont. I’ll see your Philadelphia and raise you Upstate.
The metro areas along the Empire Corridor are 4.1M (not including Poughkeepsie) versus 6.2M for Phila. Adding Binghamton and Ithaca only gets to 4.4M. Glens Falls, Elmira, etc. gets you to 4.7M. Phila. is ~310 mi from Bos; Syracuse is also ~310 mi, Rochester is 391 mi, and Buffalo is 454 mi. And of course along those entire 454 mi you only pass 5.2M including Worcester, Springfield and Pittsfield, while along 456mi from Los to Wash you pass 38.7M.
If you add anything within 75mi of the Empire Corridor (Binghamton to Syracuse) to the ridership shed of “Upstate” then you should add anything within 75mi of the NEC to its ridership potential. This doesn’t help Upstate in your analysis. Including all of the Nashuas and Worcesters and Allentowns and Dovers within 75 mi of bus travel to the NEC adds 8.5M people, or almost double the number of people in Upstate. Many of these places are closer than Binghamton to Syracuse, and many of them already have rail service to the NEC, let alone busses. Once again, you do realize that more people who are closer generate more ridership than fewer people farther away? (Maybe you don’t realize this…)
Alon seems to think crossing the border greatly reduces demand.
This isn’t something he thinks, it is something shown by the empirical evidence of international HSR lines in Europe.
Cleveland-Boston! Since the tracks and the trains will be there anyway people in Cleveland will be getting to New England through Utica. . . I’ll see your Harrisburg and raise you Cleveland.
Cleveland is ~640 mi from Boston. People will not go to New England via Utica, they will fly. There are 2.2M people in Cleveland and 1.2M in Harrisburg/Lancaster. The latter two are 200mi from NYC, which has 19.5M people in the metro area, compared to 640 mi from the 4.9M in metro Boston. Harrisburg-NY will generate more ridership than Cleve-Boston. Once again, you do realize that more people who are closer generate more ridership than fewer people farther away? (Maybe you don’t realize this…)
3,000 an hour, across Long Island Sound, was rounding up twice very very generously.
Please tell me you are not seriously arguing that the Midwest will generate more traffic to Boston than the 35+M people from NYC to DC will? Once again, you do realize that more people who are closer generate more ridership than fewer people . . . oh forget it.
The main subject of the thread is Boston to D.C. Many many times, over the years the discussion has been trimming back the 15 high speed rail routes railfans imagine emanating from Boston. It can be trimmed down to two. Which gives almost everybody acceptable amounts of service. One vaguely along Interstate 95 and another one along Interstate 90. Which has been discussed extensively over the years. Including, briefly, in these threads. It’s not my fault you aren’t paying attention.
…. you do understand that every person heading from Springfield towards Albany is a person not heading from New Haven to New York? Perhaps on your planet the peculiar behavior of space/time allows that. In these parts it has to be one or the other, it can’t be both.
Box of 24 crayons would be helpful.
Did some digging in the Bureau of Transportation Statistics data on airline origin and destination numbers for 2024.
From Detroit to Boston and Boston to Detroit combined there were 672,968 air passengers in 2024. This includes subsidiary airports like Bedford and Pontiac in the same metro areas. Baltimore saw a little over 1.3 million Amtrak boardings and deboardings in 2024.
Total air passengers to/from all upstate NY airports along the Empire Corridor (including Saratoga Springs and Ithaca but not Plattsburgh, Lake Placid, etc.) and Boston was 277,991 combined. Existing rail ridership along this corridor was 1,296,000 in 2024 (almost all to NY not Boston). Philadelphia 30th street saw just over 5M last year.
Total air passengers to and from Cleveland to Boston was 225,771 in 2024. Based on your 365 days/year and 15 hr service days this works out to 41 per hour. As you point out math is cruel, and your very generously twice rounded up 3,000 per hour is much, much higher than 41 per hour.
Add all three (Detroit, Upstate, Cleveland) together and you have 1.18M total air travel to Boston. You will never get all of those people on a train because everything beyond Syracuse is 2+ hours even at incredibly expensive almost best-in-the-world speed (faster than any Shinkansen, or Madrid-Barcelona, or anything in France). At more than 2.5 hours no HSR line gets 100% mode share vs air. The mode share for Cleveland/Detroit would be in the 15-40% range even at incredibly fast speed, at “normal” HSR speeds (Tokaido Shinkansen, Taiwan HSR, LGV Nord, etc.) you are looking at 10-25% mode share.
Of that 1.18M total air travel, you would likely only get 250-450k annual riders from those Upstate/Midwest metros to Boston. To get this requires ~1300km of HSR route; to get the high end figures requires all of that track to be 350kph, so very expensive. By comparison, bringing the ~250 km from New Haven to Boston via Hartford up to 200/250kph (so basically the same as the current Acela route via Providence) should add about a million passengers per year (extrapolating the ridership per capita of the Providence & Norwich MSAs to the population of the Hartford, Springfield, Worcester MSAs). If you also improved New York-New Haven to make travel times even better than right now would increase ridership even more. Getting a million new passengers for the cost of 250km of medium speed track is much better deal than less than 500k new passengers for the cost of 1300km of very high speed track.
Again, I realize this requires two, three perhaps even four crayons, anything north or west of Albany will be there for the New York ( and beyond ) market. … Senator Gillibrand catches the 5:00 only-stops-in-Philadephia she could be home in two and half hours…
Spending bazillions to make New Haven to Springfield realllllly realllly fast because railfannery still means the part between Springfield and Boston has to be built. Modest amounts for New Haven to Springfield is good enough. Spend the money to get New Englanders someplace other than the NEC. They won’t be going through Navel of the Cosmos, Center of the Universe, Axis of the Galaxy that is Stamford but they will get over it by the time they get to Montreal. Or Toronto. Or Buffalo. Or a lot of other places not the Northeast Corridor. If they are going through Albany, even though it won’t be quite as popular as going through New Haven, they can’t be going through New Haven. Not with how space/time works here anyway.
It crosses my mind that perhaps why you think people can get from Framingham to a gate in Logan instantly, is because that is how space/time bends where you are. Here it takes time to get from Framingham to Logan. And it takes time to get from the curb at Logan to when the plane pulls back from the gate. Five hours from the railroad station a short walk or cab ride away, in Framingham, to Detroit, looks real good.
Right at this moment, which could change any second, and likely is inaccurate because Southwest doesn’t share nice, there are 8 non-stops between Boston and Detroit. And the service is clumpy. there is a five hour gap between noon and five. When there are three flights within 28 minutes. I can imagine a train every other hour to Detroit alternating with every other hour to Toronto. Leaving Springfield at 1 works a lot better when there is a train every other hour. The train to Montreal can alternate with the train to Cleveland. It’s seven hour gap between the three non-stops from Hartford.
New Haven to New York is a sticky wicket. Stamford is in what was Fairfield County with a population of just under a million. Long Island is 8 million. With a nice straight right of way the government in one way or another owns. That when the commuter rail crayons come out can serve Wall Street, the third or fourth biggest business district in the country depending on who is counting what. And a connection to the 2,3,4,5, B,D,N,Q and R trains in Brooklyn. That makes most rail fans melting brains dribble into a quivering heap. So does that there can be a once an hour train that goes between New Haven and the Bronx, very cleverly weaved in between the express commuter trains, that gives people in Fairfield and southeastern Westchester service to other places on the NEC.
I suggest using the crayons that are blues and greens for the high speed routes and yellow, orange brown and reds for the commuter routes.
Again, I realize this requires two, three perhaps even four crayons, anything north or west of Albany will be there for the New York ( and beyond ) market. … Senator Gillibrand catches the 5:00 only-stops-in-Philadephia she could be home in two and half hours…
If the market N & W of Albany is for New York, then why did you write so much about Cleveland to Boston? Connecting Albany to DC does not requires an HSR line from Albany to Boston.
But much more unfortunately for your argument is that N & W of Albany to NY is still a smaller market than New England to the rest of the NEC.
When it comes to air travel, there are 10.7M pax per year between airports on the NEC. 8.3M are going from N England to somewhere south, of which 6.3M are going from Boston. That means Hartford, Providence and New Haven are seeing 2M pax per year to DCA, BWI and Philadelphia. Total travel from Detroit, Cleveland, Pittsburgh and any upstate NY airport to N Eng or New York is 7M. Cleveland and Pittsburgh see 1.1M and 1.2M, which means travel from Prov/Hart/New Haven is more popular than from Cleveland or Pitt to New York AND Boston (!). It’s almost as popular as Cleveland and Pitt combined, Travel from Philadelphia to Upstate adds another 0.4M, but again does not need a line east of Albany to do. Forget connections from Phila and beyond to Cleveland or Detroit, no one is going north and east to later go south and west.
As noted before, you can a lot of the people flying in the Northeast onto a train because it is possible to get travel times in the 2-3 hr range where rail mode share is high. You will capture very few of the travelers from Buffalo and beyond because the travel time will be 3-6 hrs, where HSR mode share gets very low.
Using highways, I-95, CT-15 and I-684 see about 234k cars a day on them. This is roughly traffic from New England heading to the rest of the NEC. Combined traffic on I-90 leaving Mass, I-87 at Newburgh, and I-84/80/78 leaving the NY Metro area to the west is 239k daily average. That means that travel from New England to the NEC is about as large as from New England and NY to the rest of the country.
Spending bazillions to make New Haven to Springfield realllllly realllly fast because railfannery still means the part between Springfield and Boston has to be built. Modest amounts for New Haven to Springfield is good enough. Spend the money to get New Englanders someplace other than the NEC.
Making NH to Springfield, or even NH to Boston via Springfield, really fast will cost less than doing the same for Springfield to Albany, let alone Springfield to Detroit. NH-Spring is shorter than Spring-Alb, and the terrain is a lot easier. It is much, much shorter than any line to the Midwest. Modest amounts of improvements are not good enough, because there are more people who would use the Hartford link, and because making it very fast provides a huge benefit by bringing travel markets into the high use range. Spending money to take N Englanders elsewhere is much poorer from a cost-benefit because you will spend more money but get fewer riders – there are fewer of them to begin with and at longer distances they will continue to fly.
Here it takes time to get from Framingham to Logan. And it takes time to get from the curb at Logan to when the plane pulls back from the gate. Five hours from the railroad station a short walk or cab ride away, in Framingham, to Detroit, looks real good.
No it doesn’t. In the real world there are HSR trains running right now that take 5 hours for their journey, like Tokyo-Fukuoka. They get 5-20% of passengers vs air, people take the plane instead. And Framingham is a terrible example because it is not in the center of Boston but right next to the rail line. A train from there might be better than Logan (but might not, see 5-20% above), but for the rest of the Boston metro area it will be faster to get to the airport and have a 2 hr flight than get to S Station and have a 5 hr train. Plus 5 hrs is only if you build a really expensive almost-best-in-the-world-speed rail line, more practically Detroit would be 6-7 hrs away. And if it stopped in places like Framingham instead of being an express it would be even slower, which means even Framingham has to central Boston first, at which point Logan isn’t much farther.
Right at this moment, which could change any second, … there are 8 non-stops between Boston and Detroit. … I can imagine a train every other hour to Detroit alternating with every other hour to Toronto… The train to Montreal can alternate with the train to Cleveland.
Nothing you can say will ever change the fact that more people fly from Providence and Hartford to Baltimore every year (813k) than from Detroit to Boston (672k). A million people fly from Cleveland to New York and Boston combined each year; Providence sees 1.1M air travelers to other NEC destinations. Toronto? Montreal? There are 25,032 cars on I-190 when it crosses the border in Niagara heading to Toronto, just 6,860 on I-87 at the border on its way to Montreal. I-91 south of Springfield is used by 132k cars daily.
New Haven to New York is a sticky wicket. . . Long Island is 8 million. With a nice straight right of way the government in one way or another owns. That when the commuter rail crayons come out can serve Wall Street…
I agree that intercity service to Long Island via Wall St. is a great idea. Using it to get to New England is not, the cost of the bridge across Long Island Sound would be huge.
So does that there can be a once an hour train that goes between New Haven and the Bronx, very cleverly weaved in between the express commuter trains, that gives people in Fairfield and southeastern Westchester service to other places on the NEC.
Um, these trains already exist. The Northeast Regional stops in New Rochelle, Stamford and Bridgeport in Westchester and Fairfield counties, and goes through the Bronx on the same tracks as Metro-North commuter trains. This fact confuses no one.
I’ll use small words. People in New England will be able to use the same tracks, north of Albany as people south of Albany use, to get to Saratoga Springs and Montreal. And people in New England will be able to use the same tracks west of Albany as people south of Albany are using to get to Schenectady and beyond . For instance the train from New York to Toronto can leave Albany at :00 and the train from Boston to Chicago via Detroit can leave at :10. The train for Washington to Montreal can leave at :20 and the train from Boston to Saint Louis can leave at :30… I know this is deeply deeply disappointing but none of them involve Stamford. Or, don’t clutch at your pearls too hard or fast, Connecticut.
This may come as surprise to you but they correlate with raw population numbers. It’s much easier to guesstimate things with population numbers and that the Japanese, in very round numbers, average out to one round trip per person per year. There are enough people in Southern New England close enough to large other populations for there to be TWO routes emanating from Boston.
UPS delivered something much more interesting than your cherry picking and obtusenees. I’m going to unpack that and let the cats investigate the box.
@Onux, international links in Europe are damaged by high fares on Eurostar and terrible service levels everywhere else. Also by the lack of good onward connectivity between the Eurostar and other non Paris destinations.
@Onux, with regards to Tokyo-Fukuoka, traffic is held back by the lack of a true super-express on the Sanyo Shinkansen that does Fukuoka-Shin Osaka non-stop and the extremely high fares. Other countries would charge a much lower fare for such a long trip.
I hope this isn’t a revelation. Canada isn’t in Europe. The big Canadian cities you can get to, by rail, someday far far in the future when Canada builds high speed rail, from Toronto are Montreal, Montreal, Montreal and if you want to stretch the definition of big city, Ottawa. And Montreal. And from Montreal it’s Toronto, Toronto, Toronto, stretch the definition to include Ottawa and Toronto.
The U.S. isn’t in Europe either. When Torontonians want to go some place, other than Montreal , Ottawa if you want to be generous or Montreal they look to the U.S. Travel demand is going to be ….different…
A quarter of the Canadian population is clustered around the western end of Lake Ontario.
https://en.wikipedia.org/wiki/Golden_Horseshoe
Mute demand by half it’s five million people. ….. Roughly the same as metro Boston. Or the tenth largest metro in the U.S. Put too many links in a post it gets moderated. I’ll leave it up to you to look up U.S. metro areas about half the size of Montreal. They are places railfans drool over.
5.4 million passengers on TGV Lyria given the limitations and given Switzerland doesn’t have that many tourists seems decent too?
https://moment.tech/2024/10/14/tgv-lyria-chooses-moment-to-transform-its-onboard-digital-experience/
Seems the Canadians have been stirring their stumps.
https://en.wikipedia.org/wiki/Alto_(high-speed_rail)
I’m betting a wooden nickel that after ( from the Wikipedia article )”A design phase for the project was announced with an estimated cost of CAD$3.9 billion and is expected to last between 4 to 5 years.” it will find that goodness gracious me, the REM has taken over the tunnel to Gare Centrale and it will cost too much. A different wooden nickel that the Ontariains have another fit of Real Canadian(tm)ism and decide that they don’t want nothing to do with communist trains.
Like the Highs Speed Ground Transportation Act of 1965 giving us trains that are as speedy as the ones in 1970 were the Canadians have been studying things very very studiously for decades. It keeps lots of white collar workers, in work. Like “A design phase for the project was announced with an estimated cost of CAD$3.9 billion and is expected to last between 4 to 5 years.” gives lots of college graduates lots of wan.. work.
https://en.wikipedia.org/wiki/High-speed_rail_in_Canada
…. Someday Canadians will be able to get from Montreal to Toronto or from Toronto to Montreal on high speed trains.
To be honest I wouldn’t mind some clarification here too. I thought the service pattern was:
4tph: WAS-NY-BOS (of which only 2tph stop in New London)
2tph: WAS-NY-Hartford-Springfield
With the additional stop in New London shifting the Boston trains from a 10/20 headway to neat 15min headways once the Springfield trains diverge at New Haven.
Does WAS-NY not require 6tph the whole way to Washington? What is the point of terminating at Baltimore? If anything demand would be higher between Washington and Baltimore than between Baltimore and Philly, surely. In that case there is no need for an optional stop at BWI, and in fact it would make the headways uneven.
Many many railfans think Providence is the goal. There are more people along the line through Hartford, Springfield and Worcester than there are in all of Rhode Island. And it is ever so slightly shorter. Arithmetic is a cruel cruel mistress.
The demand, east of Sixth Ave, viewed from Penn Station New York, is Long Island and New England. The demand west of Ninth Ave is the rest of North America. There are going to be trains passing through New Jersey that terminate in New York. Because “rest of North America” has more demand. Arithmetic is a cruel cruel mistress.
The point of really long trains is the capacity constraints between North Philadelphia and New York City. Six trains an hour to New England implies short stubby ones or half empty ones. Send four of them to New England they will be fuller. Arithmetic is a cruel cruel mistress.
“Many many railfans think Providence is the goal. There are more people along the line through Hartford, Springfield and Worcester than there are in all of Rhode Island. And it is ever so slightly shorter. Arithmetic is a cruel cruel mistress.”
Adirondacker is once again correct. For many reasons, the first HSR service on the NEC should be going through Hartford-Springfield-Boston, with the current Providence route as secondary service.
The real question is “given service along one of those, is there enough additional demand to warrant service along the other as well? Or maybe run a spur up to Worcester for these people? Providence-Springfield-Albany looks interesting as well (a quick looks suggest this isn’t too geographically hard, and sending this up to Montreal looks better than a direct Boston to Montreal – but someone who knows the geography better than me should double check that).
I think it’s the other way around. The existing 150 mph infrastructure, plus the political reality of serving Rhode Island, means that it should be noticeably cheaper to build the first HSR along the Shore Line / I-95. Another branch should be built through Hartford – Springfield – Worcester, but the timeline based on higher costs to upgrade the entire corridor, especially through the Boston suburbs, is more like 2 decades rather than 1.
It should, but didn’t Alon once upon a time trash that idea due to the line restrictions on the Worcester/Springfield section? Its quite curvy?
Geography/topology is an even crueler mistress than arithmetic. It’s one continuous very expensive suburb from metro Springfield Massachusetts to metro Wilmington Delaware.
The ROW is okay-ish from Boston to New London. And a bypass of every squiggle that seemed like a good idea in 1840, hugging the Connecticut coast, can be bypassed cheaply – in the vicinity of I-95. Anything else you consider is hilly if not mountainous and would involve tunnels and viaducts. Those aren’t cheap. The destination from-places-not-New-England is Boston and since it should be cheaper and easier to do it along the coast, do it along the coast.
In nice round numbers: Southern New England – Connecticut/RhodeIsland/Massachusetts have 11 million people. Almost as many as the country’s second biggest metropolitan area, Los Angeles. Or if you want to compare it to combined statistical areas it would still be the country’s third biggest. Bigger than Washington-Baltimore. Bigger than Chicago’s. 170 miles/275km of very difficult route between Boston and Albany connects them to the tracks that will already be there for the New York market. Which is also the Philadelphia market. Make it all fast enough it’s the Washington D.C. market too.
It’s only – again in nice round numbers – 60 miles or 100 km from Springfield to New Haven. Hartford is two thirds of the way between New Haven and Springfield. Upgrade the existing so a non stop train from New Haven to Hartford does it in 30 minutes and Hartford to Springfield in 15 it’s not that much different than spending 10s of gazillions of dollars to make 15 and 8. An upgrade is good enough because it’s not that long. It implies 2:30 from Hartford to Washington D.C. instead of 2:15.
New England has many possibilities. Boston to New Haven and beyond and Boston to Albany and beyond is enough. Do Boston to New Haven along the coast because it will be cheaper and easier to build. Amtrak, since it’s creation six years after the High Speed Ground transportation act of 1965 has managed to make Acela almost as fast as the fastest Metroliners were under the PRR. I don’t expect much between now and the time we are all dead.
@henrymiller74
“The real question is “given service along one of those, is there enough additional demand to warrant service along the other as well?”
Yes, absolutely. Hartford+Springfield+Worcester has about 2.4/2.5M people, similarly sized to Austin or Pittsburgh. Providence+New London is about 1.9/2M people, about the size of Austin or San Jose. If you draw HSR maps with service to any of those four cities (hint, Alon does) then you should easily feel justified providing HSR service to both northern branches of the NEC, especially since they are bounded on either side by Boston and NYC at a distance of 100-300 km, which is hugely competitive HSR distance.
@caelestor
“ The existing 150 mph infrastructure, plus the political reality of serving Rhode Island, means that it should be noticeably cheaper to build the first HSR along the Shore Line / I-95.”
The existing infrastructure along Prov-Bos works in favor of making Harford the first new-build branch, not against it. Assume that as part of an NEC HSR plan NY-New Haven will be improved to 40 min travel (as part of 2hr Bos-NY, and compared to approx. 1h40m today). That alone would cut 1 hr off of rides from Prov to anywhere NY and south. So Providence/Rhode Is. benefits a lot from a Hartford route HSR, and with existing 25kV catenary trains can take advantage of it from day 1 with no extra cost. Hart-Spr-Wor by comparison benefits little from a Shore Line/I-95 HSR because Hart-Spr still needs a transfer, there is no service to Boston, and Wor gets no service at all. For Hart/Spr/Wor to benefit requires stringing wire, and if you are going to do that you might as well improve the route to 200-250 kph. Benefiting both Prov and Hartford is better than just benefiting Prov.
“Another branch should be built through Hartford – Springfield – Worcester, but the timeline based on higher costs to upgrade the entire corridor, especially through the Boston suburbs, is more like 2 decades rather than 1.”
New Haven to Boston is something like 250 km via Hartford. China has been building ~3,600 km of HSR per year. The US can do better than 12.5km per year and put NH-Hart-Bos in service faster than 2 decades.
@Tunnelvision
“but didn’t Alon once upon a time trash that idea due to the line restrictions on the Worcester/Springfield section? Its quite curvy?”
Worcester/Springfield is curvy but also very rural, so it shouldn’t be difficult to build a new route for the 35 km or so that is very curvy. Alon thinks the bigger issue is that around Newton the Right of Way is too narrow for four tracks to allow intercity service and commuter service, I address these concerns in posts here:
https://pedestrianobservations.com/2023/05/30/can-intercity-trains-into-boston-enter-from-springfield/#comment-148408
https://pedestrianobservations.com/2023/05/30/can-intercity-trains-into-boston-enter-from-springfield/#comment-148479
https://pedestrianobservations.com/2023/05/30/can-intercity-trains-into-boston-enter-from-springfield/#comment-148476
@henrymiller74
Providence-Springfield-Albany looks interesting as well (a quick looks suggest this isn’t too geographically hard, and sending this up to Montreal looks better than a direct Boston to Montreal
It doesn’t look interesting at all. Those three MSAs combine for 3M people, but that route has no really big city in the league of Bos or NY to anchor it and drive ridership.
It may not be geographically hard compared to crossing the Alps, but between Prov and Springfield is much hillier than the flat coastal plan or river valleys between Prov or Springfield and Bos. Springfield to Albany crosses the Berkshire mountains, which are not huge but will still be much more expensive than other projects which would serve more people.
Why would you send trains from Bos to Montreal via Providence?!?!?! Montreal is North of Boston, Providence is south of it. I agree that a Bos-Montreal direct route is overkill, but if you are going to do Bos-Montreal via Albany the way would be Bos-Springfield-Albany, not going south to go back north to Springfield.
No it doesn’t because the train can get to New Haven and keep on going until it gets to Washington D.C. Like they do today.
No it doesn’t because the train can get to New Haven and keep on going until it gets to Washington D.C. Like they do today.
Amtrak’s current schedule has only two trains per day going direct from New Haven to DC, each direction. Both leave southbound before 9 am and arrive northbound after 10 pm. If you upgrade the NEC to any sort of HSR, then there will be zero diesel trains from that unelectrified branch continuing south to interfere with the schedule of the faster trains.
Let me get this straight, they are going to spend lots and lots and lots of money to build a causeway across Long Island Sound but not electrify between the high speed tracks in Springfield, which cost lots and lots of money too, and the high speed tracks in New Haven?
In the mean time supposedly ALP-45 locomotives can do 125mph using electricity and they could use those. I don’t know why they don’t but they could. Or something from Siemens.
Let me get this straight, they are going to spend lots and lots and lots of money to build a causeway across Long Island Sound
Who on earth is talking about a causeway across Long Island Sound?!?!?! Are you reading the same thread???
The discussion is between upgrading NH-Hart-Springfield-Worcester-Boston to 200/250 kph versus building a high-speed bypass of the Shore Line in Connecticut. If you do the former then Providence gets benefit from day 1 (because in either scenario you build high speed tracks from NYC to NH). If you do the former, then you have to spend even more money to electrify NH-Hart-Spring for those areas to get any benefit. And if you do electrify NH-Spring as is you are wasting money because you will have to tear a lot of that electrification out to upgrade it later. You are also leaving ridership on the table by not connecting Hart-Spring to Boston, or Worcester to anywhere. Electrifying the line as is (especially all the way to Boston) is dumb because there are or people in Hart-Spring-Wor than in Providence. You should provide high speed service to the route with more people first.
but not electrify between the high speed tracks in Springfield, which cost lots and lots of money too, and the high speed tracks in New Haven?
The only way there will be high speed tracks in Springfield is as part of a NY-Bos route. Your dreams of lots of trains from Boston to Utica, Buffalo, Cleveland and beyond are mental vaporware, there will always be much more ridership from Boston to the larger and closer metros of NY, Phila. etc. And if you did have true HSR tracks from Boston to Albany, that is even more reason to build a HSR route through Hartford first rather than just electrifying as is, because again there are more people in Hart/Spring/Wor than Providence and making them slow down on old tracks from Spring-NH with HSR on either end is dumb.
@Onux, if the NEC was competently run the two train from north of New Haven would presumably be bi-modes (and be more spread out). A bit like the direct service from London to Aberdeen or Inverness.
It’s 60 miles from Springfield to New Haven with commuter service. With connections to Amtrak and Metro North shown on the timetable. In ancient times there used to be mildly frequent through service. Then except for those odd peculiar strange city folk in southwestern Connecticut, everybody was going to drive everywhere all the time. No money was spent on trains. The fairly frequent trains stopped running. Probably because it was slower than taking the bus.
Amtrak is buying something that offends Onux’s sensibilities, that will be able to raise it’s pantographs where there is catenary. And enough cars to be able to send them through to Manhattan and beyond.
On another note, I’m curious whether upsetting the grandfathering of inadequate platform egress is an issue with these stations, similar to how it would be for a “transit” station under NFPA 130.
How bad are the NFPA 130 issues outside New York?
(And separately, the NFPA 130 issue in New York goes away if you can guarantee that two trains don’t platform on opposite tracks of the same platform at the same time.)
The T is working (slowly) on upping the Fairmount’s game but I don’t know that they are planning anything that would prevent running some Franklin/495 trains that way from Readville. That would free up some slots on the SW Corridor for Amtrak as well as Needham trains. I think the connection exists for Providence trains to go that way as well. Not every train — Back Bay is a real station with real passengers that get off there — but my guess is that more than 1/2 are going to South Station.
Need to #JustStringWires on both the Fairmount and the Franklin/495 as well, of course.
I don’t see a need to start building 16-car platforms. The NYC commuter rail system is mostly based around a maximum of 12-car sets, and Amtrak today only runs 8-9 car trains most of the time. Going to 12-car trains would be a significant capacity improvement. There are so many other projects on the NEC to tackle, like getting many stations up to 12 cars first, before trying to build mega-long stations just because some other countries do it.
While 16 car trains might be highly useful–PRR ran same in the 60s when I used them between NYP-PHL-DC–I don’t think they are as necessary at the frequencies proposed. More particularly, 16 car trains need not stop at all stations. Trenton was skipped by some PRR trains decades ago and North Philadelphia is no longer a regular stop though BWI should be upgraded to both longer platforms and four tracks. Decades ago, when I learned to use trains and transit, active door locations were announced prior to the arrival of a train. This is an AI task that should be easy to implement. much more difficult at NYP will be getting the crew changes tighter.
it has been my habit for 6+ decades to grab my light luggage and be in the exit vestibule well before arriving–the doors pop open, I step off and head for whatever stairs/escalators or sidewalk.
BTW, building a new station over in Alexandria roughly where Potomac Yards used to be as part of the Long Bridge project and making that the trade off point for diesels rather than Union Station in DC, should allow faster swaps and “plant the flag” for wire at least to cover VRE if not to Richmond. In the bargain, VRE shifting to EMUs should speed them up increasing fluidity on that route.
As to the supposed advantages of dual mode engines, when hauling the dead weight of the prime mover, fuel, and electrical equipment needed to drive the traction motors becomes zero, they might be worthwhile.
i should point out that years ago I had a cab ride in a string of PRR EMUs running between 30th St and NYP as an advance section of a DC-NY holiday train. We beat the allotted time by well over 15 minutes making all of the scheduled stops in the 89 miles just on superior ability to negotiate curves, slow down and accelerate as necessary.
While 16 car trains might be highly useful–PRR ran same in the 60s when I used them between NYP-PHL-DC–I don’t think they are as necessary at the frequencies proposed. More particularly, 16 car trains need not stop at all stations. Trenton was skipped by some PRR trains decades ago and North Philadelphia is no longer a regular stop though BWI should be upgraded to both longer platforms and four tracks. Decades ago, when I learned to use trains and transit, active door locations were announced prior to the arrival of a train. This is an AI task that should be easy to implement. much more difficult at NYP will be getting the crew changes tighter.
it has been my habit for 6+ decades to grab my light luggage and be in the exit vestibule well before arriving–the doors pop open, I step off and head for whatever stairs/escalators or sidewalk.
BTW, building a new station over in Alexandria roughly where Potomac Yards used to be as part of the Long Bridge project and making that the trade off point for diesels rather than Union Station in DC, should allow faster swaps and “plant the flag” for wire at least to cover VRE if not to Richmond. In the bargain, VRE shifting to EMUs should speed them up increasing fluidity on that route.
As to the supposed advantages of dual mode engines, when hauling the dead weight of the prime mover, fuel, and electrical equipment needed to drive the traction motors becomes zero, they might be worthwhile.
i should point out that years ago I had a cab ride in a string of PRR EMUs running between 30th St and NYP as an advance section of a DC-NY holiday train. We beat the allotted time by well over 15 minutes making all of the scheduled stops in the 89 miles just on superior ability to negotiate curves, slow down and accelerate as necessary.
They use dual modes and non dual modes in Britain. They are timetabled to do the electric runs at the same pace.
The world has moved on.
All that means is the electrics are not being used at full potential. Caltrain–San Francisco-San Jose–having gone full EMU has speeded up all train service with the pleasant result of ridership gain. Even the all stop locals are faster than in diesel times.
So the trains from London to Bristol Parkway which is 112 miles do the trip in as little as 73 minutes with two stops and a top speed of 125mph. That is an average of 92mph. Seems pretty quick to me.
Even pure DMUs can be pretty quick. The 2307 from Marylebone to Birmingham gets to Bicester North which is 55 miles away from London in 42.5 minutes. That is an average of 78mph with a top speed of 100mph. And the trains are 20-30 years old.