Intercity Trains and Long Island
Amtrak wants to extend three daily Northeast Corridor trains to Long Island. It’s a bad idea – for one, if the timetable can accommodate three daily trains, it can accommodate an hourly train – but beyond the frequency point, this is for fairly deep reasons, and it took me years of studying timetabling on the corridor to understand why. In short, the timetabling introduces too many points of failure, and meanwhile, the alternative of sending all trains that arrive in New York from Philadelphia and Washington onward to New Haven is appealing. To be clear, there are benefits to the Long Island routing, they’re just smaller than the operational costs; there’s a reason this post is notably not tagged “incompetence.”
How to connect the Northeast Corridor with Long Island
The Northeast Corridor has asymmetric demand on its two halves. North of New York, it connects the city with Boston. But south of New York, it connects to both Philadelphia and Washington. As a result, the line can always expect to have more traffic south of New York than north of it; today, this difference is magnified by the lower average speed of the northern half, due to the slowness of the line in Connecticut. Today, many trains terminate in New York and don’t run farther north; in the last 20 years, Amtrak has also gone back and forth on whether some trains should divert north at New Haven and run to Springfield or whether such service should only be provided with shuttle trains with a timed connection. Extending service to Long Island is one way to resolve the asymmetry of demand.
Such an extension would stop at the major stattions on the LIRR Main Line. The most important is Jamaica, with a connection to JFK; then, in the suburbs, it would be interesting to stop at least at Mineola and Hicksville and probably also go as far as Ronkonkoma, the end of the line depicted on the map. Amtrak’s proposed service makes exactly these stops plus one, Deer Park between Hicksville and Ronkonkoma.
The entire Main Line is electrified, but with third rail, not catenary. The trains for it therefore would need to be dual-voltage. This requires a dedicated fleet, but it’s not too hard to procure – it’s easier to go from AC to DC than in the opposite direction, and Amtrak and the LIRR already have dual-mode diesel locomotives with third rail shoes, so they could ask for shoes on catenary electric locomotives (or on EMUs).
The main benefit of doing this, as opposed to short-turning surplus Northeast Corridor trains in New York, is that it provides direct service to Long Island. In theory, this provides access to the 2.9 million people living on Long Island. In practice, the shed is somewhat smaller, because people living near LIRR branches that are not the Main Line would be connecting by train anyway and then the difference between connecting at Jamaica and connecting at Penn Station is not material; that said, Ronkonkoma has a large parking lot accessible from all of Suffolk County, and between it and significant parts of Nassau County near the Main Line, this is still 2 million people. There aren’t many destinations on Long Island, which has atypically little job sprawl for an American suburb, but 2 million originating passengers plus people boarding at Jamaica plus people going to Jamaica for JFK is a significant benefit. (How significant I can’t tell you – the tools I have for ridership estimation aren’t granular enough to detect the LIRR-Amtrak transfer penalty at Penn Station.)
My early Northeast Corridor ideas did include such service, for the above reasons. However, there are two serious drawbacks, detailed below.
Timetabling considerations
Under current plans, there is little interaction between the LIRR and the Northeast Corridor. There are two separate routes into Penn Station from the east, one via 32nd Street (“southern tunnels”) and one via 33rd (“northern tunnels”), each a two-track line with one track in each direction. The North River Tunnels, connecting Penn Station with New Jersey and the rest of the United States, face the southern tunnels; the Gateway tunnels under construction to double trans-Hudson capacity are not planned to pair with the northern tunnels, but rather to connect to stub-end tracks facing 31st Street. For this reason, Amtrak always or almost always enters Penn Station from the east using the southern tunnels; the northern tunnels do have some station tracks that connect to them and still allow through-service to the west, but the moves through the station interlocking are more complex and more constrained.
As seen on the map, east of Penn Station, the Northeast Corridor is to the north of the LIRR. Thus, Amtrak has to transition from being south of the LIRR to being north of it. This used to be done at-grade, with conflict with same-direction trains (but not opposite-direction ones); it has since been grade-separated, at excessive cost. With much LIRR service diverted to Grand Central via the East Side Access tunnel, current traffic can be divided so that LIRR Main Line service exclusively uses the northern tunnels and Northeast Corridor (Amtrak or commuter rail under the soon to open Penn Station Access project) service exclusively uses the southern tunnels; the one LIRR branch not going through Jamaica, the Port Washington Branch, can use the southern tunnels as if it is a Penn Station Access branch. This is not too far from how current service is organized anyway, with the LIRR preferring the northern (high-numbered) tracks at Penn Station, Amtrak the middle ones, and New Jersey Transit the southern ones with the stub end:
The status quo, including any modification thereto that keeps the LIRR (except the Port Washington Branch) separate from the Northeast Corridor, means that all timetabling complexity on the LIRR is localized to the LIRR. LIRR timetabling has to deal with all of the following issues today:
- There are many different branches, all of which want to go to Manhattan rather than to Brooklyn, and to a large extent they also want to go on the express tracks between Jamaica and Manhattan rather than the local tracks.
- There are two Manhattan terminals and no place to transfer between trains to different ones except Jamaica; an infill station at Sunnyside Yards, permitting trains from the LIRR going to Grand Central to exchange passengers with Penn Station Access trains, would be helpful, but does not currently exist.
- The outer Port Jefferson Branch is unelectrified and single-track and yet has fairly high ridership, so that isolating it with shuttle trains is infeasible except in the extreme short run pending electrification.
- All junctions east of Jamaica are flat.
- The Main Line has three tracks east of Floral Park, the third recently opened at very high cost, purely for peak-direction express trains, but cannot easily schedule express trains in both directions.
There are solutions to all of these problems, involving timetable simplification, reduction of express patterns with time saved through much reduced schedule padding, and targeted infrastructure interventions such as electrifying and double-tracking the entire Port Jefferson Branch.
However, Amtrak service throws multiple wrenches in this system. First, it requires a vigorous all-day express service between New York and Hicksville if not Ronkonkoma. Between Floral Park and Hicksville, there are three tracks. Right now the local demand is weak, but this is only because there is little local service, and instead the schedule encourages passengers to drive to Hicksville or Mineola and park there. Any stable timetable has to provide much stronger local service, and this means express trains have to awkwardly use the middle track as a single track. This isn’t impossible – it’s about 15 km of fast tracks with only one intermediate station, Mineola – but it’s constraining. Then the constraint propagates east of Hicksville, where there are only two tracks, and so those express trains have to share tracks with the locals and be timetabled not to conflict.
And second, all these additional conflict points would be transmitted to the entire Northeast Corridor. A delay in Deer Park would propagate to Philadelphia and Washington. Even without delays, the timetabling of the trains in New Jersey would be affected by constraints on Long Island; then the New Jersey timetabling constraints would be transmitted east to Connecticut and Massachusetts. All of this is doable, but at the price of worse schedule padding. I suspect that this is why the proposed Amtrak trip time for New York-Ronkonkoma is 1:25, where off-peak LIRR trains do it in 1:18 making all eight local stops between Ronkonkoma and Hicksville, Mineola, Jamaica, and Woodside. With low padding, which can only be done with more separated out timetables, they could do it in 1:08, making four more net stops.
Trains to New Haven
The other reason I’ve come to believe Northeast Corridor trains shouldn’t go to Jamaica and Long Island is that more trains need to go to Stamford and New Haven. This is for a number of different reasons.
The impact of higher average speed
The higher the average speed of the train, the more significant Boston-Philadelphia and Boston-Washington ridership is. This, in turn, reduces the difference in ridership north and south of New York somewhat, to the point that closer to one train in three doesn’t need to go to Boston than one train in two.
Springfield
Hartford and Springfield can expect significant ridership to New York if there is better service. Right now the line is unelectrified and runs haphazard schedules, but it could be electrified and trains could run through; moreover, any improvement to the New York-Boston line automatically also means New York-Springfield trains get faster, producing more ridership.
New Haven-New York trips
If we break my gravity model of ridership not into larger combined statistical areas but into smaller metropolitan statistical areas, separating out New Haven and Stamford from New York, then we see significant trips between Connecticut and New York. The model, which is purely intercity, at this point projects only 15% less traffic density in the Stamford-New York section than in the New York-Trenton section, counting the impact of Springfield and higher average speed as well.
Commutes from north of New York
There is some reason to believe that there will be much more ridership into New York from the nearby points – New Haven, Stamford, Newark, Trenton (if it has a stop), and Philadelphia – than the model predicts. The model doesn’t take commute trips into account; thus, it projects about 7.78 million annual trips between New York and either Stamford or New Haven, where in fact the New Haven Line was getting 125,000 weekday passengers and 39 million annual passengers in the 2010s, mostly from Connecticut and not Westchester County suburbs. Commute trips, in turn, accrete fairly symmetrically around the main city, reducing the difference in ridership between New York-Philadelphia and New York-New Haven, even though Philadelphia is the much larger city.
Combining everything
With largely symmetric ridership around New York in the core, it’s best to schedule the Northeast Corridor with the same number of trains immediately north and immediately south of it. At New Haven, trains should branch. The gravity model projects a 3:1 ratio between the ridership to Boston and to Springfield. Thus, if there are eight trains per hour between New Haven and Washington, then six should go to Boston and two to Springfield; this is not even that aggressive of an assumption, it’s just hard to timetable without additional bypasses. If there are six trains per hour south of New Haven, which is more delicate to timetable but can be done with much less concrete, then two should still go to Springfield, and they’ll be less full but over this short a section it’s probably worth it, given how important frequency is (hourly vs. half-hourly) for trips that are on the order of an hour and a half to New York.
Pedestrian Observations 
Do you think Amtrak (or long distance service in general) into Long Island could exist assuming scheduling and capacity on the NEC can’t improve to accomodate it? To reword, other than the NEC, would other “pairings” with Long Island work for intercity service? For example the circulating proposals for Amtrak service towards Scranton, or pairing with service into upstate New York.
Good question!
Pairings to the Empire Corridor could work at the level of what happens within New York City. But then on Long Island, this gets dicey unless the Main Line is quad-tracked.
I had in the past imagined using Long Island as a terminus for Empire Service trains, on the principal that no trains should terminate in Penn Station if at all possible. But it looks like the Empire Service enters Penn Station _south_ of the North River Tunnels, which I didn’t realize, so I assume an idea like that is also a non-starter?
The Empire Connection needs to be realigned to face the northern tunnels anyway, for commuter rail purposes; we discuss this in the ETA report, section 6A.
I’m sorry, I can’t get behind this logic at all. If I understand it, your argument boils down to two points:
Neither is supportable.
First, if merging a branch into the NEC was so difficult, then the house of cards should already have collapsed with the Keystone Service doing so 13 times a day between Phila and NYC, and this with the difficulty of a reversing move at 30th St. Station. Even the Vermonter makes its way down the NEC once a day, joining the NEC for its most(?) congested portion from New Haven to NY, without breaking the whole system. Indeed, in this regard, you could even say that your logic of not wanting a “branch” to interfere with the “main line” is flawed since all intercity trains from the north are already ‘branching’ from tracks shared with MNRR and its service pattern to tracks shared with LIRR and its service pattern. Just finding a slot among LIRR trains alone and sticking to it should be easier.
Penn Station should also be the easiest place to merge with the NEC, because of the multiple of tracks and platforms. It is easy to have an extra track or platform for intercity service, such that a late LI train has a space and isn’t blocking another train behind it. Your plan for Penn was 11 through tracks serving 6 platforms – an RER style setup for through running commuter rail with each approach track becoming two platform tracks would only need 8 tracks and 4 platforms, so room to spare for Amtrak.
Managing merges and branches is just part of running a railroad, people do this all over the world all of the time. Even the unbranched Tokaido Shinkansen still has to deal with scheduling around some services originating from or continuing to the Sanyo and others not.
If your argument is just that Amtrak doesn’t have a good history with tight and on-time scheduling I won’t disagree, but it you are proposing HSR for the NEC or through running LI-NJ then you are effectively proposing that Amtrak/commuter railroads get better, so they could get better at this too.
Second, the notion that load will ever be balanced N and S of NYC is just not supported by the facts. Per the 2023 Census updates, the population along the NEC north of NY (breaking out MSAs from the CSAs, and including places without direct NEC service like Poughkeepsie and Waterbury) is 12.9M. Just Wash-Balt-Phila is 15.8M, expanding to all MSAs part of those CSAs or served by the NEC (Richmond, Lynchburg, etc.) is 22.4M. Nothing will overcome that difference to allow balanced intercity load. Springfield/Hartford/N Haven are not large enough to offset the equivalent of three Bostons.
Greater Bos-Phila or Bos-Wash ridership does not reduce the difference in ridership because those trips involve riders on both sides of NY, not just additional riders on the north side. Plus, HSR gets highest ridership at trips of 2hr or less, which is DC/Balt/Phila-NY, and Bos-NY, but not Bos to farther S. Amtrak already has decent mode share DC-NY, but since 30% of 70 is more than 50% of 30, it is possible that the N-S balance will increase due to the size of the market to the S, even if ridership from the N has a greater percentage increase.
The fact is that geography means the NEC is set up to branch, and this should be embraced, not rejected. In addition to serving LI, the Empire service to Albany should be made a branch of the NEC continuing south to DC. As I have said several times, if you are going to serve Hartford/Springfield you need to electrify all the way to Boston, allowing Hartford to receive equal service as Providence and help balance the schedule (if not perfectly balance the load). LI is an obvious part of this, Nassau and Suffolk have a pop equal to the Baltimore MSA or Hartford+Providence. Since Jamaica serves Queens, you could add some hundreds of thousands more, even if they do have service today via Penn. Serving 3+M people should be an automatic yes.
In a perfect world I suppose you would serve LI via Wall St/Brooklyn. The link would give a way for trains ending at Hoboken or Atlantic terminal a way to be more useful, you wouldn’t congest Penn, Downtown/Wall St has more jobs than the Phila & Balt CBDs combined, and Brooklyn has about as many people as Nassau/Suffolk – a lot of ridership potential.
1 is more than just “merging a branch is hard.” Unlike the Vermonter example, this has the following additional problems:
1a. Very difficult timetabling on the LIRR Main Line already.
1b. Additional timetabling difficulties on the LIRR Main Line from running an intercity stopping pattern – off-peak trains today run local east of Hicksville and there are only two tracks with no passing segments.
1c. A solid alternative of making people transfer like today – both lines should be running frequently enough for untimed transfers anyway.
2 depends on what has been built, yes. The 15% difference figure assumes the extent of the system is the NEC proper plus Springfield, with the average speed between New Haven and Springfield set at 110 km/h. If there’s also Keystone service averaging 120 km/h in there, then NEC-north has 20% less ridership density than NEC-south, both sides excluding commuters – and the point is that commuter traffic density is going to be more symmetric.
There’s no real need to treat Boston-Springfield as part of the same system. The faster Boston-New Haven route will always be via Providence and New London; the idea of running some lower-fare but higher-operating cost trains on the slower line is a RENFE-style brainfart, not practiced in the systems that Amtrak has any business learning operations from except by accident.
The best way to serve Brooklyn, by far, is to improve subway connectivity from Brooklyn to Penn Station. This means keeping working on making the subway faster through slow zone elimination, signal improvements, work zone curtailment (the flagging rules were less restrictive in the 1990s and 2000s than in the 2010s and worker injury rates were if anything lower), and deinterlining at least at DeKalb; in the long run, it also means subway lines under Nostrand and Utica.
1a There is very difficult timetabling on the New Haven line between MNRR locals and expresses, downstream joins with the Harlem and Hudson lines, and Amtrak. You have written extensively about the difficulties. If Amtrak can find room to slot 6-8 tph through Connecticut and STILL slot in the LIRR timetable (although the ESA improvements at Harold make this easier, Amtrak still has to share East River Tunnel space with LIRR) then slotting in just the LIRR timetable should be doable at 1 tph.
1b Hicksville to Ronkonkoma is only ~40km. It is timetabled for just 2 tph during the day, and 6 tph peak. At the busiest time, an inter-city leaving 2 min before the LIRR would have to be going better than 114kph to catch the one ahead of it before Ronkonkoma. During the 2 tph midday forget it, with 30 min windows Amtrak can find a spot with no need to pass.
1c You could apply the same logic to the Keystone (just transfer to a faster Acela at 30th St.) yet it is advantageous based on load to have more trains running from NYC to Phila and branching off, just as it would be an advantage to have more trains running from DC to NYC and branching off to LI.
2 Total population of MSAs N of NYC is 14.9M (includes Hartford/Springfield and Empire Service to Syracuse). Total for MSAs to the S on an NEC line is 22.4M (includes Keystone). There is no way with 50% greater population that the south sees only 15-20% greater ridership. Today ridership is 233% higher to the S. HSR would improve that a lot by making the NY-Bos portion not as slow, but the population difference is still insurmountable. The way to address it is to provide service to Bos through Prov and Hartford (both branches would be ‘over served’, but not as bad as if all trains went through just one area, plus branch to LI, plus branch to Albany. Some of these options are of course long term.
Commuter traffic density has nothing to do with intercity traffic density. Symmetrical commuter density is an argument to through-route 24-40 tph from LI-NJ, not to through-route 8 tph from DC-Bos.
Bos-Springfield is absolutely part of the NEC system. The point isn’t Bos-NY traffic, but NY-Hartford, NY-Spr, Hart-Bos, and Spr-Bos. All are valid markets (Hart+Spr equals Providence in size, and being farther from Bos there is a greater inter-city market (as opposed to just taking MTA commuter Bos-Prov). This isn’t RENFE running service on a parallel line or SNCF running OuiGo as an overlay on the same route, this is a different but equally valid route. SBB runs trains from Basel to Zurich trough Aarau, but also through Brug/Turgi.
For what its worth, the first HSR route N of NY should go through Hartford. A route that covered the ~378 km NY-Bos this way in 2.5 hr would be averaging ~150kph express, which means that a 200kph line (not even a new-build HSR standard of 250kph) could do it. The enormous ridership gains would then pay for a faster 2hr or <2hr route to Bos via I-95 Prov or direct from N Haven to Bos as you please. This provides huge benefits to Prov (it already has 25kv catenary and could use the HSR tracks S of N Haven) while also providing huge benefits to Hart/Spr/Worcester. Starting HSR via Prov only helps the Hart-Spr shuttle, while not unlocking Bos-Hart/Spr routes.
For a person who advocates for national electrification for small countries/regional areas, it is baffling that you would advocate stringing wire from N Haven to Spr and then leaving the next 145km to Bos un-wired.
A tunnel from Hoboken to Brooklyn has advantages on its own merits, given the size of DT NY. Anyone who advocates for a Bos North Sta-South Sta link (as you have many times) could not possibly object, given that DT NY has three times the jobs of DT Bos. Once you have that it makes perfect sense to provide subway transfer free intercity service to Wall St (500k jobs!) and Brooklyn (2.7M people! Would you skip Pittsburgh and Cleveland on the way to Chicago?). Suggesting to just improve subway connections is like saying we can close Baltimore Union Sta., lets just extend/improve the Balt subway so they can transfer at BWI. Plus, so much of Brooklyn is served by the 6th Ave or Broadway lines which don’t directly connect to Penn Sta at 34th/Herald Sq, so you are asking for a longer trip and double transfer compared to a station in DT Brooklyn or Atlantic Terminal.
Yeah, no. The slow part of NYC-Bos is NYC-New Haven. It takes about 4h on Acela today, and New Haven is (in timetable terms) about the halfway point, rather than the roughly third of the way to Boston it is in distance. So any serious plans to speed up NYC-Boston needs to get New Haven-NYC sub-hour speeds anyways (with attendant benefits for everybody going through New Haven).
Beyond that, though, the Boston-Providence route is already largely usable at HSR speeds, so all you need is reconstruction of New Haven-Providence to get acceptable times, which is going to be far cheaper than building New Haven-Springfield *and* Springfield-Boston to acceptable standards. There’s definitely utility to a New Haven-Hartford-Springfield corridor, and separately a utility for Springfield-Worcester-Boston, but it’s not a better option for the *first* HSR route for New Haven-Boston, especially not in an environment where the benefits of that route only accrue at the end of the project, rather than having intermediate stages that provide intermediate benefits.
@jcranmer,
That Bos-Prov is usable at HSR/near-HSR speeds now is an argument FOR building a new HSR/near-HSR route through Hartford before a new route from N Haven to Prov.
If you build NY-NH at HSR speed, then you immediately cut Bos-NY travel time to 3hr, with benefit to Prov. directly and Hart/Spring connections. If you then build a 200kph+ route NH-Hart-Spring-Bos, you can drop Bos-NY time to 2h30m, while vastly improving NYC-Hart/Spring and opening up the Bos-Hart/Spring market. Then you can build a 300kph+ route from NH-Prov (or all the way to Bos) and drop Bos-NY time to 2hr or less. This is the staged approach with benefits at each step (first to Bos/Prov, then Bos/Hart/Spr, then Bos/Prov).
If instead you build a full HSR line through Prov, you probably don’t get a sub 3hr Bos-NY journey any sooner (after addressing NY-NH, building a greenfield alignment through SE Connecticut should be just as long and difficult as upgrading the Hart-Spr-Bos route), and at the end of all of that you still can’t get from Bos-Spring/Hart or from NY-Hart without a transfer, and to go Worcester-NY you have to backtrack to S station.
Look at the map at the top of Alon’s post. See the big grey oval stretching from NH to Spring with no red line in it? See the dark green area with a red line in between NH and Prov? The gray represents people who ride trains, the green is trees that don’t. Collectively the Hart/Spring/Wor area has ~50% more people than Providence (this is true whether you measure metro areas, urban areas, or New England County/Township areas (a unique Census product adapting N England’s colonial area legacy boundaries). Giving Bos a trip to NY in 2hr while leaving the 2.4M people in the CT river valley and central MA with no HSR service is a bad tradeoff. Giving Bos a 2:30 trip to NY would be a game changer (faster than DC-NY today!). Doing so while putting Hartford within 1:15 of NY and Bos is the win-win solution.
The difference is that I-95 is a 120 km corridor flat enough not to require tunnels whereas the New Haven-Hartford-Worcester-Boston route is longer and isn’t happening without tunnels.
@Alon,
Would tunnels be needed to get a 200kph route? Remember I am not suggesting the final full-speed HSR route for the NEC be Hartford. It just needs to provide 2h30m service NY-Bos. If you design NY-NH to be 0h40m (in line with a 2h total time Bos-NY express) then NH-Hart-Spr-Bos just needs a design speed of 175kph to average 125kph express and finish the journey in 1h50m.
In any event, I would say that a few tunnels and ~230km of upgraded track to provide service to 2.4M people is a good deal. That is a population equal to the Austin, Pittsburgh, or Sacramento MSAs, and larger than metro Las Vegas, Cincinnati, Cleveland, or Indianapolis.
There are 11 million people in Massachusetts, Connecticut and Rhode Island. There can be TWO high speed routes in and out of Boston. One to New Haven and one to Albany. If there is one from Providence to Harford that screws Worcester and Springfield. And eventually doing things like Montreal, Toronto… Cleveland, Detriot…
Two and half hours is too slow. If it takes two and half hours to get to New York – from Boston – people will still fly to Philadelphia, DC, etc. It facilitates things like “4 hours to Pittsburgh” which gets lots of connecting flights out of Northeast airports.
It’s about 150 miles to go from Boston to New Haven via Worcester, Springfield and Hartford, and somewhat less to go via Providence. However, you only need to build ~70 miles of new track between New Haven and Kingston, since the track from Kingston to Boston is generally good for 150mph speeds already. 70 miles of new track only gets you from Springfield to Newton (beyond which you’re fucked anyways for any new ROW that isn’t underground).
Which is easier? 70 miles of easy greenfield track, or 70 miles of somewhat harder greenfield track plus upgrades to 70 miles of existing track that’s also supporting dense local stopping patterns? That’s what the choice boils down to: if you’re going through Providence, your job is already half done.
Chiltern runs a reliable service with local stopping trains to High Wycombe and express trains to Birmingham with only one track per direction and limited passing places.
Um Um Um…. they “branch” off the NEC when they go to Sunnyside Yards. The LIRR trains are merging in. So are the SEPTA, NJTransit and sometime any decade, Metro North trains.
PSA is supposed to open in 2027.
So the additional merging and diverging will be sooner!!
Adirondacker, for once we agree. The idea that Hudson Tunnels-Penn-E River tunnels/LIRR-Harold-Hell Gate-New Haven/MNRR line is no big deal, but Hudson Tunnels-Penn-E River tunnels/LIRR is a super complex problem just doesn’t pass the logic test.
Perhaps for the few dozen people, a month, who want get the twice weekly flight to Nowhereastan. Railfans over estimate the importance of airports.
It’s also the connection point to much of Queens…
They aren’t going to be taking Amtrak from Woodside, assuming Amtrak stopped in Woodside, to get to JFK.
In other posts you whine moan and complain that during the peak of peak periods there are three tunnels serving one direction. Make up your mind about what is which.
The LIRR Main Line east of Harold runs 3+1 peak direction. Grand Central Madsion and Hunterspoint trains are bled off for their termninals. The tunnel to Penn Station always run 2+2.
Why do they need to change trains is someplace other than Jamaica? ….. they do, at places farther east.
You can change trains in Woodside. It’s how Long Islanders get to events in Flushing Meadow Park, whether it’s Citifield or one of the other destinations.
Because PSA won’t serve Jamaica, so no opportunity for PSA/ESA transfers.
Why would anybody except an addle brained railfan want to do that?
It’s too bad you can’t juggle more than one thing at a time. The whole point of Penn Station Access is to keep people OUT OF GRAND CENTRAL. With the side benefit that it will make their trips faster.
Because connections are always good. Right now every MNRR train goes to GCT, it used to be that every LIRR train went to Penn. Now some LIRR trains don’t go to Penn and soon some MNRR trains won’t go to GCT. That means for some people you won’t be able to take the next available train to get where you want to go, you would have to wait for a (slightly) later train. With a PSA/ESA transfer, someone in Stamford wanting to go to East Midtown (or wherever along the N Haven line) can get on a PSA train heading to Penn if it is the first to come, then cross platform to an LIRR train that will take them to GCT like they want, faster than if they waited for a GCT bound MNRR train. Same for someone on LIRR wanting to get to Penn when a ESA train pulls up.
Connections almost no one will use are pointless.
Why would that be a problem? Just like sending the Ronkonkoma train THRROOOOOGH Penn Station to Port Jervis is never going to be a problem why would an Amtrak train have a delay?
The entire point of through-running systems is to set things up in a way that reduces cascading delays. The crayon bit at the end of the ETA proposal that I linked in my comment to DW goes over how the multiline crayon separates the system into pieces that don’t interact with one another – and how, in intermediate stages, some lines just don’t touch the through-running core but instead divert to Brooklyn or Hoboken or the non-through-running parts of Grand Central until additional tunnels are built.
“the multiline crayon separates the system into pieces that don’t interact with one another”
OK, so just ensure that the Mainline/Ronkonkoma trains are the ones that through route to Newark/NEC, and then the Amtrak trains are not interacting with commuter trains on a different route.
The Amtrak trains and the commuter trains will be interacting with each other until sometime far far in the future when there is separate set of intercity tunnels across Manhattan. Even then there might be some sharing between the New England trains and the commuter trains on the eastern ends of the LIRR and Metro North.
I feel like the recent posts are just build up for a banger about NEC high speed rail… every time there is a post, I hope it’ll be the one, but its just more buildup (good buildup). Looking forward to it!
Oh, it’s not going to be a post – it’s going to be a full report.
I’m actually for this and your argument that slotting a train on LIRR is hard, also proves the point that LIRR schedule needs to blown up and redone.
LIRR has added excessive amounts of padding. Additionally, the LIRR trains run at glacial speeds. Consider that a non-stop LIRR train from Grand Central Madison to Jamaica, needs over 20 minutes to cover 10 miles. Average speed is around 30 mphs on along tracks where MAS is significantly faster and top speed just isn’t reached anywhere.
Perhaps adding Amtrak trains to Ronkonkoma would the “bitch-slap” that LIRR needs? 🙂
Most likely, LIRR will “bitch-slap” Amtrak to be slower than they do their own trains. It’s their ralroad, which they do a lousy job of running to begin with, and things got much worse once Grand Central service started. The Main Line 3rd track didn’t help much at all.
As it is, Amtrak more often than not dilly-dallies on the MN New Haven Line at about 50MPH on the local track and loses 10 minutes on the way.
Physics is a cruel mistress. Volts times amps equals watts.
The LIRR tried running the dual mode third rail/diesel locomotives, at higher speeds, using third rail. It didn’t work out. They and Amtrak switch over to diesel as soon as possible. Those locomotives are ancient. Things may have changed over the past few decades.
But volts times amps equals watts hasn’t. A megawatt divided by 750 volts is 1333 amps. A locomotive needs multiple megawatts. How many third rail shoes does this mythical locomotive need to accelerate to the track speed limit?
….Maybe Bombardier/Alstom could slap some third rail shoes on the NJTransit electricfied multilevels. The Keystones could go to Washington DC. Revived “Clockers” could originate at Suburban and go to Ronkonkoma. Or slap some third rail shoes on Silverliners.
Things have moved on. And bi-mode trains are more of a thing now – https://en.wikipedia.org/wiki/Electro-diesel_multiple_unit.
And there are even ones that do third rail and run on diesel – https://en.wikipedia.org/wiki/British_Rail_Class_769
The proposal is apparently for an all electric locomotive. The existing third rail/diesel locomotives switch over to diesel as soon as they can. They tried running them, at higher speeds, via third rail and it didn’t work out well. That was decades ago. ….. Physics is a cruel mistress and volts times amp equals watts.
The Amtrak plan is for a train that’s several minutes slower than LIRR trains even while making fewer stops, so what “higher speeds.”
Something more than creeping into or out of a Manhattan terminal.
People do travel by train between the southern half of the NEC and Long Island. That does not mean sending Amtrak to Long Island is commercially viable. It is not, and would be a waste of train miles.
LIRR operates at least 5 Penn Station trains per hour all day. LIRR has 110 stations and 10 branches. LIRR shares the Moynihan Concourse with Amtrak, making transfers much easier. Amtrak run Regional trains up and down the NEC every hour, plus several Keystones to Philly. That means travelers will go when they went and to/from where they want on Long Island and not arrange their life around the 3 Amtraks for one branch and 5 stations. Nobody headed to Stony Brook, Patchogue, or East Hampton will settle for Ronkonkoma and call a taxi.
If an Amtrak train terminates in NYPS, then it should be sent back to Washington or Virginia within an hour two. Sending it to Long Island and back, possibly not until the next day, is a waste of train miles and poor equipment utilization.
There are also all sort of technical reasons that make it difficult for Amtrak to run on the LIRR, including differences in the signal system pulse codes, shore-horning it in around LIRR service, neither Amtrak nor LIRR is a Swiss watch operation, Mineola is only on the 3 track main line’s local track, wasting diesel fuel when there is 3rd rail (Airo trains will not have 3rd rail shoes, not even the Empire Corridor ones), either qualifying Amtrak crews on the LIRR or qualifying LIRR crews on Amtrak equipment, and servicing the equipment in Ronkonkoma. Notice Amtrak only proposes sending federally supported NEC trains, not state supported Empire Corridor trains. NYS-DOT would never fund such an operaion when MTA runs there just fine.
Railfans and Planners like the idea of running equipment all over the place, even when it serves no purpose. This is about as silly as running an Amtrak Boston train to Bay Head, Raritan, or Dover on NJT.
I agree with most things, but I disagree that people would prefer to pack into a 3+2 seating train with luggage for an hour, interrupt their trip to transfer to their intercity train. Just as many people adjust their schedule to catch an express train vs a local train, many people would happily hold out for a direct train where they can setup their laptop, order a glass of wine, and ride in a higher level of comfort.
Airlines continue to add more business class seats, more premium economy, and more economy plus to their planes because customers expect a higher quality experience.
California HSR is prototyping seats that go beyond just 1st class and 2nd class seats.
Additionally, airlines are flying non-stop to more destinations with smaller fuel-efficient planes like A321XLR.
We can argue as to how large a direct Ronkonkoma to DC or Ronkomoma to Albany market is, but it seems foolish to dismiss it.
Perhaps LIRR Infrastructure could use better signalling system or more power feeding the 3rd rail. Maybe the Ronkonkoma station needs another platform with two more tracks to serve intercity trains?
There are more people on Long Island than there are in Metro. D.C. Does that mean it’s silly to send Amtrak trains all the way to Union Station in D.C.? They could all just transfer to MARC in Baltimore… there are almost as many people in Nassau and Suffolk as there are in Baltimore.
Just because Long Island has 3 million people does not create a false equivalency of it to the nation’s capital or Baltimore. Washington is one city and destination. Long Island is a huge suburban sprawl 120 miles long of over a hundred villages and hamlets. Comparing the MARC Penn Line to the LIRR out of Penn Station is also ridiculous.
Long Island has 8 million people. 3 million live east of the border of New York City. Queens is more densely populated than the District. Roughly twice as dense. And roughly 4 times the population.
It’s why the NAR HSR plan makes sense. Even at the silly price.
But more likely to get that lunar railway than here on earth where it could connect up >30m people. So much for Amtrak Joe.
A tunnel of that length does not actually make sense. If you build an I-95 bypass between New Haven and Kingston, grade-separate some junctions (like Shell), and run a track laying machine on the New Haven Line to replace the ties and increase the cant, you can do New York-Boston in two hours. Every concrete-heavy megaproject has to be judged against this baseline, and not against the current trip time.
@Alon
2h is not 1h40.
And it is not supporting those 8 million Long Islanders, forcing them to make long treks into Manhattan.
As for a tunnel, there are probably cheaper options because isn’t most of the sound fairly shallow? Japan has built longer tunnels to link far lower population centres.
It’s not 8 million, it’s charitably 4 million – Brooklyn and western Queens don’t gain anything from this.
Then, the New York-New Haven trip times go as follows:
Grade separation at Shell but no other ROW deviations, intermediate stop at Stamford: 0:52
Grade separation at Shell and the usual ROW deviations I’ve mapped, intermediate stop at Stamford: 0:45
Grade separation at Shell and the usual ROW deviations I’ve mapped, nonstop: 0:42
Sunnel, intermediate stop at Jamaica: 0:45
Sunnel, nonstop: 0:42
The speed zones with the Sunnel route are 160 km/h to Jamaica (Jamaica itself is just east of a 140 km/h curve), 250 km/h to Hicksville, 100 km/h at the curve east of Hicksville, 140 km/h at the curve between Bethpage and Farmingdale, 300 km/h to Ronkonkoma, 200 km/h through the curve from east of Ronkonkoma to the ROW used to run north to the Sunnel (which already requires some takings, though probably less than the Darien bypass), 320 km/h in the ROW and the Sunnel, and then 210 km/h and 150 km/h in the last few km to New Haven as in the NEC alignment that it rejoins.
Japan built the Seikan Tunnel, yes, at a cost of $18 billion in inflation-adjusted 2023 dollars, but ridership has been disappointing. The decision to build the tunnel was made in 1971, but in the subsequent 17 years of construction, Japanese economic growth slowed and also air travel increased; the expectation was that rail-ferry-rail traffic between Tokyo and Sapporo would keep increasing, but instead people switched to flying. Tokyo-Sapporo can get some modal split back when it becomes an all-Shinkansen route, but most people will probably keep flying, as they are on Tokyo-Fukuoka.
(For what it’s worth, you may be delighted to hear that none of the French rail projects so far is a white elephant like that, except maybe the Chunnel, and that’s the fault of Maggie plus British security theater and not just that of SNCF.)
Why would you need more tunnel to get from Manhattan to Long Island? They already have six tunnels to Manhattan. One across the Hudson would be useful.
Shorehaven Long Island to New Haven Connecticut is a bit longer than the Chesapeake Bay Bridge Tunnel. Since there aren’t any naval bases west of there, for the Defense Department to fret about, it wouldn’t even need tunnels. It could be all bridge. Roll it into a toll bridge for road traffic. Getting people from …. east of the Throgs Neck Bridge… who want to drive to New England out of Queens, the Bronx, Westchester and Fairfield would be a very good thing. It might even be people east of the East River because traffic in the Bronx, Westchester and Fairfield is execrable most of the time.
“Long Island” means Nassau + Suffolk – 3 million. Nobody is going to take Amtrak to Jamaica only to scatter into the rest of Queens and Brooklyn. They’ll do that when they get to Penn Station. How would anyone get to Brooklyn ? Backtrack on the LIRR to Flatbush or schlep on the J train to East New York and WIlliamsburg ? Forget it. The rest of Queens means only the E train or a bunch of buses at Archer/Parsons. Railfans on a Saturday excursion is about all who will do it.
As for Long Island, that’s 3 trains a day to 2 stations in Nassau and 2 stations in Suffolk. What about everywhere else, and throughout the day ? Again, they’ll sort themselves out in Penn Station, not cool their heals in Jamaica, an open air station with little seating and no amenities.
The only reason Stephen Gardner came up with this, and the rest of “Connect US” (I call it “Conned US”) is political payback for his Guardian Angel, Senator Schumer, and to distract and entertain advocates from Amtrak’s shabby service delivery and because Ronkonkoma has a big, white elephant of a new train yard. An Amtrak train would run just as slowly as any LIRR train, inluding getting stabbed for 5 minutes for no reason outside of Jamaica and Hicksville while the so called Control Centers figure out where their trains are or stick to their stubborn track assignments.
Three million people is metro Baltimore. Or more than metro anything in New England not-Boston. That doesn’t change that Brooklyn and Queens are on Long Island.
How are they supposed to get from Manhattan to Jamaica to scatter to the buses? Until there is a tunnel from New Jersey to Brooklyn, most people in Brooklyn would be going to Manhattan for west/southbound trips. If the high speed route goes via Long Island, some of them would want to go to Jamaica for trips to New England.
Why would it go through Hicksville when the former Central of LI ROW is still there and owned by the government in one form or another? You have to re-imagine the LIRR and integrate that with every-ten-minutes service to Western Queens through Brooklyn to equally close and densely populated suburbs in New Jersey. And decide how closely spaced the trains are under the Hudson between Jersey City and Manhattan. 26 an hour gives you every ten minutes on four branches and two intercity trains. “Wall Street” is a third biggest central business district in the country and assuming the tracks are there for commuter service running intercity trains on them is “free”. Unlike anything to Saint Loius or San Antonio.
….. squinting at the maps, LIRR Main line east of Yaphank is owned by the government or very lightly developed. Shaming Brookhaven Labs into the awfulness of LEED certified new office buildings might help.
Because Amtrak does not know the Central RR of LI right of way exists, they came up with this, nobody else cares, and it would be a legal and costly nighmare to reactivate it. or that would have been the 3rd Main Line track to begin with.
The Central Branch doesn’t really save you any time – it’s only around 1.5 km shorter from Floral Park to Farmingdale, and doesn’t save you the Hicksville curve because the curve to the branch at Queens Interlocking is about as tight. And then it’s a two-track ROW whose inner section has local commuter rail traffic.
And it’s not France. Or Japan or Spain. New Rochelle, Stamford and Bridgeport can have twice an hour service to points west of Manhattan just like Worcester, Springfield and Hartford can. While you are reimagining LIRR service if there are six trains an hour across the Sound that leaves “slots” for Suffolk County express trains to super express to Jamaica – change for Wall Street or Grand Central etc.
If you’re building a bypass then it has to go through Hartford, one way or the other. The I-95 bypass is a stupid idea which deserved to be killed in the 2010s and deserves to be killed again on the merits (95 wasn’t designed to be a rail corridor, zero new riders over on-corridor improvements which are fully capable of delivering NY-BOS in more than 2 but less than 3 hours at a price tag that leaves us with plenty of spending money to hit places that actually want rail, a much bigger than zero number of new lawsuits from the local landed gentry who are already mobilized to fight this one and much more capable of fighting this one than anyone north of them, multiple feasible alternatives in the world where we’re laying 150 km of brand-new right-of-way at any cost providing a much larger ridership potential for what I believe you yourself worked out to be roughly 7 minutes) even if everyone agrees on its technical feasibility to permit New York-Boston in two hours.
And for the record, if New England is large enough for two HSR lines out of Boston, then I-95 is about the fifth line after the one to Providence and Hartford, the one to Springfield and Albany, the fever dream under Long Island Sound, and the fever dream along I-84.
The question is not, as someone will be along to scream about shortly, whether Providence is more important than Worcester/Springfield – it’s whether Hartford is more important than New London, which maybe sees no service anyway due to the bypass putting its station some 15 stories up, and in that case the question is “is it worth it to add another HSR stop in Connecticut or not?” I think it is, but I also think you could easily just cut Rhode Island out of the equation entirely and easily run a single high-speed rail line from Boston to Springfield to New Haven if people are that concerned about the cost of land in rural eastern inland Connecticut.
No it doesn’t. Just like in the past if you wanted to go to Hartford, you took a train that went through Hartford instead of one that went through Providence, in the future people who want to go to Hartford can take a train that goes through Hartford. Which is a different train than the one that goes through Providence.
And building high speed rail next to the Connecticut Turnpike isn’t going to ruin it’s bucolic charms no matter how much astroturf NIMBY’s whine about it.
You can’t reconfigure things in Connecticut and then whine the existing is no good on Long Island. The curve in Hicksville looks slightly tighter. There wouldn’t be any curve in Farmingdale. It could blast through at more than 300 kph. Eyeballing it from satelite images it’s four tracks wide. It would have to be four tracks wide because the commuter train can’t be stopped on the same track in Deer Park at the same time the train to Boston is coming through at 325 kph.
You are getting into silly talk. The LIRR’s speed limit is 80MPH.
I checked the trip times at various top speeds – give me some credit. Jamaica-Hicksville at 250 vs. 300 km/h is something like 20 seconds of time difference. The Hicksville curve can be rebuilt but it’s nasty and the same is true of Farmingdale; even the 200 km/h curve from east of Ronkonkoma to the ROW would involve more takings than on any Connecticut bypass except Darien.
It would have to be on dedicated tracks because the train stopping at Garden City or Brentwood would be in the way. Or the one moving between Wyandanch and Pinelawn. Or.. If you go down the West Hempstead Branch and use the former Central of LIRR ROW there is no curve in Farmingdale. No slowing down at all between Floral Park and Yaphank. The undeveloped land in the general vicinity of the Long Island Expressway and William Floyd Parkway is going to be easier to condemn than anything in Darien.
I think you can improve the corridor up to Springfield/Boston, plus run some Long Island service (although I’m not 100% conversant with all operational issues). As for time & slower speed to Long Island: yes it’s slower with the congestion issues mentioned. But a handful of trains from both the NEC and the Empire would enhance access, and perhaps take some strain off Penn Station (more through running trains).
Also, I appreciated more the importance of Jamaica for transfers.
Long Island trains run off the high-numbered tracks, and “thorugh run” as it is to West Side Yard, not dwelling in the station more than 10 minutes, with most arrivals on tracks 20&21, which do not even access the Hudson river Tunnels. There are basically no conflicting opposing train movements from 20&21.
Empire trains “run through” to Sunnyside and back from station tracks 5 through 8.
LIRR and Empire paths seldom cross except through tunnel line 1& 2, which LIRR makes little use of, usually just rush hours when there are not many Empire trains to begin with, at most one per hour. So there is no strain to be relived by thru-running them, and would add to ladder track congestion if they tried.
It seems like there’s sufficient space to quad track the main line out to at least Mineola with minimal changes to existing infrastructure other than at Merillon Ave and New Hide Park (unless they went back to triple track there), and no land acquisition. Since that would alleviate conflicts with Port Jefferson Branch trains, do you think that would be sufficient to allow Ronkonkoma Amtrak service?
It definitely would be, and lead to changes in LIRR service too.
It does, actually. 12 million people is enough for ONE high-speed line, which means your ONE line has to serve as many people as possible, which means hitting Hartford and Providence. I’m sorry that you have to share your train and can’t have two separate routes, but, as you yourself like to say… IT’S NEW ENGLAND. It’s not Honshu. It’s not even New York to Philadelphia, the one place on the line you maybe actually can make the case for a brand-new bypass line next to the right-of-way you already have.
If you can justify building any new right-of-way then you can justify building it to where people actually want to go, and it turns out train lines aren’t the same thing as highways and you actually can’t just draw a line down the median and expect it to work – in other words, it’s still a brand new right of way and it doesn’t actually matter that a freeway happens to be there currently.
Otherwise, upgrading the existing shoreline is good enough, and 2.75 hours end-to-end still beats flying and driving every hour of every day, and you can take the hundreds of billions you didn’t spend to do the same thing in 99 other places, or maybe even build a brand new right of way if you want one between some other city pair that doesn’t currently have train service.
There is life west of Ninth Ave. Two hours and fourty five minutes between Boston and New York would need very very very fast trains to make it to Washington D.C. in three hours. Or even Philadelphia.
It’s 11 million people. Less if the people in Fairfield County Connecticut are using alternate trains that toddle through there instead of going via Long Island. If 12 million people isn’t enough for two high speed lines then neither is Chicago. Or any place else in the U.S. not Los Angeles.
Life west of Ninth Avenue doesn’t care where their train originated or how long it took to get to them in Penn Station, because it’s still going to be the same 90 minutes from there to Union Station either way.
Boston to Washington in 4 hours is still faster than flying. The drive takes 8. The people will survive life with the legacy shoreline if the new line can’t be built.
People who want to get from Boston to Hartford don’t care if the train went through Worcester or Providence.
Which high speed rail corridor does Chicago get? Detroit? Saint Louis? Cleveland? 12 million people aren’t enough people for two corridors the 10 million people in metro Chicago definitely aren’t worth three.
Correct, just like New England isn’t big enough for multiple bespoke high-speed rail lines, Chicago isn’t either. They get one rail line which goes east to Detroit, then turns south, and goes onto Toledo and Cleveland. Just like how everyone has to share their one train in New England, everyone has to share their one train out of Chicago, too. Sorry to fans of St Louis or Green Bay, but there’s just not enough there yet for HSR in either location – or if there is, it’s in another east-west line connecting all those small CSAs that line up. Otherwise, they can get by with a train running at 110 MPH. So can the Ohio corridor, which desperately needs rail – regional intercity, conventional speed rail. So does Pittsburgh. So do all these other places that suddenly become attractive investment options if you’re only willing to let go of the magic talisman that says “HSR.”
So can New England if we don’t want to spend the money to connect Hartford to Providence.
They get one rail line which goes east to Detroit, then turns south, and goes onto Toledo and Cleveland.
HSR would go east to Toledo, then branch to Detroit and Cleveland. Going to Detroit first then south turns an ~525km journey Chi-Cleveland into a ~665km one, adding an unnecessary hour to the trip. Also your “everyone gets one line and shares” mantra seems to have the root intent of not building too much expensive HSR track, but Toledo is only ~85km from Detroit, which means your route has 55km more track (and expense) than going to Toledo and branching (525km Chi-Cleve + 85km Tol-Det = 610km).
None of them are the navel of the universe that is Hartford. By his own logic Toledo and the rest of Ohio don’t get anything because the Midwest blew it’s wad on Chicago-Detroit.
Detroit is the second largest metro area in the Midwest. It’s on the way to Toronto. Michigan owns ROW. And Detroit is the second biggest metro are in the Midwest. Just like you can almost spit from Niagara Falls New York to Niagara Falls Ontario you can almost spit from Detroit to Windsor Ontario. Where someday perhaps maybe the Canadians will have high speed rail tracks to Toronto and Montreal. Connecting the two biggest metro areas in the Midwest to Toronto is more important than Toledo. Going across Ontario is much shorter than going through Cleveland if you are aiming for Detroit and upstate New York and New England.
Detroit is around 100km from Toledo. It wouldn’t add an hour.
120 km of tunnel-free, mostly in a preimpacted corridor rail is not hundreds of billions; try $5 billion, saving around 24 minutes (cf. the Frederick Douglass Tunnel at $6 billion saving 2.5 minutes).
It’s 120km with at least two water crossings, one of which is the New London problem, but in fairness I probably should have said “tens of billions” instead of hundreds. Regardless, I have no faith in it actually being deliverable at $5 billion.
It’s 167km from New Haven to Hartford to Providence. Discounting the New Haven to Hartford portion we’re left with about 110km, trading one of the crossings for a tunnel, and trading fighting with NIMBYs for fighting with the Sierra Club. I’m willing to entertain the idea that there’s a marginal cost increase, but only that – and maintain that, say, $6 billion here is better than $5 billion on I-95.
Yes, that’s why I’m saying $5 billion and not $4 billion. The numbers come from costs of European lines that have water crossings, for example the LGV Méditerranée with its four bridge crossings of the Rhone plus some tunneling elsewhere. New Haven-Hartford-Providence is considerably more than this because it does need some tunnels – not a lot, but enough, including a short one in Providence itself, that it’s visible in the topline cost.
Any amount spent here is visible in the topline cost is my main objection. Sure, it’s “24 minutes saved” – but it’s also the least important 24 minutes you can find anywhere on the line. Boston to New York is going to be equally viable at 108 minutes instead of 84 and it’s going to be equally viable at 2 hours or 2.5 hours; running a greenfield alignment at a slight additional premium in time or money or both (you can probably get away with a much cheaper tunnel if you go south down the rail trail instead of straight west through Olneyville and you can maybe also get away with no tunnel if you clear out the 10 freeway instead) becomes worth it for the extra ridership it represents.
Otherwise, it’s better to spend it where less time is saved but for far more people or where it will enable us to run far more trains or even where it will extend rail service to less economical put more politically imperative regions who need to be paid off to get them on board anyway – Georgia, Tennessee, South Carolina.
There’s no such thing as viability – there’s ridership as a function of trip time, and this function not only has a negative slope but also has a positive second derivative. If spending X on getting from 3:00 to 2:30 is justifiable on a benefit-cost ratio, then it is also justifiable to get from 2:30 to 2:00.
There is no saturation in this relationship (although, to be clear, ridership at 0:00 is not infinite, due to access and egress time). Rather, in countries that have high-speed lines going all the way between major cities, the cost of increasing speed is really about what it takes to raise the top speed, and so the cost of saving a minute increases as the trip time decreases. In contrast, if a line is mixed, with some fast and some slow segments, then the cost of building fast segments to replace slow segments is usually fairly constant: construction costs per km are usually the same along the line if the topography is the same, and if an area is more difficult, it’s probably because the topography is more difficult and then it means that the slow segment is especially slow, so the higher cost per km also translates to a larger time saving per km. For example, the 10 billion € budget of the Halle-Bamberg high-speed line across the mountains between Thuringia and Bavaria led to a two-hour trip time saving; at typical German high-speed rail construction costs, reducing the trip time further from four hours to around 2:30 would cost maybe another 10 billion €, barely more per minute saved, and the reason it’s not done is that Germany in general is underinvesting in rail infrastructure.
People in New England want to go places other than New York. Apparently great big thundering herds of them in Hartford just yearning to see Providence, and vice versa. Who won’t do it if it takes ten minutes more.
@Alon
this function not only has a negative slope
Although Transit Hawk is generally completely wrong in his analysis of what does for doesn’t warrant HSR, to be fair the time/ridership slope is not even. It is very steep between 2:00 and 4:00, and then shallow on either side. Ridership gains per minute below 2:00; once you reach 2:00 air mode share is 0%, so there are not many riders to gain. Similarly, below 4:00 you have lost almost all riders already (isn’t Tokyo-Fukuoka 5% rail, 95% air or such?) so improving from 5:00 to 4:30, say, isn’t doing much. Just as it is more important to make slow areas fast than fast areas faster, improving a line from 4:00 to 3:30 is much more important than improving from 2:00 to 1:30. So while you are technically correct that there is no limit of viability, as a practical matter travel times below 2:00 are not that beneficial.
In this regard a 2:30 Bos-NY trip time would be game changing, Acela gets 83% mode share DC-NY at just 2:45.
These 2:00/4:00 rules of thumb of course assume a reasonable intercity distance of a few hundred km, longer or shorter distances would operate differently. Also, those time/ridership curves were developed with the Shinkansen in Japan running at 210-220kph and the TGV in France at 250 – the advent in recent decades of trains in the 320-350kph range should shift things.
The question of Bos-NY speed is also a factor of the size of destinations beyond Bos. The quicker one can reach NY the higher the ridership to Phila/DC and vice versa. Bos-DC could be a 2:30 trip at Chinese express HSR speeds, stopping only in NY. But this would never happen because of how NEC cities are arranged. So the question becomes would the expense to get Bos-NY down to 1:30, Bos-Phila down to 2:35 and Bos-DC down to 3:00 be worth if for the number of riders gained over 2:00/2:45/4:00.
It’s not in fact correct that the slope gets shallower at 2:00. Competition with air saturates around 2:30 or so, but as speeds increase, there’s better competition with cars and also induced trips. Cascetta-Coppola find that the ridership impact of the opening of new HSR in Italy is consistent with an elasticity of ridership with respect to travel time of about -2, with four out of seven city pairs examined having been below 2:00 even before the new HSR opened. The elasticity of modal split is much shallower, only -1, the balance of extra ridership coming from induced trips.
I’ll add that focusing exclusively on competition with air has been a serious failing of HSR operations in parts of Europe, especially Spain – where SNCF and DB’s monopoly services aimed at serving everyone, RENFE imitated air in an attempt to outcompete the air shuttles between Madrid and Barcelona, leading to a situation with seven different train classes on this city pair each with its own pay scale. Ridership was around half what one would expect from the TGV (or as far as I can tell the ICE) relative to city size and trip time, until liberalization led to more service and lower fares, finally matching what the state monopolies here and in France have always done.
The meme that HSR works best at “three hours or less” comes from the 60s and 70s. When airfares were regulated and airlines had to compete on service. And there was no security. And you could be at the airport curb 15 minutes before your flight. When traffic was a lot less congested and parking was easier.
There is no “beyond Boston”. Rail enthusiasts in especially frothy fits of fannery imagine high speed rail to Nova Scotia. There are potato fields and tree plantations east of Portland Maine. Or north of it.
Sending a train across the lightly populated hills and valleys of Connecticut to connect Providence and Hartford at screamingly fast speeds means the train isn’t in New Haven. Or even Waterbury. How much does it cost to get screamingly fast speeds from Hartford to New Haven? Alternately a much cheaper ROW could be built along the coast and the great big thundering herds of people who apparently have an unmet desire to travel between the two could get to Hartford at more conventional speeds – from New Haven – much cheaper. And still be faster than driving.
Wlimington is almost exactly halfway between New York and Washington D.C. it’s not going to take an hour to get between New York and Philadelphia and then only half an hour to get to D.C.
There is a lot of “beyond New York”. Metro Portand is a bit smaller than metro Harrisburg. Metro Lancaster, which is a bit smaller than metro Portland is along the way. Pittsburgh-Boston is around 600 miles which is four hours at an average speed of 150. Harrisburg to Pittsburgh will be all new build and it might be a bit less than 4. Three hours BOS-WAS means Boston-Richmond is 4. And there are umpteen small markets hiding in there like Providence-Philadelphia. That aren’t there in Chicago-Detroit-Toledo-Cleveland. But then Chicago-Detroit is a slightly longer than Boston-New York which means Toledo and Cleveland are too far away.
The meme about HSR working at three hours is not from the 1960s, because at the time the only place in the world with HSR was Japan, which didn’t have much domestic air travel. (Anthony Perl points out the contrast in New Departure: the TGV rescued SNCF from decline, whereas the Shinkansen was built before there was any decline in favor of cars and planes). It’s a piece of SNCF dogma from the 1990s and 2000s, which reflects how SNCF managers travel; based on this dogma, they believe HSR isn’t competitive at four hours, even though the TGV has better modal split than flying on Paris-Toulon. In Germany the dogma is instead that trains should take four hours.
Are you sure that isn’t a piece of A. Levy dogma about SNCF:-)
Perhaps they had that in the early days (which was probably before the airline management came on board?) but it doesn’t really jibe with the plan to link every corner of the Hexagone by TGV, which has to be 4 decades old policy. Not to mention reaching beyond the borders deep into most neighbours. They continued technological advances to reduce the travel times to these distant points.
Further, SNCF management have had a different public stance:
The reality has always been runing a couple random TGVs a day from Paris to some provincals’ towns, sometimes. Outside of that, you’re out of luck.
Des actions, pas des mots.
12 million people is enough for ONE high-speed line
This is a foolish way to think about planning HSR. It disregards distance, density, and the size of areas that line might connect to.
The proper way to evaluate system size is as route-km/100k population. Among countries with 300kph+ track, this figure ranges from 1.27km/100k (S. Korea) to 8.9km/100k (Spain, a system widely believed to be overbuilt on the back of an economic bubble). Taiwan is 1.39, Italy is 1.56, Belgium 1.79, Germany 1.93, Japan 2.17, China 3.19, France 4.13. Many of these countries have HSR lines under construction or planned that would raise their route density to between 2.64 to 5.78 km/100k.
There are some limitations to evaluating this way. Japan isn’t very high, despite the extensive Shinkansen network, because HSR does not go to Shikoku are barely reaches Hokkaido (although this is being extended) so while those areas contribute to the denominator of “per 100k” they do not have tracks contributing to the route-km. If you only divided HSR route length by the population of prefectures that were served, the Shinkansen network would appear more dense. Conversely, evaluating at too small of a scale has problems too. If you evaluated HSR along the NEC at the county level, there would be no HSR from DC to NY, because the tracks go through Cecil County in Maryland for 29 km, but the 100k population there “should” only have a few km.
At a macro level though, one should not look at the population of CT-RI-MA alone when considering the route length within those states. New York city plus Westchester is just under 10M, but the NEC crosses them in only ~50km, does that mean that we have to build HSR from Yonkers to the Rockaways to use up their “route budget”?
Totaling the population of the states served by the NEC (analogous to the country-level figures I gave above) gets to 70M people. At the current level of S. Korea that is 892km, which is just under Bos-Wash + Hart-Spr-Bos (964km). At the level of Italy you would have 1097km, which would allow adding the Keystone Corridor (1132km). At Japanese density, HSR length would be 1525km, equivalent to all of the above plus HSR to Richmond and Albany (1535km). Expand to include the length that Japan has in service plus under construction and you would get 1855km, enough for all of the above plus a line to Portland/Brunswick in Maine, and a new route Manchester/Nashua/Concord NH (1896km) [service to Maine and N. Hamp., where most of the population is in the southern 4 counties and are part basically part of Greater Boston, would also bring the population base to 73M.] At the current level of HSR in France, the greater NEC would have ~2900-3000km of route, which would add Springfield to Albany, Albany to Buffalo, service to Ronkonkoma, and Richmond to Newport News (2893km).
If you total by MSAs instead of states, you get 46-57M depending on if you include places like Albany in your count (population in consideration grows as you expand the network). This knocks each of the route totals above down a tier – the full NEC plus Hart. would be provided at Italy level route density.
In other words, it is not necessary to pick a one or two routes “out of Boston”, or aim for a compromise route Bos-NY. In any reasonably built out system there will be enough track for both N. Haven-Prov-Bos and N. Haven-Hart-Bos, plus major corridors like Richmond, Keystone and Albany.
When considering branches off the current main line, the order based on current or likely riders per km would be:
Hart-Spring-Worcester-Bos
Keystone Corridor
Empire Service to Albany
DC-Richmond
Bos-Brunswick
Albany-Buffalo
Richmond-Newport News
What’s the purpose of running a 300 kmph train line 157 kilometers? Acceleration isn’t infinite. You’re not going to instantly hit 300 for a 29 minute ride from Springfield to Boston – nor do you need to, nor would you even want to since your own analysis depends correctly on stopping the train at least once.
Actually, it’s the height of foolishness to bring in a calculation like density per route-km and then use that as justification for maximizing speeds in order to bypass most of those people. You’re not running local trains on that brand new HSR line – you’re just building your very fast bypass so “NEHSR” can sail by cars in the median of the Mass Pike at 300 while the person in Worcester trying to get to their job downtown either pays the premium for NEHSR Unlimited or continues riding the same bad commuter rail service they suffer with today, which might have gotten an incidental lift in the places where HSR was forced out of the bypass and onto the legacy alignment but for the most part remains untouched.
This is the problem. You just want to build a lot of HSR route-miles and hope for comprehensive rail reform to just fall into line behind it; you’re basing your analysis on all the places we could go if we had the Shinkansen but disregarding the rest of Japan’s comprehensive rail network as is the case for every other world peer you cite; your main concern is HSR as replacement for flight traffic and I would assert that is in large part because when you make that your main concern it means “ground” transportation no longer needs to factor in at all.
And you’re basing your analysis on the idea that it has to be HSR – it doesn’t. Let’s go back to that 157 mile route through the mountains of western Massachusetts. Let’s even generously assume we can hit 50% average speed as a function of top speed when accounting for stops and assuming “slow zone” penalties are uniform at all speeds. Okay, you can build “HSR” and get a 65 minute ride to Springfield. Or, you can fix the tracks you have already – at a much lower cost – and you have to give up your 300kmph dreams if you do, but you can still have 180kmph, averaging 90, and if you do that you have a painfully, abysmally slow, low speed… 105 minute ride. What’s the expected travel time in regular traffic according to Google maps again?
Washington to Richmond, 68 minutes or 113 and plenty of money leftover to pick up Hampton Roads (73 minutes out of Newport News at these speeds – or, 3:06 to Union Station; plus a completely separate service offering 110 minutes to Norfolk via Suffolk for those who really hate the bus). You might even have enough left over to fix VRE’s fundamental problems.
Hey, speaking of Richmond, you’ll not be delighted to know that they have seen the light on this one. They’ve let go of the magic talisman. Instead of obsessing over “SEHSR,” they are hard at work purchasing up abandoned rail lines and buying active tracks from CSX, building strategic passing tracks within the right-of-way that already exists, and it’s all in service of creating fast trains – but, critically, not HSR – to Richmond and from there to Raleigh. It’s really kind of embarrassing for New England that Virginia (and North Carolina) are leading the way on this one, but here we are.
There’s a whole country of possibility just waiting for us. All you have to do is put down the crayons and turn away from the glorious light of European Railroading.
What’s the purpose of sending one 117 km from Providence to Hartford?
Because there’s a perverse insistence on building a “bypass” here, on having “HSR to Boston” – and because money continues to flow into this black hole regardless of what either of us or any of us say. This is why I say that if a bypass is built, it must go through Hartford to Providence – if there’s $5 billion to waste on trains running next to 95, it can go towards drawing the line through the hills and valleys to keep trains moving at high speed where the concern trolls don’t have 200 years of Coastal New England money and connections.
Because the $5 billion should be spent buying up freight track in Massachusetts and stringing wires for MBTA Commuter Rail and expanding regional service to all the potato farms and mountain dwellers who don’t warrant lurid HSR crayon lines but do warrant a reliable conventional rail service on the tracks that are already there, but if we have to spend it on a “bypass” of a segment of line that absolutely doesn’t need to be bypassed, then the entirety of Hartford is worth the extra billions and the 10 minutes of everyone’s life in Boston.
But if you’re trying to gotcha me into saying this and that are different, I’m afraid to disappoint: Hartford-Providence is the consolation prize I’m aiming for, because I’m not going to get what I really want – which is to have the New Haven to Providence Study canceled outright and the money used on literally anything else.
Buying freight tracks means giving CSX money far in excess of what the tracks are worth, just because it can name its price. It’s turning Boston-Springfield into a much more expensive project than it needs to be, with some ridiculous plans like double-tracking the line with one track for freight and one for passenger rail rather than one westbound and one eastbound. Electrification is great, yeah, and also comes from a completely different source of funds, not really fungible with intercity rail funds except on the Providence Line, where the cost is the lowest because it’s already wired.
There are certainly challenges involved in buying track, some of which you’ve highlighted here. But the answer is not to concede and shift efforts towards building another line somewhere else – it’s to get CSX on board with better passenger-freight interactions. Here is another place where I’m willing to overpay on a lower margin initiative as opposed to overpaying for the overpouring of concrete in the form of a new bespoke HSR bypass whose primary purpose is to avoid interacting with the existing (underutilized!) right-of-way.
I mentioned thankless and unsexy work in another comment and this is generally what I mean by that – actually taking on the freight railroads, regulatory reforms at every level, local-first initiatives. A program for a national rail overhaul that ends with trains running from sea to shining sea, something that benefits all of us; and not the latest in a series of “visions” we’ve been pumping out for 20 years with nothing to show for it but L after self-inflicted L.
It’s only a different source of funding because of politics and because of the initiatives being written stupidly. We/the feds are certainly capable of authorizing spending on/redirecting funding to nationwide electrification among many other initiatives we could be pursuing (like hiring more crew or buying more trains for the routes we already have and the tracks with capacity for them already). We’ve chosen to spend that money on “identifying corridors” instead, and more broadly on continuing the search for the showcase centerpiece HSR project that the politicians can cut the ribbon for and say “look at what we delivered.”
And it’s never truly too late to make a different choice.
Um um. how is the trains supposed to get from Providence to Hartford unless you build something someplace else.
The train serving the overwhelming demand for Providence to Hartford can go through New Haven a lot cheaper. It has to be faster than driving not faster than flying.
That’s right! It could be cheaper! In fact, it could cost us $0 to use the existing tracks and continue to run the trains where they already run today. At worst, it costs a couple hundred million to rehab the bridge that’s already there and pay off the Coast Guard to get them to stop whining about “bridge slots.” Otherwise, it’s perfectly possible to run trains on the tracks that already there, and we know it’s possible, because we’re already doing it.
But we don’t want to spend $0. We want to spend billions and billions to have “HSR” running on its own special tracks, so because we have to do that, we will simply have to build one route that serves everybody as best it can, and everyone will just have to learn how to share it, and everyone who gets left out of that route will have to wait for 2100 when we’ve filled up the first line and can now justify a second one.
Or we could just spend $0.
To be fair HS2 would have been a cheaper Alon if they used the existing platforms at Euston as you have suggested before and terminated the sleepers out in the countryside. Hopefully that would also have the advantage that you could serve additional destinations such as Oban and also run shorter trains so you could stop at Birmingham New Street.
It would also have been cheaper if the trains were 14 cars long not 16 so they could use Birmingham New Street or Moor Street instead of a brand new station in Birmingham.
The same does to be fair apply to the North East corridor in the US, especially as presumably a fair chunk of it is nowhere near capacity.
Flying on the North East Corridor sucks as well, I believe the reliability is super bad. Certainty the one time I did it my flight was late.
We get that you are a frugal puritan who thinks Hartford is the center of the Universe.
….. Paris is the same-ish population as Southern New England. It’s truly excessive the way there are multiple high speed routes. They should have picked Paris-Lyon and given up. Because it’s not Hartford.
…… or excessive that Chicago has I-90 and I-80. New England should have been content with using the New Jersey Turnpike to get to the Pennsylvania Turnpike to eventually get to Ohio. But then the Penna. Turnpike doesn’t go through Hartford. It should have been I-84 through Scranton.
expanding regional service to all the potato farms and mountain dwellers who don’t warrant lurid HSR crayon lines but do warrant a reliable conventional rail service on the tracks that are already there
This is a well intentioned idea that ends up being more of waste of money than building the expensive HSR lines between major metros. The HSR will get riders, the regional service to farms and mountains will not. Tracks and catenary are heavy and expensive to build and trains are large and expensive to run, which means they do very well when there are large numbers of people to move. If there are not, you either end up running the train infrequently, which means in a world where cars exist no one rides it because the wait is too long, or you send up spending huge subsidies to have it come often enough, which means you are wasting money and energy on very empty trains.
As I note below, the French do this and their rural network goes largely un-ridden while the Japanese have been cutting their rural network (I think up to 50% of total lines in Hokkaido) in the absence of subsidies. The Swiss come closest, but Switzerland has unique geography where you cannot go from A-C by car and beat the train going A-B-C; since all traffic goes down one valley then up another, everyone goes A-B-C and the train is more attractive.
The way to address transit for regions that don’t warrant major rail lines is via subsidized bus service. The roads already exist and roads are cheaper to maintain than track. Busses can be scaled to meet service demand better than trains (again trains scale and do best at heavy loads). At low frequencies it is better to use 2-4 employees on one train as drivers for 2-4 busses instead since that means higher frequency and shorter waits for the passengers. Unlike trains which are stuck to the route of their rails, busses can use any intersection to change direction, which means instead of only providing people trip options of A-B-C-D east to west you can also offer A-X-Y-Z north to south, or A-B-E-F or whatever else meets their transportation needs.
Haddenham and Thame Parkway and Princes Risborough are both rural stations and both I am sure make a profit.
The trick is to run a serious amount of service, both have two trains per hour all day and the trains have an average speed of 50-60mph.
Yes, but Haddenham and Thame Parkway and Princes Risborough are on the Chiltern Main Line. It is London-Birmingham service that is filling trains at those frequencies, which then leads to better ridership at the rural stations.
<a href=”https://www.youtube.com/watch?v=wlMwc1c0HRQ”>Hmmm … Your ideas are intruiging to me, and I wish to subscribe to your newsletter</a>
Building a 300+kph line doesn’t mean that the trains all go 300kph. In fact in no line anywhere does any service average the top line speed. I never mentioned anything about maximizing speeds, I used countries with 300kph lines as a way to identify modern comprehensive HSR networks for a comparison of how much HSR could be built. All of those countries also have HSR lines with a top speed below 300kph. Service on the NEC should absolutely follow the Shinkansen model of Express/Nozomi (just DC-Balt-Phil-NY-N Haven-Bos), Limited/Hikari (current Acela stops), Regional/Kodama (all stops). So none of those people being used to frame the size of the network should be “bypassed” because even if some trains go by at high speed, others will stop.
While you are correct that high speed isn’t needed for 157km journeys like Bos-Spring, it is needed for longer journey’s like Bos-NY or Hart-DC or Rich-NY. If you accept an extra 40 min from Bos to Spring (157km) then by the time you reach DC you have added an 3:08. Your 105 min ride to Spring ends up making Bos-DC 8:12, when the Acela does it today in 6:38 and the NE Regional in 7:30-8:00. Your “this-speed-isn’t-so-bad-because-it-is-faster-than-driving” is actually slower than what is provided today, not some sort of upgrade. For what it’s worth, in the case of Springfield, I have always stated that I believe for an initial NEC HSR route through Hart-Spring-Bos that very high speeds are not necessary and we should in fact just upgrade the existing line to 175-200kph to allow Bos-NY in 2:30 (assuming other upgrades to get NY-N Haven down to 0:40). Faster speeds can be a further phase with a new-build very high speed line.
DC to Rich is 179km, so if you do it in 68 min that is 155kph average, not 90kph, when I thought you were arguing for 90kph average service.
I understand your argument for widespread upgrades to medium speed service, in many ways this is the approach taken by Britain (which now gets average speed on some of its major lines almost equal to the original Bullet Train), the Swiss (who prioritize timed connections over absolute speed), or Germany (with its short segments of NBS greenfield HSR lines mixed with other stretches of ABS upgraded lines). The problem is two fold. First, over longer distances the higher speeds are necessary to get the ridership. SNCF was in ridership decline before the TGV started, upgrades to the Paris-Lyon line were not going to help, but HSR did. Today that line is nearing capacity, the old Paris-Lyon route still exists, but no one rides it, so the French are planning a new LGV Paris-Lyon through Orleans. In Japan the Tokaido Main Line still exists, but it is used only for freight and commuter services, through intercity services no longer exist on it, everyone rides the Tokaido Shinkansen instead.
Second, is that the widespread comprehensive medium speed networks you refer to don’t do well. France heavily subsidizes regional trains, but it was recently found that 50% of the service miles carry only 2% of passengers. The TGV is full, the it-doesn’t-have-to-be-HSR is not full. You speak of Japan’s comprehensive rail network as a precursor to the Shinkansen’s success, but Japanese rail companies have been closing rural and regional lines for years due to low ridership. The Shinkansen is still full, and they are expanding the network.
I’m not sure why we would want to turn away from European railroading when all of their railroads have much better efficiency and performance metrics and all get much better ridership. I’m also not sure why you would cite European railroads as an archetype when HSR was invented in Japan, has its fastest top and average speeds in China, its largest network and ridership in China, and I specifically cite the Tokaido Shinkansen as a model for the NEC given both are linear corridors with many metros along the way.
If we don’t agree on definitions then we are never going to get anywhere other than sniping at each other while talking around each other. And in general I think there is still a fundamental disagreement here, in large part because the definition of HSR is a lot like the definition of obscenity – “I know it when I see it.” You mention using the presence of a 300kmph line as a way to identify modern comprehensive HSR rail networks, which is fine – but then is the size of that network everything which interacts with a line that at some point touches its top speed? Is it just the line or corridor that has that top speed segment? Is it actually some lower speed definition, say, 200kmph, and the comprehensive network is based on that standard?
I’m not asking you to commit to defining things in a reply to this comment, either. I’m explaining that my conception of HSR is very specifically going to be the CA model, which is building for the most part a brand new greenfield alignment through the desert with minimal track sharing in the terminal regions only as necessary. This is also the Shinkansen model – a key component of the Shinkansen which has gone generally unremarked on throughout our many arguments is that the Shinkansen doesn’t share tracks with anything except other Shinkansen trains; it has three classes of service it is concerned with and nothing else is present on the line. Once you give up that physical separation of infrastructure, you no longer actually have a Shinkansen-style service, even if you have Shinkansen-style stopping patterns; and conversely, adopting the Shinkansen or CAHSR model for places like the Keystone Corridor or Empire Corridor or service to Springfield means building that physically separated HSR infrastructure which can never be used for other traffic. (Which is a choice – and I’ll restate my “8 or 0” position on CAHSR as very much being that it’s a radical but not necessarily incorrect choice. You cannot make this kind of infrastructure investment unless you are trying to produce a singular beats-the-rest-of-the-world train line that gets everyone not in California excited enough to come up with another couple hundred billion to build another one for themselves, much like how nobody ever talks about what it was like to build the Shinkansen but only talks about other places new extensions of it can go – but to do that, CAHSR needs to be the Tokaido Shinkansen and must not be an hourly or twice-hourly “pilot” service.)
The other “HSR” projects moving forward in this country, for certain arguable definitions of forward, are also generally adopting the dedicated infrastructure model. Brightline West won’t even consider connecting to LA area terminal tracks at all, preferring to plant its “LA” Station in Rancho Cucamonga where its dedicated high-speed tracks will end, Texas Central is similarly planning on all greenfield and on using the actual Shinkansen rolling stock, every canceled Obama-era HSR plan (with the notable exception of the NEC) were all greenfield as well. I am hardly building a strawman here when I say that for most Americans, the public perception of HSR plans in America is that of dedicated greenfield alignments and that arguing against that kind of project – especially when it’s an I-95 bypass for the one part of the NEC where there isn’t a compelling argument for doing that (e.g. New York – Philadelphia, which could legitimately be said to be at sufficient capacity to justify a parallel bypass line) – is something I will never stop doing even if we come to an agreement that this vision of HSR is not what you are advocating for and not what you want.
But, again, that’s not the only definition of HSR even in the American perception of it. There’s another distinct possibility, which is informed largely by European HSR systems and also the Acela Express, that being a mixed-use legacy corridor upgraded to support high speeds. The main problem with this is not really with the corridor upgrades – which I wholeheartedly support – it’s with running HSR as a separated service, and especially as a premium separated service (this is an Acela failing but it is hardly exclusive to Amtrak.) There is no reason to have coach class on the Avelia Liberty be “business class” and there is also no reason that Avelia Liberties should not be used to run Regional service between Boston and Washington – no reason except for the fact that definitions matter and perception matters and Amtrak is neither interested in nor investing in a vision of HSR that includes regional services. It does not want the public to look at the Amtrak Airos that are soon to replace the Amfleets and think “this is high-speed rail,” regardless of whether or not we settle on a definition of HSR which includes them. And in this paradigm, at best, marginal benefits then “trickle down” to the regional and local services on those shared lines, should they exist – but ask me about how well it worked out every other time I was promised something would “trickle down” to me.
You correctly point out that medium speeds, competitive with driving, would be a downgrade to today’s NEC services. That’s true, but what’s also true is that it would be an upgrade pretty much everywhere else, notably including Richmond, where I indicated that a 90kmph average speed would produce a 113 minute trip to Washington (as opposed to 68 minutes at very high speeds), but this would be leaps and bounds ahead of existing trip times which are averaging roughly 135 minutes (give or take 10 minutes based on how padded any given run is) to reach Richmond “Staples Mill Road” and then another 26 minutes(!) to go 10 additional km to the Main Street Station for the 3 trains that do so. The story of Amtrak in all parts of the country that are not the NEC is this one: very slow, very infrequent, even when quite a lot of money is plowed into initiatives that purport to fix both flaws and don’t. (See again: Amtrak Midwest.)
And I would put forth that this is the story that is still being written, even on the NEC, where the fruits of HSR will result in Acela 2.0 and not the Acela Regional. It’s the story that is going to continue being written for a long time in the future, with every new chapter a treatise on why this line or that project failed, why America just can’t seem to find room in its heart for trains, why money shoveled into what projects do get off the ground never seems to buy anything of value, and why rail expansion projects which produce 2 to 5 daily roundtrips are critical flops.
And the only way to change that story is to close the book and pick up a different one: the story of a nation already blanketed by railroads setting forth on a path to modernize them – properly. Not through new HSR lines or service expansions, but rather through fundamental reforms: to how railroads are regulated, to how railroads are operated, to the companies that have proven unacceptably poor stewards of this infrastructure, to how railroads are funded. We need to let go of the idea that all we need is to put HSR on the American map, and embrace the idea that rail needs to improve at every level, on every corridor, in every use case. We need more North-South Rail Links and less NEC FUTUREs, more daily departures on Amtrak’s many corridor services and less “Corridor ID Project”s, electrification as a national mandate instead of something we might sometimes add incidentally when we’re not distracted by the siren song of batteries, more trains to the big city from nearby neighboring municipalities and more trains running from one end of the city to the other. Rebuilding local transit and getting people invested into alternatives to using their car on any random day of the week, not just as a replacement for longer distance traveling.
But every year that goes by where none of that happens but we have initiative after initiative for Visions of Amtrak Connecting US through a Modern American High-Speed Rail Network while the corridors crumble around us and the local subways are catching fire only deepens my objection to HSR. I’m not impressed by being invited to compromise, lectured on how HSR is complementary to local transit, educated on the intracies of Shinkansen stopping patterns or network sizes as a function of population density because none of that is what’s actually on offer here. And every year I grow less and less impressed by the perennial invitations to dream bigger and reach for the future in an advocacy environment too concerned with the everdistant horizon of a decade from now and more and more entrenched into the position that we need to reject the promise at the end of tomorrow and instead put all of our efforts into what we can fix today instead.
I didn’t start off vehemently anti-HSR, but 24 years of HSR politics in America sure did get me there – and it sure didn’t get me anywhere else, because those trains ain’t runnin’.
You cannot mix slow and high speed trains on track. The high speed trains end up catching up to the slow trains. You can run a mix of stopping patterns in some cases, that means running much less trains as the train slowing down to stop needs to have the following train that isn’t slowing far enough behind to not run into it, and then after the following train passes enough time to accelerate again (and then more deceleration time when we start over)
That said, we should in general build all new track to 300km/h standards even if we don’t run at those speeds. It lets our future generations (possibly investing in other track that meets their needs) use now legacy track. Upgrading track is nearly as hard as building all new (somewhat less takings, but they are still there)
Also even though you should plan to run all trains on a track at the same speed, that doesn’t mean you always don’t allow mixing. There are several emergency situations where you may choose to run all trains slow to get past the emergency. You may also use your high speed tracks as a low speed route while doing overnight maintenance on the low speed tracks.
Lyndon Johnson signed the High Speed Ground Transportation Act of 1965. The Pennsylvania Railroad and Budd Company were proposing the second generation of Metroliners would be capable of 160 miles per hour. 60 years later there are a few miles of track in Rhode Island and New Jersey that can support that.
Railfans and concern trolls lose sight of that. The fast automobile catches up to the slow automobile I don’t know why they cannot conceive of that happening on a railroad.
In some wildly optimistic service scenario there might be five or six high speed trains through Worcester-Turnpike an hour. The MBTA express train from Worcester can dive into the high speed tunnel after Framingham. That solves problems with the fast Worcester train catching up to the slow Framingham train
You cannot mix slow and high speed trains on track.
Except in Japan the highest speed Nozomi (220kph average) and the slowest speed Kodama (129kph average) both use the Tokaido Shinkansen even though it is only a two track line, and JR Central still manages up to 16 tph. There are passing tracks at stations, and slower trains will wait there then leave immediately after a faster train. The slower train reaches another passing track before the faster train catches up, then waits again. This is pretty basic.
if the slower train is loitering around on the other side of the island platform it’s not on the track where trains are moving. Most railfan’s brain deteriorate into quivering jelly when it comes to contemplating that the express leaves at :00, the limited leaves at :05, the local leaves at :15 and the train that uses an alternate route leaves at :20. Most places in North America don’t have enough origins and destinations for something that complicated. And even fewer have “alternate route”
if the slower train is loitering around on the other side of the island platform it’s not on the track where trains are moving
It’s only off the track for a few minutes while the moving train passes it. Right before that it was on the track with the moving trains with the faster train behind it and catching up. Right after that it will be on the track with the moving trains with the faster train ahead of it and pulling away. So yes, you can mix slower and faster trains on the same track.
it’s off the tracks. And the only place in North America this might be a problem is between Philadelphia and New York. Where there is enough room for six tracks. Where all of them will be able to use any track and any platform that is long enough. They can schedule the Philadelphia-New York non-stop that originates at Suburban and terminates in Brooklyn to use the commuter express tracks.
The problem is it’s more expensive to do that, so you have to justify it with the “HSR” service, and then it’s the “HSR” project where everything – even the unrelated things – is beholden to “HSR.” And then you look at the price tag and get sticker shock so you start making little compromises like carving it up into phases or offering less service or downgrading parts of the project to “provisioned for a future…” and all the while the price is still going up and you’re getting blasted from every direction in the press and maybe some techno grifter drops a squat on your project with some futuristic gadgetbahn that will definitely render this whole thing obsolete… Clearly a hypothetical situation right? Nah.
But this all is in service of something Richard Mlynarik justly points out in the followup article’s comment section: the “consensus building” and the “synergy” and the “compromises” only ever happen one way – they promise it’ll serve everyone, they insist that the only way to build out a modern functional regional rail network in any given state is to staple it onto the back of HSR, then somewhere between 0 and 1 of those 2 things materializes and it sure isn’t the humble regional train.
And so here I am, demanding compromises where actually this time HSR has to give something up instead of local transit. Here I am, insisting that everywhere gets exactly one line, so that we all go into this with the exact same expectations about what gets built and what gets scrapped. Here I am throwing cold water on everyone’s delightful crayon artwork not because I’m a concern troll or delighting in being Negative Nathan (that person’s name is adirondacker12800, just so we’re clear) but because I want it to be abundantly clear for all of us as to the reality of the situation being that every km of HSR track you fight for is 100 km of smaller scale improvements that won’t happen as a direct consequence of fighting for HSR instead.
You can either support local-first rail improvement, or you can support HSR. If you claim to support all rail spending, you’ve chosen HSR.
It must be odd then for you to observe that those countries that have embraced HSR are also the same ones with the best intracity rail transit. Is there any country that has HSR without good intracity transit?
For the US, one functional HSR line could transform the public’s attitudes to rail in general. Spending the same amount of dollars on all that small scale stuff will have exactly zero impact, except on the tiny minority ie. its current users. BTW, it needs to be real HSR not some half-assed low-speed (<250kph) compromise that doesn’t work for anyone yet still costs almost the same. “Make no small plans … etc.” (Burnham) Does anyone doubt that HSR between LA and the Bay Area would be wildly successful if it does it in 2h40m and is approx. cost competitive with air?
This is quite literally insane talk.
You know what utter fraudulent criminal bullshit was peddled to the voters of California in 2008? ”HSR between LA and the Bay Area … in 2h40m and is approx. cost competitive with air”, that’s what.
For a total cost $33 billion.
To be complete in 2020.
Remind me again what year it is?
Remind me again which planet we are on?
No doubt at all this has “transform[ed] the public’s attitudes to rail in general.” None at all! How could it not have done so in the past fifteen years?
You’ve chosen $250 billion. In 2050. Or 2060. Maybe. Maybe. Whatever. Eventually the public will appreciate these mighty, not-small works, and … not despair?
Great plans! Great success! Keep building and building on great success!
Stop freaking out and read what was written. “If”.
It only seems a fantasy to Californians like you but wherever such a line has been built it has been wildly successful. FWIW I suppose the most comparable is Madrid-Barcelona at 620km and covers not dissimilar challenging terrain, fastest time 2h38m.
I wonder whether you supported the SNCF plan of decades ago to link LA-SF. Which is when you should have built it. Or were you as frothing mad about that too?
My sole point is that if a proper HSR line could be built the insane resistance would fade away.
That’s like having a sole point about unicorns shitting rainbows. I mean, sure, go ahead and think about that to yourself all you like, my dude, but why bother typing?
Oh and at this point resistance isn’t “insane”. It has been 100% rational since 1996, when PBQD (today’s WSP), acting solely in the interest of “maximum corporate profit without ever delivering anything“, utterly and fatally fucked over the routing (Pacheco, to go with Palmdale, and later gaming for maximum freight rail/freeway impacts all the way along 99). It was game over then, and things have only gotten worse.
Everybody who has supported this disaster and who isn’t on the payroll has been conned and robbed blind.
But the problem is, California isn’t about to open even a proper HSR line. It’s about to open (in 6 years, massive asterisk MAYBE) an “Initial Operating Segment” from Fresno to Bakersfield, which will underperform, cause the actual phase 1 to be canceled, and leave us with a train every half hour from nowhere to nowhere that everyone will shake their heads over, decrying once again an example of how “rail doesn’t work in America.”
Richard has far more right than I do to be angry on this one, and I feel the need to actually apologize to him here:
I’m sorry for supporting CAHSR. I was wrong to do so. In previous comments, including on this very thread, I maintained – and still believe – that there was merit in actually building a singularly transformative rail line somewhere in America to make Americans believe in an interstate rail program. But for that to actually happen, it needed to be singularly transformative. Hence, my insistence on “8 TPH or nothing at all,” my insistence that “good enough” in the context of mature rail networks abroad would actually be a massive failure in California – and that if you are promising me the Shinkansen, you have to actually deliver the Shinkansen. It’s become clear that they’re not going to do that, though.
It’s even more clear that the people still supporting HSR don’t see a problem with that – see previous responses to my position “educating” me in how getting maximum return on investment for our 9-figure capital project in the form of running trains to its theoretical capacity was the “wrong goal” and how I/we need to learn to live with 2 or 4 tph because many European HSR routes do just fine with 2 tph and 96 daily roundtrip flights “only” translates into something like 36 HSR trainloads anyway.
I do still believe that if we ever open a train line in this country that offers a service equal to or better than the Shinkansen, that would be enough to erase all past sins. Unfortunately, that’s not going to be CAHSR, and the opportunity for it to have been CAHSR left the station sometime in the mid-2010s.
So, exactly what I wrote.
I do still believe that if we ever open a train line in this country that offers a service equal to or better than the Shinkansen, that would be enough to erase all past sins.
I don’t understand, this seems like a complete flip of everything you have said before. An HSR line with Shinkansen-style service is exactly what everyone has been arguing fore along the NEC, and you up to this point have been saying we shouldn’t do it?
Because what you’re arguing for is not the Northeast Shinkansen for as much as you insist on Shinkansen-style service for it – we have, in fact, already rightly ruled out the Northeast Shinkansen option that was on the table and we rule it out again every time another North Atlantic Rail comes along, in part because the classic grift is to promise something legitimately world-changing and then deliver somewhere between 0 and 10% of it at 100% of the world-changing price tag. (See: CAHSR)
Instead, what we are creating is Acela 2.0, at great expense, and possibly using the Shinkansen for operational cues – some of which are things that should have happened anyway, e.g., allowing someone in Penn Station to get on the next train to Stamford with their ticket no matter what class of service it is or what logo is on the train, running more trains at every service level – but these things, the components of this project that I actually like, are considered incidental at best or even unnecessary. Some of the things in Alon’s forthcoming report on an integrated NEC timetable are going to require the kinds of small-scale incremental improvement projects I’ve been pushing for this whole time such as junction fixes, which I support, but which don’t need to be predicated on HSR and in fact shovels should have been in the dirt for them already regardless. Other things, chiefly the 120 km of useless bypass track in Connecticut, I very much don’t support.
(I also think, in terms of national politics, most rail detractors who say “rail doesn’t work in America” already exclude the NEC from that, so there’s a political question of whether anything done here can be truly transformative on the national psyche – but this is not a poli sci blog and most of the people here are New England focused so.)
And above all else, again, what I’ve said is: if you’re staking the future of rail in America on a singular transformative project to make people believe, it has to be something that is actually transformative then. Not “Shinkansen-style,” not dragging the NEC into the 21st century, not even Alon’s forthcoming timetabling project for maximizing NEC service if everyone gets on board: you have to deliver the Tokaido Shinkansen or something better than it, which I have been assured is a ridiculous thing to ask for and as far as the built environment around the NEC is I generally agree with that.
So fix Metro-North, fix the dispatching, fix the regulations and the funding structure and the fare tables; all of that is what I want, and little to none of it requires a $24 billion cash outlay that’s too much money to spend on incremental improvements when actual transformative change can be had for less.
I’ll circle back here to a previous comment by Michael, about how the small-scale stuff “wouldn’t have an impact” – I disagree. $24 billion is more than enough to put all of Amtrak Midwest on 2 tph plus Ohio services that are competitive with driving. None of it would be HSR. That doesn’t make it “not transformative.” In fact, defeating the popular meme of the Northeast being the only place rail can work would be extremely transformative, and demonstrating that you can deliver a better rail experience than driving at 1% of the cost of an American Shinkansen means that when everyone wakes up and goes “shit, when do I get my train to the city, that thing is cool” it’s much easier to keep momentum going.
You obviously don’t understand different classes of service. If any ticket is good on any .. vehicle… there is only one class of service.
Yep, you sure got me! Obviously referring to the local train and the express train as two different classes of service is wrong and stupid. “Class” only ever means coach or business or comfort+ or any number of other stupid brands that dictate the feel of the seat or whatever.
Surely it can’t be that equating faster trains versus slower ones to literally first and second class tickets is stupid and one of the biggest problems with Acela 1.0. No, I’m the one who is stupid. Good job! You owned me! Great work!
Acela has two classes of service ….on the SAME train. Neither of them coach.
I’m very aware that you desperately need me to be wrong, because it’s a problem for you that I’m not.
And to a large extent, I’m even sympathetic with your position, because in so far as you’re saying we just don’t build enough and what we do build we overpay for, you’re completely correct.
No, the problem is where “just build more” is the sum total of the ‘solution’ provided and there’s a complete repudiation of any attempt to even consider anything that isn’t putting a shovel in the dirt. We do have to build more – but we can’t simply build our way out of the problem. We have to get better at building effectively – but we do that through a strong state public works arm managing a steady stream of small projects and we can’t do that if all we do is lurch from megaproject to megaproject, throwing record-setting amounts of money into a black hole, learning nothing.
And the problem is that even though we spend $10 for every $1 that a capital project should cost, we are still at least spending that money; for every $10 we should be spending on maintenance and operation, we spend $1 – or less.
And so yes, by those metrics, by the actual reality of US politics and infrastructure, HSR isn’t justified anywhere except for California. Because yes, maximizing utilization of infrastructure is an important goal, and that means running more trains, not more route-miles. $5 billion to buy 120 km of high-speed parallel track because it “saves 24 minutes” but doesn’t serve anybody who wasn’t already served by the underutilized and under-maintained legacy tracks half a mile of way is absolutely a waste of money, and in talking about it like it’s the baseline and not in and of itself a concrete megaproject, we are not solving the problems of the US rail network – we’re perpetuating them.
Because there’s an entire country outside of the Northeast Corridor and it’s already full of serviceable rail lines that should be seeing hourly service, but any plan to extend rail to new places or increase rail frequencies is announced in terms of new trips per day.
Because $43.5 billion dollars was granted to passenger rail investment, of which the one functional train line in the country today got half of it and the rest of the country has to fight over table scraps.
Because we’re all collectively choking to death on a built environment that prioritized the automobile over everything else, but getting people out of those cars is seen as a lesser goal than getting them off of planes.
Because actually fixing the problems with rail in this country requires rolling up our sleeves and doing a lot of thankless and unsexy work, which doesn’t spark joy the way lurid fantasies of a nation blanketed by 300 kmph track drawn in colorful crayon lines does.
That’s why I’m right, regardless of how desperately the build-just-build-baby-BUILD contingent needs me to be wrong.
(Can you answer in-thread in the future?)
The politics of the NEC vs. the rest of the US buckets is to some extent beating a dead horse – the money’s allocated between these two tranches and the goal of our project is to get HSR on the NEC on the available budget.
The other issue is that it’s very easy to spend money and have nothing to show for it. The splitting of the rest-of-US money between a bunch of different lines serves that purpose: some money to California HSR but still not enough for the Initial Operating Segment (let alone LA-SF), some to Brightline West but nowhere near enough to complete the line, etc. So the program will not produce any complete line that can get serious ridership, and people will learn to associate it with failure, as they did the Obama-era effort for California HSR and for the medium-speed lines rebranded as higher-speed rail.
(I will try to – I’ve been avoiding it in cases where the thread is already too deep and the reply chain is maxed out, because most of the problems I’ve experienced with comments going to the wrong place happen at that point.)
Spending money and having nothing to show for it is the main thing that motivates me to be here, and to pick a name like Transit Hawk; but to be clear, it’s the “nothing to show for it” that I have an issue with – I’m more than happy to spend the money to get the actual return. In particular, much of the efforts to upgrade medium-speed lines to higher speed rail are the kinds of things I want to see, and my repeated complaining about a lack of operational investment is because that work produced Amtrak Midwest, whose most frequent service is the Hiawatha that has 6 round trips per day despite tracks with capacity for – at an absolute minimum – 2 round trips per hour; this is the primary reason I consider that entire program to be a failure and I suspect that the majority of people who want to be on the train are instead getting in their cars because the train never shows up. Put another way, I insist including on the time spent waiting at the station because your train was inconveniently timed as part of the overall speed of the line and hence going from 1 train per 3 hours to 1 train per hour represents a savings of up to 120 minutes.
And I do get that coming into Northeast focused projects and saying “actually the money needs to go somewhere else” is the kind of fundamental disagreement that is never going to be bridged over – but even on the NEC, my primary concern is or would be delivering maximum frequency over maximum speeds; which $5 billion to build a parallel line next to the existing one which isn’t or wouldn’t be fully utilized does not do. I do make the exception for when spending that money creates a service which doesn’t exist at all currently, hence why I’m willing to allow for Hartford-Providence (it’s certainly not because I like either municipality: to be clear, I don’t) but in general my position has always been and will always be that every dollar spent on new track at this point has to be justified against spending that dollar on either rolling stock or crew or wires – or, as in Virginia and North Carolina, purchasing the tracks that exist already.
But I also get that it is beating a dead horse at this point. I’ll keep doing it… but I get it.
Amtrak should put you in charge of all schedules and routings. While they are doing that, why not rename ‘Amtrak’ to ‘RailAlon’? Cheers Mate!