Amtrak's New Acela Trains Are Here. They're Moving Slower Than the Old Ones

32 JumpCrisscross 63 8/29/2025, 3:54:44 PM wsj.com ↗

Comments (63)

pempem · 3h ago
This is so intensely misleading.

The trains are new. The wires etc they run on aren't. When there's funding to update those, the trains will run faster.

If this was done in the opposite way, we likely would still be working on infrastructure (it's a bigger project) and the trains would still not run faster cuz they would be old.

We need to hold the fourth estate to better account.

legitster · 3h ago
The B1M put together some good content on the particular infrastructure problems that affect the route: https://www.theb1m.com/video/northeast-corridor-baltimore-tu...

The route relies on some really old infrastructure, including single span bridges and tunnels that predate TBMs.

octaane · 4h ago
octaane · 4h ago
Important context buried at the bottom:

"Passenger rail analysts said that it isn’t unusual for rail operators to pad their schedules when introducing new trains, especially if they will run alongside older ones. For example, during a rollout, a railroad might schedule more generous dwell time at stops, they said."

Also, the new schedule is only about 15 mins longer than the old one...this is a big nothingburger.

docdeek · 3h ago
You’re right on this point, but my read of the article was that you can buy the fastest trains in the world but if your infrastructure is not up to enabling them, they won’t go much faster than the ones you have already.
eqvinox · 3h ago
The headline — which, let's face it, will be the only thing a whole bunch of people are going to read — really makes it out like they bought shitty trains, or even that trains can't be better. Which is just untrue.
CGMthrowaway · 3h ago
>"Passenger rail analysts said that it isn’t unusual for rail operators to pad their schedules when introducing new trains, especially if they will run alongside older ones. For example, during a rollout, a railroad might schedule more generous dwell time at stops, they said."

Why? For marketing purposes?

scienceman · 3h ago
I would guess to account for any unfamiliarity from the operators (new systems, etc) and allowing more time to sort out any other kinks. (they also pad the schedules for old trains too -- this is to accommodate small slowdowns without causing cascading delays).
bobthepanda · 3h ago
If you run the new trains at maximum potential then they will just catch up to the old train in front of them and then have to maintain the old train’s pace. So during the transition period you couldn’t really run the new train faster anyways.
jasonpeacock · 3h ago
New trains: padded schedule

Old trains: padded schedule

So really, all train schedules are padded - which makes sense, you need buffers to absorb variance in performance to have reliable schedules.

eqvinox · 3h ago
> Old trains: padded schedule

No — Old trains: schedule based on experiences from having ran them for at least a year (i.e. all seasons)

New trains' buffers are larger because you don't know e.g. how shit the brakes are when you have tons of leaves on your rails. (Yes this is an actual thing¹.)

[¹ Ed.: in case anyone is incredulous at the leaves thing: https://www.groupe-sncf.com/en/group/behind-the-scenes/traff... ]

SoftTalker · 3h ago
The contact surface area of a steel wheel on a rail is about the size of a dime. That's what you have to work with to stop the train.
CGMthrowaway · 2h ago
To drive the train. But to stop the train, surely the contact surface of the brakes with the wheels is more important?
jimktrains2 · 2h ago
Just like on a car, if your break pads are capable of completely stopping the wheel, but the wheel is not capable of creating enough braking force with the surface, then you slide. If you lubricate between the wheels and surface, which is essentially what leaves do on rails, no amount of at-wheel breaking power will stop you.

This is what antilock Brakes in cars prevent, they pulse the brakes to allow the tire to regain traction, preventing slippage and loss of control.

tialaramex · 1h ago
Note that a modern train also has the same system, only more so, a car would typically have 4 wheels, I think the Acela probably has like 96 or something.
tialaramex · 1h ago
There's also some "infant mortality" stuff, the EMUs (Electric Multiple Units, which I suppose is roughly what this is?) in my country have not good failure rates in their first weeks and months but they get much better (typically best in class due to fewer moving parts than say a diesel) in mid-life.

Brake performance is a thing you can simulate and measure on a test track, which presumably happened at least months ago. However loss of adhesion, the reason braking stops working on contaminated railheads is also impacted by moisture, a bunch of dry leaves won't make anywhere close to the same problem as the same leaf material after a nice gentle drizzle, not really rain per se, not enough to actually wash the rails clean, but ensuring the leaves turn into a thin mush that makes braking next to impossible.

RAIB report 12/2023 about an incident near me talks about that, the driver maybe makes some dubious decisions, but ultimately he brakes, and it does nothing, so he brakes harder, still nothing, maximum braking, still nothing, select emergency braking (same effect but hey, it's there to be used right), still nothing - oh shit, we've passed the danger signal and I see another train, time to leave. He actually spent ages in hospital because he tripped trying to flee and was trapped in the wreckage, but on the other hand if he'd just sat there frozen he might well be dead 'cos his train did indeed smash into the other one so the side where the driver is sat smashed into another train at like 50+mph.

JumpCrisscross · 3h ago
> the new schedule is only about 15 mins longer than the old one...this is a big nothingburger

Then why the new cars? Are they cheaper to run? If so, the real price of a ticket (or losses [1]) should go down.

[1] https://www.amtrak.com/content/dam/projects/dotcom/english/p...

adeelk93 · 2h ago
The train was designed with a 20 year service life and is approaching 30. We do need new trains, even if they didn’t bring about any improvements.
OneDeuxTriSeiGo · 2h ago
The new cars hold far more people, are faster, and they are supposed to be paired with comprehensive upgrades to the extremely aged infrastructure and tracks along that route.

Both the cars and the overhaul on the route were funded by the Biden admin (primarily with funds allocated by congress between 2021 and 2023) with the intent that modernisation to the route would gradually allow the cars to run at speed over the next few years. Now it's unclear what the status is of this infrastructure overhaul with the current administration suspending funds for the previous admin's projects.

craftkiller · 3h ago
Quoting from the brief article you didn't read:

> Amtrak previously has said the new trains could potentially shave 20 minutes off the travel time between New York City and Washington, D.C., reducing the duration to around 2½ hours

> because of Amtrak’s old infrastructure, the new trains can only travel at top speed during certain portions of the journey [...] Infrastructure upgrades planned over the coming years will improve trainspeed and reliability, the railroad said.

> Each new train will hold 386 passengers, an increase of 27% compared with the current fleet.

Therefore: Faster trains carrying more people. Just not faster on day 1.

bobbylarrybobby · 3h ago
US rail is so pathetic. Not the rail’s fault per se, but it's pretty embarrassing for a a supposedly wealthy country to not be able to offer its citizens cheap, easy, high speed travel across the country.
toast0 · 3h ago
We do have cheap easy high speed travel across the country. It's just not rail.

In the US, rail is for freight. In the US, cross country distances are too large for terrestrial transport to be high speed. A flight network allows for much better speed, without needing to have continuous infrastructure.

Rail in general requires pretty specific alignment and limits on turns and slope; high speed rail has even tighter tolerances. It's very expensive to build that, especially through rocky terrain, and through existing development.

Flight networks are much more flexible. If you have room for an airport, and demand, you can get direct flights to/from anywhere. Roads are good too; because of their utility, they have a large network which tends to offer better routing than rail, and piecemeal upgades work for roads; switching routes is much easier for cars than trains, and roads can be quite rough but still usable at low speeds.

saubeidl · 3h ago
China is about the same size as the US. Here's a map of their HSR network: https://www.travelchinaguide.com/images/map/railway.jpg

Flights are nowhere near as efficient as rail, not to speak of the environmental issues.

It's factors like this that show how the US is falling behind in international competition.

toast0 · 2h ago
Yes, China is about the same size as the US, but a route from Beijing to Kashgar is 23k feet up and 21 k feet down. A route from Washington DC to Los Angeles is about the same distance, but has 77k feet up and 77k feet down. A lot more elevation to work through. Much of the east side of that map is relatively flat. And there are bottlenecks getting to the west side, as there would be in the US, but the US has a lot more relative population on the west side than in China.

I'm sure that rail wins over flights in terms of efficiency for passenger miles traveled. However, you often have to travel more miles to get to your destination. A direct flight can save a whole lot of time. Flights are much faster, although security theater adds a lot of time[1]; but for longer routes, or routes with many stops or connections for rail, the single segment operation of a flight starts to win.

Also, China has the advantage of authoritarianism which allows it to more easily get right of way that's well aligned for HSR. While it was once easy to get land for railroads, when there were active land grants, now you'd need to do a lot of work to get well aligned land to add a rail line; where they're still active, existing rail right of ways through developed areas are fine enough for slower service, but expanding the right of way to better align the tracks is a lot of work. In theory, you could eminent domain, but that process is long and expensive.

[1] In a world where you think it's possible to have government intervene to expand passenger rail in the US, it's equally possible to have government intervene to streamline the security theater in flights; or they might add the same security theater to passenger rail.

M95D · 36m ago
They're building a rail line Taipei to mainland? :-O
jcranmer · 2h ago
That's not a map of their HSR network, that's a map of their rail network.

Openrailwaymap plots tracks by max speed, and you can see the map of China here: https://openrailwaymap.org/?style=maxspeed&lat=30.9587685707... --the tracks that are red-to-purple are the ones that are high speed.

saubeidl · 2h ago
I'm having a hard time reading that map - but this one also goes by speed and shows the extent of the HSR lines: https://en.wikipedia.org/wiki/High-speed_rail_in_China#/medi...
legitster · 3h ago
The area in question (Northeast) is about the only place in the US that matches the population density of the parts of Europe served by high-speed rail.

It's not a question of wealth but of efficiency. Even if you took every single meter of high speed rail that France has built in the last 50 years, you would barely cover 2/3 of the distance of a single line between New York and LA.

rafram · 3h ago
New York to LA would be a terrible high-speed rail route. Just way too far. But the US has many city pairs - SF and LA, LA and Vegas, Denver and Salt Lake, Chicago and Detroit, Pittsburgh and Philly, and so on - where high-speed rail could easily replace most flights if it were built.
legitster · 2h ago
"Premature optimization is the root of all evil".

Some of those routes sound nice, but they don't usually have enough daily traffic between them to warrant the massive infrastructure costs. LA to Vegas would rival the distance of the longest European high speed lines, but with many fewer stops between and a fraction of the passengers between. Outside of connector flights, I don't even know how many people even regularly travel between Chicago and Detroit.

Part of the advantage places like Japan and Europe have over the US is not just that they are dense, but that their population skews to one massive metropolitan center. So it's easy to design a hub model where traffic patterns are easy - in and out of one area. You see this in the success of regional commuter rail networks in the US.

(SF and LA is a great candidate though, and it's failure a highlight of how hard it is to build in the US)

rafram · 54m ago
> LA to Vegas would rival the distance of the longest European high speed lines, but with many fewer stops between and a fraction of the passengers between.

Many fewer stops: that's a good thing. A fraction of the passengers? No, that's absolutely false. Brightline estimates 50 million trips between LA and Vegas every year [1], which is more than, for example, the total number of tourists to Paris every year.

> their population skews to one massive metropolitan center

Japan, sure, being a long and skinny group of islands definitely helps. But is that actually true of Europe? Germany has a great high-speed rail network (current on-time performance issues aside) with a very pluricentric population.

[1]: https://www.brightlinewest.com/overview/project#:~:text=Toda...

legitster · 36m ago
On their website, total traffic is 50 million, of which Brightline only anticipates about 9 million riders annually.

Their pricing is expected to be over $100 a person one-way: https://www.sfgate.com/travel/article/ticket-prices-leaked-h...

So this puts it at pretty low ridership and much higher ticket prices than equivalent lines around the world. (I'm very supportive of the project! It's just worth being realistic about the economics here.

My understanding of Germany's high speed network is that it is pretty comparable to the Acela - high speed in certain corridors, reasonably affordable, and dense population along the corridors. Which is pretty realistic goal for parts of the US. But often we compare to France's network which really outperforms nearly every other country outside of Japan.

tantalor · 3h ago
Florida has high speed rail

https://en.wikipedia.org/wiki/Brightline

legitster · 3h ago
It's not technically fast enough to be called high speed rail: https://en.wikipedia.org/wiki/Higher-speed_rail

But it is worth calling out that good regional rail services exist in the US, and they don't even have to be that fast to be successful!

allturtles · 3h ago
With an average speed of 70 mph, the Brightline isn't high-speed as it would be defined elsewhere in the world. Neither is the Amtrak Acela, for that matter.
dc396 · 3h ago
Yes, US rail is pathetic, at least for passenger travel. However, a quick search on Google Flights shows a one-way ticket from LA to DC costs $98.00 on Frontier via Atlanta (YMMV) so I don't think it's correct to say there isn't cheap, easy, high speed travel across the country.
legitster · 3h ago
It's also worth pointing out that even at TGV speeds, a coast-to-coast train ride would take ~20 hours. It would have to be sooo much cheaper than an equivalent plane ticket to be worth it.
ACCount37 · 3h ago
If we price the extra travel time at the level of US minimum wage, $7.25/hr, the extra ~15 hours of travel alone would add up to a price tag of $100. And if you don't have a bullet train, then the time-economy gets even worse.

That sure seems like it explains a lot about the state of US passenger rail.

saubeidl · 2h ago
You could do a sleeper train one can work on. No loss of working time.
ACCount37 · 2h ago
Which makes the train considerably less passenger-dense, and thus makes the train tickets more expensive. In return for reduced value hours: hours that aren't as productive as real office hours, or as entertaining as real vacation hours.

I'm not sure if this would be enough to make coast to coast passenger trains more viable.

bombcar · 3h ago
This is exactly it.

The USA has solved two problems: cheap and easy travel across town (in your car) and cheap and easy travel across the country (in a plane).

Both COULD be done by rail, and some places do - sometimes on the same rail network.

And when the ONLY option for transcontinental travel was rail - the USA has some impressive shit.

Given the limitations and money available it’s surprising it’s as good as it is.

mperham · 3h ago
> cheap and easy travel across town (in your car)

It's only cheap and easy if you discount omnipresent traffic jams, 50,000 deaths per year, ever-worsening climate change and the inability to walk/bike anywhere anymore.

Aside from that, it only costs the average American $1,000/mo to keep that cheap and easy transportation!

mrpopo · 3h ago
It's also 700kg of CO2, one of the best ways to worsen climate change per dollar spent
dc396 · 2h ago
Ignoring the point that climate sensitivity wasn't in the parent comment, AFAIK, airplanes generate 0.16 kg/km, whereas trains are around 0.1 kg/km and container ships are at 0.016 kg/km. However, passenger ships and gas cars are at 0.25, diesel cars at 0.28 kg/km and rockets are over 1.0 kg/km, so it appears planes are in the middle, not really "one of the best ways" to worsen climate change.

Sure, through a simple analysis of CO2 kg/km, trains are better for the climate for long distance travel, but they are vastly slower (average time from LA to DC is over 80 hours), which has knock-on effects, e.g., sleeping at home averages 0.25 - 0.32 kg/night, whereas staying at a hotel averages 10-40 kg/night, eating at home averages 2.3 kg/meal vs. 3-8 kg/meal, etc.

SoftTalker · 3h ago
What is it per passenger-mile, that's what you would have to compare to e.g. a car, bus, or train.
lapetitejort · 3h ago
246g of CO2 per passenger kilometer [0]. Versus 35g for national rail

[0]: https://ourworldindata.org/travel-carbon-footprint

idiomat9000 · 3h ago
The rivers are to good as mass good transport medium. They make goods transport bx rail economically unviable for the core country, leaving only a ring of economic routes.
ianburrell · 3h ago
That is weird thing to say when US has the best freight rail in the world. Lots of containers are transported by rail from west coast to east coast ports. The competition is trucks, not barge.

If anything, that is one reason that passenger rail sucks because the freight companies own the rails. They delay passenger trains and don’t get punished. The freight companies don’t spend on maintenance, and have pulled up double track.

eqvinox · 3h ago
[Regardless of validity, which is questionable—]

Why are you bringing cargo into this? The argument was about passenger travel…

JumpCrisscross · 3h ago
> US rail is so pathetic

National passenger rail.

We still have the world's largest rail network [1]. It's just focussed on freight. (We also have world-class municipal and regional rail in the New York area.)

[1] https://en.wikipedia.org/wiki/List_of_countries_by_rail_tran...

jcranmer · 3h ago
> (We also have world-class municipal and regional rail in the New York area.)

If by world-class you mean something that most of the world would consider a rail system, yes (in comparison to most US cities' attempts which aren't really worth even pretending constitutes regional rail). If you mean something that most of the world would consider as something worth emulating, nope.

saubeidl · 3h ago
I'm not sure municipal rail in NY is quite world-class, tho it is by far the best the US has to offer.
2OEH8eoCRo0 · 3h ago
Most people don't want trains. Most places where people want trains have them (high density places).
mrpopo · 3h ago
Trains are just the most efficient way of moving people between cities. They benefit everyone, even people with cars.

You have a business trip and need to go from A to B by yourself? Take a train, it frees your brain and the highway for people traveling in groups or with lots of luggage.

Incidentally it also avoids moving 2 tonnes of material for no reason.

bombcar · 2h ago
A passenger train in the USA weighs about 1000 tons, plus another 150 or so tons of locomotive, before we add the passengers (seating 80 per car and about 11 cars for 880 passengers).

That’s about a ton per passenger.

kjkjadksj · 25m ago
To be fair an EV would be at least two tons per passenger. I guess a 1990s honda civic is still one if the most efficient means of vehicular travel.
ReptileMan · 3h ago
>Trains are just the most efficient way of moving people between cities.

If the cities are closer than 600 km. Trains are good for movement inside a typical state, but rarely between states.

2OEH8eoCRo0 · 3h ago
You're preaching to the choir but the average person doesn't care.
abap_rocky · 3h ago
One of the most recent inter-city routes to open, the Borealis service between Chicago and St. Paul, far surpassed expected ridership levels in its first year servicing 212K passengers over a projected 155K [1]. This comes despite the fact the trip would be faster not only by air but also by car. I doubt we'd see this sort of overperformance if "most people don't want trains".

[1]: https://www.news8000.com/news/amtraks-borealis-line-celebrat...

dfxm12 · 3h ago
Most people don't want trains.

Care to flesh this out?

I can only find data that contradicts this:

https://railpassengers.org/happening-now/news/releases/new-p...

https://media.amtrak.com/2023/08/data-finds-overwhelming-sup...

Most places where people want trains have them (high density places).

This doesn't address OP's comments. To remind you, they mentioned "cheap, easy, high speed travel across the country". I would grant Amtrak is easy. It's not the other things though...

garciasn · 3h ago
It’s a chicken or egg problem. Unfortunately, because train travel is slow and expensive, it’s simply less expensive and faster to fly.

If the rail infrastructure was upgraded to allow for faster travel and the costs were lowered, Americans would find them more desirable—-guaranteed.

ryoshu · 3h ago
I'll take the Acela over flying any day. It's cheaper and less of a hassle. Time-to-commute is a wash when you do the airport security theater dance + delayed flights on the tarmac.
kylehotchkiss · 3h ago
I've found that looking at USA as a middle-income country with extremely wealthy regions helps me better understand our country. The supposed wealth doesn't seem relevant in the hollowed out towns between the coasts. It's generally not being sent or spent there in any meaningful way.

Much of the actual wealth is in the northeast corridor, where the Acela is improving the quality of their product, and in California, where Southern CA and the Bay Area both have fairly nice regional trains that go as fast as the right of way will let them. I think we've done OK connecting places that matter, and as flying continues to enshittify and people will choose rail for comfort.