Home High-Speed Rail On Amtrak, infrastructure upgrades and costs

On Amtrak, infrastructure upgrades and costs

by Benjamin Kabak

Amtrak released its 2012 agenda today. You can read the press release in all of its PDF glory right here. One of the key initiatives concerns its profitable Northeast Corridor service, and I wanted to highlight it for a brief moment. Says the press release:

160 MPH HSR UPGRADES IN NEW JERSEY In 2012, Amtrak will advance design, engineering and other pre-construction activities for a $450 million project funded by the federal high-speed rail program that will boost top train speeds from 135 mph to 160 mph along a 24-mile section of the NEC between Trenton and New Brunswick, New Jersey. The project supports the goals of the Gateway Program and includes upgrading track, electrical power (frequency converter capacity and additional substations), signal systems and overhead catenary wires to permit the faster speeds and also reconfigures track switches at the western entrance to New York Penn Station to mitigate congestion issues. Major construction work will begin in 2013 with project completion expected in 2017.

So let me get this straight. Amtrak is going to spend nearly half a billion to improve their trains’ top speeds by 25 miles per hour along a 24-mile-long section of track. And it’s going to take another five years for this project to finish. No wonder an American high-speed rail line — let alone an entire network — can’t get off the ground.

You may also like


Lawrence Velázquez January 11, 2012 - 4:38 pm

Let’s make an unrealistic and certainly incorrect simplification, which is that trains travel at their top speed along that entire length of track.

24 miles ÷ 135 mph = 10 2/3 minutes

24 miles ÷ 160 mph = 9 minutes


Eric F. January 11, 2012 - 5:22 pm

Right, but the Acela Express estimates a 65 minute travel time from NY – Phl. Nine fewer minutes would represent a reduction of nearly 15% in travel time. It would take about 5% off the duration of a NY-Washington trip. Those are pretty significant travel time reductions.

Eric F. January 11, 2012 - 5:30 pm

Oops, my bad math! I see you are talking about a reduction of less than 2 minutes! Yup, not so much. Although I bet there will be some system reliability improvements that help out matters beyond headline travel time.

Douglas John Bowen January 13, 2012 - 1:31 pm

Eric F. is on the money here: System reliability, along with corresponding system speed improvement potential for all rail traffic on this stretch, not just Acela trains. And for those intent on saying such an expenditure isn’t worth it: What other kind of catenary upgrade would please?

Nathanael January 16, 2012 - 7:55 pm

To be very clear, the increased top speeds are essentially marketing. It’s included because it’s easier to get money by advertising high top speeds.

The project is for:
– upgrading track
– electrical power (frequency converter capacity and additional substations)
– signal systems
– and overhead catenary wires
– reconfigures track switches at the western entrance to New York Penn Station

Amtrak already needs to upgrade the signalling, replace the overhead catenary, replace the switches at the west end of NY Penn, add additional electrical substations, and add an additional frequency converter (or change the entire system to a different frequency) — and repair the tracks.

So they got a maintenance/upgrade project funded by advertising it as a top speed project. Good for them.

Martin June 23, 2012 - 4:33 am

You are missing the point – it is replacing life expired catinary,signaling and electrical equipment over 100 years old.

jim January 11, 2012 - 4:39 pm

Yeah. Close to $20M/mile to retrack, repower and resignal to essentially high speed rail standards. That’s more or less the ballpark for work that’s to be carried out on currently operating track. They’re building to 300 km/h standards, but their current rolling stock can only manage 160 mph.

Adirondacker12800 January 11, 2012 - 5:10 pm

Or 5 million track mile – it’s a four track railroad roughly from Newark NJ to Wilmington DE. The part about new and improved substations should affect reliability greatly. So should reworking the switches in Penn Station.

Alon Levy January 11, 2012 - 6:18 pm

The bulk of HSR cost is civil infrastructure, i.e. things Amtrak doesn’t even have to do.

jim January 11, 2012 - 8:34 pm

Powering and signaling an already tracked California Initial Constructable Segment is estimated at approximately $10M/mile with no trains running along it at all.

Alon Levy January 12, 2012 - 7:00 pm

Sure, but Amtrak doesn’t need to do even all of that. It needs to modernize the catenary, primarily. The substations are another matter – the voltage needs to be switched to 25 kV 60 Hz, but my understanding is that the project doesn’t do that. As for signaling, I’ll try to look up whether that portion of the NEC is already ACSES-ified. I believe it is, but I’m not sure.

Nathanael January 16, 2012 - 8:14 pm

Signalling: that section has to have the PTC upgrade, which I think is called “ACSES II” (though I may be mistaken).

I am really unclear on what Amtrak is doing with the frequency stuff. It looks like they’re maintaining the 25Hz frequency, which is silly, *but* sort of makes sense, because they really don’t want to add extra frequency conversion points, and they have no funded scheme to convert Penn Station.

So my question is whether they are redoing the wiring to be *ready* to convert the frequency — immunizing the signalling against 60Hz, checking the clearances and making the wire diameters suitable for 25kV 60Hz, etc. If they are, I’m OK with that. If they’re not, it would be stupid to lock themselves into this frequency with new work.

Alon Levy January 16, 2012 - 11:20 pm

I had a nice presentation about ACSES, which claims it is PTC (which it is). It also listed which parts of the NEC are ACSESified – about half of the route does, but it’s a discontiuous half, and I don’t remember whether the Jersey portions are. Alas, I can’t find it anymore.

Nathanael January 17, 2012 - 11:41 am

Yeesh. I also don’t have the report which says which portions of the NEC are fully ACSESified. Wish I did. 🙂

al January 12, 2012 - 1:54 pm

Higher speeds through switches would be a boon for getting trains 10-14 cars long through quickly. If NJ Transit and AMTRAK could get together and upgrade signalling, the North River Tunnels can run 40 tph. Both Secaucus Junction and NY Penn have enough tracks and platforms to run that many trains through the tubes. Upgrades to catenary along that section will reduce downed wires events. There is also enough space for a siding near the NJ entrance (near Tonnele Ave) to the North River Tunnels for rescue locomotives. The same applies for on the other side. There’s a track between the southern pair tunnels that LIRR use to access NY Penn that dead ends. You can fit a pair of locomotives there too.

Bolwerk January 11, 2012 - 4:40 pm

In all fairness, this is a pretty hefty upgrade. It’s probably past due, since it’s not unlikely much of that infrastructure dates to the 1930s. It all needs to be done while the railroad more or less continues to operate. It is also surely much more expensive than it needs to be, but not shockingly expensive by U.S. standards.

I think it’s much worse that just the year-to-year maintenance costs on the NEC are some of the highest in the world. And the NEC isn’t even that busy by European or presumably Japanese standards.

zz January 11, 2012 - 4:42 pm

These projects are expensive because this country did not have the foresight to reserve sufficient right-of-way for future improvements, or alternative rights-of-way for development of new corridors. We must now upgrade our existing system incrementally in the same ROW as an existing system, without disrupting service. What do you expect?

In Japan, where rail companies seize lots of extra real estate to cash out its value over time, and in Europe, where they don’t have suburban sprawl taking up every last developable square inch, miraculously it’s much cheaper to make intercity rail improvements.

Alon Levy January 11, 2012 - 6:22 pm

What do I expect? For a start, I expect Amtrak to be able to shut down an underused four-track mainline two tracks at a time. The traffic south of New Brunswick is such that everything could be accommodated on two tracks, except maybe during rush hour (I haven’t checked) and even then three would be enough. More to the point, let’s look at how much (or, rather, how little) it cost Germany, generally a high-cost country, to upgrade Berlin-Hamburg to 230 km/h, giving it an average speed of about 190 km/h. If I remember correctly, it was €18 million/km, including grade crossing elimination.

Also, in Japan, rail companies don’t seize extra real estate for intercity rail. They buy it out and develop it for commuter rail.

Adirondacker12800 January 11, 2012 - 8:19 pm

Who says they aren’t going to be shutting down a track at a time?

Nathanael January 16, 2012 - 8:17 pm

Look at the fairly recent FRA rules about clearance between running tracks and tracks having work done. Then check the track centers on the NEC. Serious non-Track-Renewal-Train work on either of the middle tracks == single-tracking the corridor 😛

simhedges October 13, 2012 - 12:56 am

“in Europe, where they don’t have suburban sprawl taking up every last developable square inch”. Look at population density in the UK compared to the US – it’s a lot higher. If it’s not suburban sprawl it’s villages and small towns every 2 or 3 miles – and in between are “Areas of Outstanding Natural Beauty” which much be protected where possible. Why is why the 100 mile long proposed new HS2 line from London to Birmingham will cost about $40bn. It makes the Amtrak upgrade of the NEC look positively parsimonious.

Eric F. January 11, 2012 - 5:27 pm

“No wonder an American high-speed rail line — let alone an entire network — can’t get off the ground.”

I would argue that The Big O did implemented his HSR-love in a very poor way. If he applied just a tiny fraction of the 2009 stimulus bill spending to the NEC, he could have accomplished quite a few projects, many of which would be nearing completion by now. I share your frustration, but it represents not some metaphysical lacking in the U.S., but rather a strategic failure by a guy who likes HSR in theory but doesn’t really understand how his own country works or what kind of infrastructure is currently existing in it.

Bolwerk January 11, 2012 - 9:21 pm

Maybe, but the same can be said for most of the U.S. rail industry and political establishment. And NIH syndrome pretty much precludes hiring people who know what they’re doing, like the French, Germans, Japanese, or even Chinese. So the competency isn’t here now, and gov’t regulations are preventing us from developing it.

Also, spending additional money in the northeast is mostly a political no-no. The Empire Corridor would have been sense for HSR too, but that would have involved returning to New York some of its own money.

Nathanael January 16, 2012 - 8:23 pm

Actually, we got some good stuff from the HSR money between Poughkeepsie and Schenectady. Signalling upgrades the whole distance, double-tracking the part which wasn’t, and a fourth station track at Albany-Rennselear (which should allow late trains to stay out of the way of on-time trains).

So far, no politicians are paying attention to the accelerating increase in passenger traffic demand between Niagara Falls, Buffalo, Rochester, Syracuse, Utica, Rome, and Schenectady, but maybe they’ll notice eventually. The Schenectady-Albany double tracking will help those trains’ reliability some.

Scott E January 11, 2012 - 7:17 pm

I’d bet this project is all about the swich reconfigurement near Penn and the replacement of decaying infrastructure along the 24 mile stretch. The “high-speed” component is just a mirage in order to get some federal funding.

Nathanael January 16, 2012 - 8:25 pm

Absolutely right, and good for them to do so. Although it will allow for 160 mph running, largely by replacing decaying infrastructure with more modern replacements (such as constant-tension catenary).

Alex C January 11, 2012 - 8:21 pm

Until the FRA enters the 21st century, high-speed rail will never get off the ground here. They don’t even acknowledge modern high speed rail in their regulations. As of 2011 (correct me if this has been changed) high-speed trains absolutely had to be Acela-style with locomotives at the ends, forget EMUs.

Stephen Smith January 12, 2012 - 12:45 am

Well, they can be EMUs, as long as you vacant passengers from the end cars and fill them instead with weights.

Subutay Musluoglu January 12, 2012 - 7:52 am

While I agree with many of the sentiments expressed here, the fact of the matter remains that the NEC has major state of good repair issues and this should be seen as a down payment. The catenary on the stretch in question dates to the PRR’s original installation in the 1930s. It is long past the date of replacement. Much more needs to be done and it won’t come cheap. There are significant issues with bridges, catenary, track, switches, and traction power substations all up and down the NEC. Estimates for bringing the NEC up to SOGR are estimated to be at least $5 Billion. In this context, the amount expended on the section between Trenton and New Brunswick (also known informally as the “Raceway”) might seem a little high, but it is not unreasonable. It should not be seen as merely reducing running times (which I agree is not that much). That is not the objective here. Upgrading the catenary, tracks, and other NEC components will increase reliability and go a long way towards reducing the frequency of those dreaded announcements that are heard all too often at Penn Station – …”service has been suspended due to downed wires.” Or that “the power has failed.”

I too feel strongly that the HSR program has been misguided at best. While many of the projects around the nation designated as HSR are worthwhile in their own right, they should never have been labeled as such. There should have been a distinction between corridors that are essentially freight lines in need of upgrading, and dedicated true HSR lines. This administration has laudable goals when it comes to passnger rail, but they have committed a huge mistake by sprinkling money all over the nation and lumping it together as HSR. All it has done is given anti-rail critics more ammunition.

Douglas John Bowen January 13, 2012 - 1:40 pm

Actually, any “lumping” may have confused not only rail advocates, but rail opponents as well, and anti-rail critics have been slow, very slow, at times even absent, when it comes to “higher-speed rail” projects such as St. Louis-Chicago, or Empire Corridor upgrades.

Contrarily, one could argue that the Obama Administration took a lesson from Isaac Asimov and both his “Robot” and “Foundation” series of works. Right or wrong, the administration has advanced not one but two initiatives, HSR and higher-speed rail (HrSR). HSR, sure enough, has come under heavy criticism, but (again) HrSR is moving ahead in many places, at speeds it might not move were it a solo target.

Anonymous January 12, 2012 - 8:54 am

Ben misunderstands the issues here. They are upgrading the old catenary wire to constant tension. The most important result of these upgrades will be the improvement in reliability of service with pantographs not being broken by sagging wire during the hot summer months causing countless delays (This is what Metro North has been doing north of NYC and what caused a big problem for them in the summer). The increase in speed is a secondary effect. It is quite bad that Amtrack’s PR department does not understand what is going on because things like this are exactly what politicians are looking for to cut funding.

Peter Laws January 12, 2012 - 12:56 pm

1) No Amtrak train is profitable. There are several NEC-based services that cover their cost of operations. None of these trains come anywhere near covering the cost of the infrastructure they run on. Please don’t be like politicians and conflate capital and operations! 🙂

2) As noted in other comments, the infrastructure in question is well past it’s “best by” date and needs to be rebuilt, as does most of the rest of the corridor west of Connecticut (which also needs rebuilding but is not owned by Amtrak). That some traffic will be able to hit 160 mph is a happy by-product.

3) The work on those switches west of PSNY will likely save more time than the change in the top speed on the track being renewed. However, it’s easier for the reality show generation to understand “the trains will now go 160!” than for them to try do math and understand that it only saves 2 minutes.

Peter Laws January 12, 2012 - 2:04 pm

And by “as does most of the rest of the corridor west of Connecticut (which also needs rebuilding but is not owned by Amtrak)” I meant “west of New Haven in Connecticut”. 😛

Clarke January 13, 2012 - 3:49 pm

Connecticut grade crossings, anyone? The speed hit trains take when they have to crawl across local roads is definitely more drastic than a few minutes in NJ.

Nathanael January 16, 2012 - 8:28 pm

There aren’t very many (11). Most of them are surrounded by track with other speed restrictions (curves etc.) — it’s not really the grade crossings slowing the trains down. The most egregious example of this is Mystic. Obviously it would be great to have a fast path through Mystic, but just you try….


Leave a Comment