Amtrak's own hypothetical national track infrastructure discussion

Amtrak Unlimited Discussion Forum

Help Support Amtrak Unlimited Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Then again - since some of the HST that exist are entirely raised - building like that would solve the problem of grade crossings

In Europe, they're actually sunk into the ground in many cases, making automotive overpasses easier.

In the US, the best case for a brand new build barring other better ROWs is the median of existing freeways. The biggest problem there is the changes in grade and/or curves which would probably lead to more elevated construction.
 
1. The North American rail system is so Third World that expecting to build 320 km/hr (200 mph) high speed rail is perhaps too much of a stretch for any known method of financing it. After all, you need to crawl before you can walk, and you need to walk before you can run marathons. If you can build a system that has 100 - 130 km/hr (60-80 mph) end to end average speed, it can be perfectly competitive with driving, and a lot less expensive to build. Most of this can probably be done by relatively minor expansion of existing infrastructure and additional track maintenance. In the real hinterlands where there are only one or two long-distance trains a day, what is mainly needed are more and longer sidings.

2. That said, there are a few places where the existing rail routes are either so curvy and have such high grades or have out of the way routes that bypass current population centers that some new tracks would be in order. I'm thinking mostly of the routes that cross the central Appalachians, particularly the old B&O main the Capitol Ltd. uses and the old PRR main used by the Pennsylvanian. The slow running over the mountains limits the potential for Baltimore/Washington - Pittsburgh corridor service and Philadelphia - Pittsburgh corridor service, and the old PRR route completely bypasses State College, PA, about the largest population concentration between Harrisburg and Pittsburgh and full of potentially train-riding students. Again, we don't need 160+ km/hr (100 mph) running, it just needs to be as fast or faster than driving.
Great comment. In a recent discussion, a point was made of how much bang for the buck you could get by making 30 mph track 50 mph, and 50 mph 79. Much easier than being obsessed about 110. We can do this. Reasonable speed, reliable, multiple frequency trains are the answer.
 
A decade ago, a BNSF manager once told me that some of the higher revenue trains will pull in $250,000 revenue (not profit/net income, just sales revenue) per departure.

I did some digging a while back (and shame on me, I keep forgetting to bookmark the links where I find this info because I’d love to go back to it and get more information) that gave sample rates of freight train revenues. I want to say that the revenue is in the neighborhood of a couple hundred dollars per car-mile. Basically, they take in a ton of revenue. I seem to recall my calculations that if a long-distance Amtrak train were to match the revenue of a freight train on a per-mile basis, it would require several hundred passengers paying higher-bucket sleeper fares.

They money they get from Amtrak in track access charges is a drop in the bucket, in comparison. Even if delays (and associated penalties) brought their Amtrak income to zero, it still wouldn’t be more painful, financially, than having to reduce freight traffic to let Amtrak through on time. This, right here, is the crux of the issue for why freight railroads don’t like Amtrak. An Amtrak train can take up more than one freight train slot (because of higher speeds, plus, sometimes, the need to accommodate a train going the “wrong way” on a long, single-track railroad when the freights could otherwise fleet their trains), yet brings them very little revenue to do so. Amtrak OTP does okay when the railroad is not full, but when freight traffic picks up to the point where they need every possible slot, OTP goes down the toilet. And it doesn’t matter how “friendly” the railroad supposedly is, either.

Remember when the Empire Builder, running on “passenger-friendly” BNSF had the best OTP of any long-distance train (early to mid 2000s)? Then, suddenly, BNSF realized that their little-used freight line in northern North Dakota (so little used that they were ready to abandon a part of it through Devil’s Lake, and Amtrak was drafting up plans for a reroute) was useful for all of the oil fracking going on in North Dakota, and almost overnight, the Builder’s OTP dropped to nearly zero. It got to the point where Amtrak had to add an extra consist to the Seattle & Portland end because otherwise trains had no chance of departing the west coast on time.

Take a freight train pulling in $250,000 on a route half the length of the Empire Builder, and then try to see what a passenger train would have to earn in fares to even come close to that, and the math becomes very difficult to pencil out in favor of the passenger train in any scenario where both share the same tracks and you are facing potential traffic that is meeting or exceeding those capacity limits.
So, $250,000 fines for each instance of delaying a passenger train would equalize that calculation pretty quickly. Seems advisable.
 
2. That said, there are a few places where the existing rail routes are either so curvy and have such high grades or have out of the way routes that bypass current population centers that some new tracks would be in order. I'm thinking mostly of the routes that cross the central Appalachians, particularly the old B&O main the Capitol Ltd. uses and the old PRR main used by the Pennsylvanian. The slow running over the mountains limits the potential for Baltimore/Washington - Pittsburgh corridor service and Philadelphia - Pittsburgh corridor service, and the old PRR route completely bypasses State College, PA, about the largest population concentration between Harrisburg and Pittsburgh and full of potentially train-riding students. Again, we don't need 160+ km/hr (100 mph) running, it just needs to be as fast or faster than driving.
I've crayoned-out a route through State College -- basically requires a tunnel under State College proper (5-10 miles, probably cut and cover) and tunnels from there to Lewistown (5-10 miles to pass under Penn Roosevelt State Park), while the rest of it is easy. It would take much more work to straighten out the rest of the line (West of Altoona and the curves between Lewistown and Harrisburg) but I think the value of stopping at State College alone would be quite substantial. PA state government has not, so far, agreed. (*sigh*)
 
Construction technology has improved, but physics hasn't changed. 19th century technology is NOT why railroads follow rivers rather than go straight over mountains.

If you substitute "straight under mountains" for "straight over", there is a famous tradeoff related to speed (and early-19th-century technology was slower, obviously). Early designs aggressively avoided grades in favor of curves, which is right for low speeds and heavy loads.

But for high speeds, you have to aggressively avoid curves in favor of grades; high speed trains can handle steeper grades, but curves provide permanent speed limitations.

Some of those 1830s-1840s routes are *absurdly* curvy, following every bend in the river or the mountains to stay as close to flat as possible. (And to save money, of course.) Nobody would build it that way now; they'd put in a cut or a tunnel or a bridge or something, and would accept a small grade to eliminate extreme curves. Horseshoe Curve is an example of what designers avoid now, but not the most extreme example. By the 1890s,, railroad construction was already prioritizing straighter over flatter and a number of lines had numerous bypass tunnels built for exactly this reason; lots were done in the 1920s-1930s.
 
I've crayoned-out a route through State College -- basically requires a tunnel under State College proper (5-10 miles, probably cut and cover) and tunnels from there to Lewistown (5-10 miles to pass under Penn Roosevelt State Park), while the rest of it is easy. It would take much more work to straighten out the rest of the line (West of Altoona and the curves between Lewistown and Harrisburg) but I think the value of stopping at State College alone would be quite substantial. PA state government has not, so far, agreed. (*sigh*)
Wonder what that would all cost? Might be cheaper to "move" Penn State to Altoona....😊
 
I've crayoned-out a route through State College -- basically requires a tunnel under State College proper (5-10 miles, probably cut and cover) and tunnels from there to Lewistown (5-10 miles to pass under Penn Roosevelt State Park), while the rest of it is easy. It would take much more work to straighten out the rest of the line (West of Altoona and the curves between Lewistown and Harrisburg) but I think the value of stopping at State College alone would be quite substantial. PA state government has not, so far, agreed. (*sigh*)
I've lined up a similar route myself, but you also need to deal with Bald Eagle Mountain on the northwest side of State College, unless the new line is run down the Nittany Valley to the water gap that leads into Tyrone. It might be better to run the line directly from State College to Port Mathilda, because there is an existing rail line between Port Mathilda and Tyrone that could be upgraded. Also, the terrain in the Nittany valley is pretty rolling, so any rail line placed through it would require a lot of cuts and fills. So it might be better to have another tunnel, under Bald Eagle Mountain, and a ~400 ft. elevation difference between State College and Port Mathilda.
 
Last edited:
Wonder what that would all cost? Might be cheaper to "move" Penn State to Altoona....😊
Well, Elon Musk's Boring Company will provide the revolutionary technology needed to make this project much cheaper! :)

The costs could well be paid back by the generalized economic benefit of rail service to the entire region -- and we're talking about the possibility of a multiple-train-per-day Keystone West service between Harrisburg and Pittsburgh, not the current one train per day. Of course, they would probably need to run one or two trains per day over the old line, so Huntingdon doesn't get totally shut out -- Huntingdon doesn't even have bus service, in fact, they don't even have taxicab service, which surprised me, as it's not that small a town.
 
I've lined up a similar route myself, but you also need to deal with Bald Eagle Mountain on the northwest side of State College, unless the new line is run down the Nittany Valley to the water gap that leads into Tyrone. It might be better to run the line directly from State College to Port Mathilda, because there is an existing rail line between Port Mathilda and Tyrone that could be upgraded. Also, the terrain in the Nittany valley is pretty rolling, so any rail line placed through it would require a lot of cuts and fills. So it might be better to have another tunnel, under Bald Eagle Mountain, and a ~400 ft. elevation difference between State College and Port Mathilda.
FYI
https://www.statecollege.com/news/l...n-use-after-years-of-sitting-dormant,1480507/
 
I've lined up a similar route myself, but you also need to deal with Bald Eagle Mountain on the northwest side of State College, unless the new line is run down the Nittany Valley to the water gap that leads into Tyrone. It might be better to run the line directly from State College to Port Mathilda,
Yes, that was my route. :) From west to east, Port Mathilda to State College, more or less following state route "Business 322" under State College. Then punch a base tunnel from Boalsburg (or maybe Tusseyville) to Woodland. One more tunnel and a few bridges gets you to Burnham, where the existing rail line from the east ends. Then you can run through from Harrisburg to Tyrone *via* State College. Huntingdon (and only Huntingdon) loses service.
 
Courts? No, Congress would have to do it. The courts would happily enforce it, but it would have to be passed by Congress first.

No, I mean the courts. Even if congress assessed such a penalty (doubtful), it would definitely face many challenges as to the constitutionality of such a penalty.
 
Great comment. In a recent discussion, a point was made of how much bang for the buck you could get by making 30 mph track 50 mph, and 50 mph 79. Much easier than being obsessed about 110. We can do this. Reasonable speed, reliable, multiple frequency trains are the answer.
The US could follow India's example and shoot primarily for 100moph on all trunk routes, with 125 on select segments. This is entirely separate from the HSR side of things which is handled by an entirely different organization, with entirely different sources of funding so as not to interfere with the more widespread service provision goal for the trunk routes. But that sort of thing is unlikely to come to pass. We are too used to haphazard wasteful course of action involving trying everything that does not work before chancing onto something that does.
 
No, I mean the courts. Even if congress assessed such a penalty (doubtful), it would definitely face many challenges as to the constitutionality of such a penalty.
No, it wouldn't. Ever looked into Constitutional law? Crystal clear it's constitutional. Not debatable. Textbook regulation of interstate commerce.

I agree that it would be a very heavy lift to get Congress to pass such a law.
 
No, it wouldn't. Ever looked into Constitutional law? Crystal clear it's constitutional. Not debatable. Textbook regulation of interstate commerce.
Just because a federal law passes the interstate commerce constitutional hurdle doesn't mean it wouldn't pass other hurdles.
 
I've crayoned-out a route through State College -- basically requires a tunnel under State College proper (5-10 miles, probably cut and cover) and tunnels from there to Lewistown (5-10 miles to pass under Penn Roosevelt State Park), while the rest of it is easy.
A new HSR line shouldn't go straight through an established downtown of a city. In this case it would probably be most sensible to put the State College station somewhere south of downtown (PA26 / PA45 / Whitehall Rd. areas?) and encourage smart development around it. The university's campus transit could link up to it there.
 
Recently, a rail company in China announced that it is now offering high-speed rail cars with bogies that can adjust to train tracks of different gauges. If these trains work as advertised, then rail travel between Asia & Europe via rail is more convenient as changing bogies between border crossings takes up to 3 hours.

 
Recently, a rail company in China announced that it is now offering high-speed rail cars with bogies that can adjust to train tracks of different gauges. If these trains work as advertised, then rail travel between Asia & Europe via rail is more convenient as changing bogies between border crossings takes up to 3 hours.


Sounds like it would be a good plot point for an action-adventure movie involving a chase of a high speed train, a border crossing with a gauge change, and a bogie changer malfunction. (I guess you'd need a brake failure, too, so the train can't slow down for the customs check.) I mean, it's kind of inpractical to have fight scenes on the tops of moving railway cars, what with the trains going 300 km/hr and 25 kV catenary in the way, so I guess movie producers do need to find and alternative narrative. :)
 
The number that I read for installing new track is $1 million per mile so if you could even use the long abandoned Milwaukee road ROW West, and lay track, a dedicated Chicago to Seattle route would cost over $1 billion. Then add in the signaling and station costs. I do not see Amtrak ever owning more of its own track. That is also a reason why I say the big "corridor route" plan that Amtrak management is pushing is a pipe dream. The freight railroads, except perhaps for the small branch RR lines, do not want more passenger trains on their tracks.
 
There is a difference between opening new corridors and actually building a new right of way. A mile of track may cost about $1 million to build, but that's doesn't include the cost of buying the land. The biggest barrier to Amtrak getting it's on ROW is the cost of land in areas where train service is needed. Rebuilding an existing double track line to passenger standards would run between $2.5-$2.7 million per mile. Throw in a few sidings and the Class 1s would consider allowing more corridors trains in their tracks. Cause track improvements they don't have to pay for or very little for means a lot. Getting your average track speed up from 38 miles per hour to 55 or greater even over a 200 mile stretch is a game changer. And the cost would be what $600 million from the public not including stations? It's a win win for passenger trains and the railroads.
 
Correct this radical assumption:

Amtrak has Priority (right of way) if operating ON TIME on Freight tracks ?!?
If operating late the Priority is void -

Now however if the lateness is due to the Freight traffic causing the lateness then
what kind of redress is made available ?

Passenger service railroads ALWAY had Priority right of way except during war time -
National Security movements of critical materials.

Of course passenger rail service is not what it used to be and Amtrak building it's own
trackage is not going to come into play until the country is wall to wall residences similar
to the NEC.

I think I would prefer throwing blindfolded darts at a revolving piñata dart board to resolve this !
 
Track cost per mile is not considering the costs of bridges and tunnels. That really ups the costs. Just look at the CA HSR project. That project is having to build the many bridges and does not seem to be getting any road funds to prevent grade crossings.
 
Back
Top