cross country high speed corridors

[Joel N. Weber II]

I would very much like to be able to get from Boston to California by sleeping car in about 24 hours or less. TGV technology ought to make something like this possible.

However, it's unlikely that a high speed track all the way across the country is going to be funded purely to run sleepers. On the other hand, most travelers who have a bit of experience with trains seem to find that if their trip happens to be doable in about 3 hours or less by train, the train is a better choice than the plane. So I've been looking at the question of whether you can build a lot of these track segments which individually will make sense to the average airplane traveler, which will happen to connect to each other for the convenience of the long distance train traveler.

If we as a country were to collectively decide that where a major city is within about three hours or less by high speed train of another major city, that building TGV quality track makes sense, I believe it is possible to come up with a route that goes all the way across the country by high speed train. I'm assuming that the train will average 170 miles per hour from station to station, and that the major cities that can be considered are the top 30 US Combined Statistical Areas.

Where I've used track miles, they come from the Amtrak timetable; a high speed route might end up being a little shorter. Where I've used highway miles, the numbers are a bit rough, in that I've gone with whatever Google Maps thinks is the city center to be used when an address within the city is not specified, and a track route might not follow a highway route at all. Still, a rough estimate is better than no data.

The times between Denver and adjacent cities are a bit on the long side; then again, the Denver airport has managed to make itself huge to help keep itself competitively slow with the extra travel time those trains encounter relative to the rest of the train segments on this list.

One possible routing:

Philadelphia is the 8th largest Combined Statistical Area.

Philadelphia to Pittsburgh is 353 track miles on the current route; that trip would be roughly 2 hours at 170 MPH.

Pittsburgh is the 18th largest Combined Statistical Area.

Pittsburgh to Cleveland is currently 140 track miles, which would be under an hour at 170 MPH.

Cleveland is the 15th largest Combined Stastical Area.

Cleveland to Chicago is currently 341 track miles, which is right around 2 hours at 170 MPH.

Chicago is the 3rd largest Combined Statistical Area.

Chicago to St Louis is 297 highway miles, a bit under 2 hours at 170 MPH.

St Louis is the 16th largest Combined Statistical Area.

St Louis to Kansas City is roughly 250 highway miles, roughly 1.5 hours at 170 MPH.

Kansas City is the 22nd Largest Combined Statstical Area.

Kansas City to Denver is 603 highway miles, a little over 3.5 hours at 170 MPH.

Denver is the 14th largest Combined Stastictal Area.

Denver to Salt Lake City is 534 highway miles, but that's a pretty indirect route. However, 534 miles at 170 MPH is a bit over 3 hours.

Salt Lake City is the 27th largest Combined Statistical Area.

Salt Lake City to Las Vegas is 420 highway miles, about 2.5 hours at 170 MPH.

Las Vegas is the 25th largest Combined Statistical Area.

Las Vegas to Los Angeles is 271 highway miles, a bit over 1.5 hours at 170 MPH.

And Los Angeles is the 2nd largest Combined Statistical Area.

There are some other possible variations on this theme. I was looking at Pittsburgh to Philadelphia because that's where the tracks go now, but Pittsburgh to NYP and Pittsburgh to WAS tracks might also be viable options for routes that would take less than three hours at 170 MPH. I think Pittsburg to Columbus to Indianapolis to St Louis is a viable option for meeting all of the basic criteria, but if not all relatively nearby large cities get high speed track connecting them, focusing on the very largest cities is probably best, even if that makes the cross country route a bit less direct.

 

[wayman]

Salt Lake City to Las Vegas is 420 highway miles, about 2.5 hours at 170 MPH.

They may look close on a map, but they're a world apart, culturally! :lol:

 

[George Harris]

We had the beginnings in the early 1950's but without funding from Uncle Sugar, there was no way for the railroads to keep it going. At that time long sections of the ATSF between Chicago and LA had 100 mph speed limts and long sections of the UP across Nebraska and Wyoming and a lot of Ogden to LA allowed at least 90 mph. Other than the ICRR and ACL most lines in the east thought 80 mph was fast enough.

 

[the_traveler]

and long sections of the UP across Nebraska and Wyoming and a lot of Ogden to LA allowed at least 90 mph

And now UP can't (or at least doesn't want to) run Amtrak timely!

 

[Walt]

I'm assuming that the train will average 170 miles per hour from station to station, ...

Gosh, how would one even protect the tracks, for a train traveling at that speed?

The ol', stop at a crossing, look both ways, would no longer work well since at that speed, a train could easily go from "no where in sight" to "smash, crash, derailed", before the double tractor/trailer could safely cross.

Would such a train need some real, serious, braking power too?

I just can't imagine the existing style of infrastructure being able to handle speeds like that. Would one need enclosed, tunnel like, tracks?

 

[AlanB]

I'm assuming that the train will average 170 miles per hour from station to station, ...

Gosh, how would one even protect the tracks, for a train traveling at that speed?

The ol', stop at a crossing, look both ways, would no longer work well since at that speed, a train could easily go from "no where in sight" to "smash, crash, derailed", before the double tractor/trailer could safely cross.

Would such a train need some real, serious, braking power too?

I just can't imagine the existing style of infrastructure being able to handle speeds like that. Would one need enclosed, tunnel like, tracks?

According to FRA regs, you can't run trains that fast on tracks that have grade crossings. Such a line would need to have zero grade crossings or the train would have to slow down for any grade crossings, so your scenario of a double tractor trailer, much less a car getting hit, could never come to pass.

On the other hand there would still be a worry for animals and people just wandering across the tracks, and that would be very hard to combat.

 

[jackal]

I assume such a HSR as Joel is proposing would be built from scratch, not using any current tracks (at the very most, possibly using portions of some current ROWs if they were straight enough and the current railroads were willing to give them up, though even then the tracks themselves would need to be rebuilt).

The new tracks and ROW would, as it is in European high-speed railway, be completely grade-separated as well as fenced off to discourage trespassing.

It would be an absolutely massive undertaking (especially to build the ROW to 170mph standards in built-up or geographically-challenging areas), but if a majority of the air traffic between these cities were diverted to rail, it might potentially be fairly cost-effective.

 

[puck71]

Something like this would be awesome and could take away (or at least lessen) one of the biggest downsides of long-distance train travel in the US - time. The best way to make it happen I think would be to start on one of the routes (probably LA to LV, but maybe east coast) and see how it does. If it does as well as I think it probably would, then expand from there.

 

[Joel N. Weber II]

I assume such a HSR as Joel is proposing would be built from scratch, not using any current tracks (at the very most, possibly using portions of some current ROWs if they were straight enough and the current railroads were willing to give them up, though even then the tracks themselves would need to be rebuilt).
The new tracks and ROW would, as it is in European high-speed railway, be completely grade-separated as well as fenced off to discourage trespassing.

It would be an absolutely massive undertaking (especially to build the ROW to 170mph standards in built-up or geographically-challenging areas), but if a majority of the air traffic between these cities were diverted to rail, it might potentially be fairly cost-effective.

The older French TGV tracks that are only good to 200 MPH have a minimum curve radius of four miles. The newer stuff they're building has an even larger curve radius in preparation for the faster speeds they hope to run someday.

The parts of America that I've been looking at in Google Maps don't tend to have even a 10 mile stretch that is free of curves that would fail to meet the four mile radius standard. (I'd love to see someone come along and tell me that I'm wrong and I haven't been looking in the right places, and that actually, there's a nice existing 100 mile long stretch that's already straight enough for 200 MPH operation the whole way without all the construction costs associated with establishing a new right of way, even if a track had to be bulit from scratch on that right of way.)

Part of this is because historically, US railroads were generally focused on keeping the grade down to about 1% or 1.5% as much as possible, and they didn't want to tunnel more than absolutely necessary, and if they had to have lots of curves to make that work, that was OK. Unfortunately, that doesn't result in a route that works for high speed passenger trains.

So while maybe there's an existing right of way that's out there that's straight enough, I think basically you'd probably need to build track from scratch on a new right of way if you want to do 150-200 MPH.

When I talk about an average speed of 170 MPH, I'm actually thinking of 200 MPH track through areas that's barely populated, and the existing right of way (where I'm guessing 45 MPH might be a reasonable expectation) through densely populated areas. I think I have seen something that says that the French have actually achieved 176 MPH station to station averages, using existing rights of way in densely populated areas, and new 200 MPH track in less populated areas.

I'm starting to think that when going through nearly flat farmland, elevating the tracks about 15 feet above the surrounding land would be really nice. It would improve the view for the passengers, it would get the tracks at about the right elevation for crossing roadways on bridges, and it might provide some protection against flooding if the land suporting the railroad is stable enough when the surrounding land is flooded.

 

[Joel N. Weber II]

On the other hand there would still be a worry for animals and people just wandering across the tracks, and that would be very hard to combat.

If you can get people out of their automobiles and on to trains, there's a good chance that the reduction in fatal automobile accidents will more than make up for the deaths along the tracks.

People who die in car accidents also often aren't doing something that's easy for a careful person who's risk averse to avoid, whereas an intelligent person who doesn't want to be killed by a train running on a fenced, grade separated track has a good chance of being able to choose to avoid trespassing.

 

[Joel N. Weber II]

Something like this would be awesome and could take away (or at least lessen) one of the biggest downsides of long-distance train travel in the US - time. The best way to make it happen I think would be to start on one of the routes (probably LA to LV, but maybe east coast) and see how it does. If it does as well as I think it probably would, then expand from there.

I'm starting to think that the inital route with the most potential is probably one that starts just west of the greater New York City area (possibly taking NJT's route as far as High Bridge to get through the densely populated area, with the 200 MPH track starting near High Bridge and heading west from there, but I'm not sure if that's really the best way to get out of New York City). It would then head in a straight line to somewhere a bit to the north of Pittsburgh, with tracks so that the trains that aren't super-express could slow down and head south a bit and stop at the existing Pittsburgh station, and the super-express train wouldn't have to go any slower than 200 MPH while bypassing Pittsburgh. The track would then head northwest towards Cleveland, again with slower track to the existing station, and a bypass that's 200 MPH the whole way around Cleveland, and then the 200 MPH track would head west towards Chicago, and eventually connect with the existing tracks feeding into Chicago once it reaches a heavily populated area.

Then, when there's money to expand on that, additional tracks could be built connecting to Harrisburg (possibly with the idea that a train heading south from this new track to Harrisburg would either continue to Philadelphia or to Pittsburg via the existing route with all the local stops rather than finding its way back to the 200 MPH track after stopping in Harrisburg), and an additional track could be built just west of Pittsburg that would continue southeast to Washington, DC at 200 MPH.

The New York City Combined Statistical Area has a population of almost 22 million; the Chicago CSA almost 10 million; the Cleveland CSA almost 3 million; and the Pittsburg CSA almost 2.5 million. The Las Vegas CSA is smaller than any of these, at almost 2 million; the Los Angeles CSA has almost 18 million. If you assume that both Cleveland and Pittsburg should be close enough to New York City that planes would become obsolete for this route if such a high speed track is built, and if you assume that number of trips is going to be proportional to population, then New York City to Pittsburgh ought to have more riders. Plus, people who go to New York City tend to be comfortable with trains. The downside is that New York City to Pittsburgh is a longer route than Los Angeles to Las Vegas and thus will probably be a more expensive experiment.

It would also be interesting to try to get a good estimate of what fraction of the New York to Chicago market Amtrak would capture if Amtrak could do the trip in five hours. And of course, maybe it will be possible for Amtrak to exceed 200 MPH on a good portion of that route someday, which is certainly something that should be planned for when it's built. Are there numbers on Amtrak's share of the Boston to Washington DC market, which would be a pretty good indicator of how willing people are to take 5-7 hour train trips (as NYP to CHI would end up being with this 200 MPH track) where a faster plane is available?

 

[jackal]

According to Wikipedia, "by some reckoning, Amtrak has captured over half of the market share of travelers between Washington and New York." (The source is an article in Trains magazine, June 2006, if anyone wants to verify that claim.)

Since we're still in the experimental stages and there is no way this will be built without a lot more proof that HSR works, I wonder if it would be more cost-effective to revamp the NEC to allow these 200mph speeds. Yes, it will require a lot of land taken under eminent domain and will upset a lot of people, but it would probably still be cheaper and would probably have a better ROI than tunneling through the Appalachians. If we could achieve a 170mph average between WAS and NYP, that would virtually guarantee 100% marketshare to Amtrak between those cities. (The current average is somewhere around 75mph, IIRC.)

Once Amtrak has complete dominance there, then we can start looking at expanding westward...

 

[AlanB]

Once Amtrak has complete dominance there, then we can start looking at expanding westward...

Actually I'd look to expand south first, where they already have and interest and plans for higher speed trains.

 

[wayman]

Once Amtrak has complete dominance there, then we can start looking at expanding westward...

Actually I'd look to expand south first, where they already have and interest and plans for higher speed trains.

Slowing down the definition of "high-speed" a bit, moving to the South where everything is a bit more laid-back...

What's the likelihood of Amtrak--once given the permission to initiate new routes--working with Norfolk Southern to add higher-speed (not "high-speed", but at least trains that can reliably clip along at 79 rather than poke along at 45) between Atlanta and Jacksonville, Memphis and Chattanooga and Atlanta, and other enormous corridor capacity and infrastructure projects in the south which NS is undertaking? Obviously, NS is doing this to increase their freight traffic capacity and efficiency, but if Amtrak could benefit from ATL-JAX with one reliable fast train a day that would be fantastic. NS isn't exactly upgrading the Crescent's route, but parts of the Crescent might benefit and certainly connecting routes to the Crescent would benefit Amtrak.

If those connecting routes run on brand new well-designed corridor track and show off how reliable train travel can be given the money for solid infrastructure, that could be a huge demonstration and bring more money in that direction....

 

[Joel N. Weber II]

According to Wikipedia, "by some reckoning, Amtrak has captured over half of the market share of travelers between Washington and New York." (The source is an article in Trains magazine, June 2006, if anyone wants to verify that claim.)
Since we're still in the experimental stages and there is no way this will be built without a lot more proof that HSR works, I wonder if it would be more cost-effective to revamp the NEC to allow these 200mph speeds. Yes, it will require a lot of land taken under eminent domain and will upset a lot of people, but it would probably still be cheaper and would probably have a better ROI than tunneling through the Appalachians. If we could achieve a 170mph average between WAS and NYP, that would virtually guarantee 100% marketshare to Amtrak between those cities. (The current average is somewhere around 75mph, IIRC.)

Once Amtrak has complete dominance there, then we can start looking at expanding westward...

There are a finite number of seats available on the Acela trainsets. I have to wonder if just buying two more business class cars for each existing Acela trainset and buying 10 or 20 more Acela trainsets would help Amtrak's market share more cost effectively. Or maybe even just some new electric locomotives so that Amtrak could add some trains with Amfleet Is that make only Acela stops. That would drive down Amtrak's ticket prices, which would make Amtrak attractive to more travelers. I know I've heard a coworker complaining about how expensive Acela has gotten lately, and there's a trip I'm contemplating later this year where I may choose the Regional in at least one direction due to Acela's high prices. That suggests to me that there are travelers who are more sensitive to price than any mode preference who will prefer the plane simply because Amtrak doesn't have cheap seats (because of stupidity in the equipment acquisition process, not because of underlying energy costs).

And I bet if you found some way to upgrade most of the track to 200 MPH but made sure that Amtrak's number of available seats was the same as now, you'd find that Amtrak's market share wouldn't grow much.

Annoying people who will scream NIMBY is far more politically dangerous than spending lots of money, as far as I can tell.

I also think the long term plan should be for the existing NEC to turn into largely a commuter route, and if we want to spend real money on achieving 200 MPH for a good part of the route, we should build a completely new route that will probably skip most city centers entirely. Such a route could be built under the Deleware River, the Chesapeake & Deleware Canal, and the Deleware River expensively but with minimal legal paperwork, or perhaps it could be built 10-30 miles to the east or west of the existing route, largely over land, to be completed 30-40 years from now whenever the political process and eminent domain process can be completed.

(The MBTA recently spent a decade building the roughly 20 mile Greenbush Line on an existing but not active right of way. If we want any of this to actually get built at any useful point in our lifetimes, we need to be looking to agressively prefer options that will avoid as many potential delays as possible.)

I also think that Acela as it exists proves that there's demand for high speed rail. You don't need 100% market share to know that a service works; a service that has even 20% market share is valuable to travelers.

If we want to have any chance of having a 24 hour or less Boston to California sleeper 20 years from now, we probably should be trying to get the New York City to Chicago and quite possibly also Los Angeles to Las Vegas segment under construction soon, so that they'll have a chance of being done 10 years from now, and hopefully once Americans have an opportunity to experience those routes, it will become obvious that we should fill in the Las Vegas to Chicago segments.

 

[George Harris]

The older French TGV tracks that are only good to 200 MPH have a minimum curve radius of four miles. The newer stuff they're building has an even larger curve radius in preparation for the faster speeds they hope to run someday.. . . .

I'm starting to think that when going through nearly flat farmland, elevating the tracks about 15 feet above the surrounding land would be really nice. It would improve the view for the passengers, it would get the tracks at about the right elevation for crossing roadways on bridges, and it might provide some protection against flooding if the land suporting the railroad is stable enough when the surrounding land is flooded.

Joel: The basic French design was for 300 km/h = 186 mph

The original standards for Taiwan High Speed were written by them based on

"Initial operation at 300 km/h, but not precluding ultimately running at 350 km/h"

350 km/h = 217 mph

I think they kept the curve radius unchanged from what they used for 300 km/h in France.

For that system, the minimum curve radius was 6,250 meters = 20,500 feet (real close) = 3.88 miles

It also allowed an "exceptional" radius of 5,500 meters = 18,000 feet (real close) = 3.42 miles.

If we think faster, then say at 250 mph, you would want to have a curve radius of 36,000 feet = 6.82 miles or larger.

For your second point, for almost the whole country the "100 year" flood elevation can be found, and if you get a few feet above that you will be above the even more extreme floods. Usually, 15 feet above the prevailing ground is way more than necessary for that.

However: In Taiwan there was also the political promise that the line would be on structure through the southern farming area. As a result, a continuous viaduct nearly 100 miles long was built. In general, the deck was about 25 to 30 feet above the ground.

Minimum clearance over roadways in this country is defined as 14'-6", and for major roads usually is 16'-0" to 16'-3" Add an allowance of about 7 to 10 feet for bottom of structure to top of rail, and you get the track elevation at about 25 feet above the ground. Then, if the roadways are built clear of flood elevation, in the areas subject to flooding you will be higher than that.

As to speeding up the northeast corridor: It is way past time to start spending the money somewhere else. When we get a few other areas to an equivalent level of service, then we can start thinking about doing more there. I am thinking about such things as California high speed, or even Chicago - St. Louis, Miami - Tampa Bay, Seattle - Portland, etc.

 

[Joel N. Weber II]

The third paragraph of the intro to the TGV article in Wikipedia states

TGV trainsets travel at up to 320 km/h (200 mph) in commercial use.

And a couple sentences after that, Wikipedia links to this article which states:

France's electrically-powered fast trains have been operating since 1981, daily reaching speeds of 320 kph over some 1,600 kilometres (1,000 miles) of track.

Mostly I'm posting this because I want to demonstrate that I'd made some halfway reasonable attempt at gathering accurate facts. (I hadn't bothered to check wikipedia's citation prior to this post, but the 320 km/h claim does seem to be accurate. Google's calculator seems to think that 320 km/hr is more like 198.8 MPH, but I don't think that 1.2 MPH discrepeancy has a meaningful effect on the passenger's experience.)

I'm not sure getting lost in splitting hairs over 300 vs 320 vs 350 km/h has much point, however. The basic problems involved in achieving any of these speeds are similar in needing to find a right of way that is very straight. Whether your minimum radius is 3.5 miles, 4 miles, 7 miles, or something even larger than that just in case faster speeds become possible later, you're stuck with the same basic problem that existing rights of way in the US have curves that are too tight. Any of these speeds would require trainsets faster than the existing Acela trainsets, and meeting FRA safety standards on such fast trainsets is something that has never been done, though I expect it's an entirely solveable problem.

The real problem to be solved is figuring out how to get the American public to decide that building these new high speed routes is worth the money.

 

[Joel N. Weber II]

Here's another Telegraph article that lists slightly different speeds than the other Telegraph article:

Currently, average travelling speeds for the TGV are around 186 mph, but trains on the new Paris-Strasbourg line are to run at 198.7 mph.

But those numbers look an awful lot like the numbers cited outside of that article as maximum speeds, which makes me think the reporter may have been confused when using the word average.

 

[Green Maned Lion]

Joel, I have a considerable respect for you as an intelligent and forward-thinking individual, but allow me to pose the following question:

Our government can't keep our basic infrastructure (ex. the northeast corridor or even the interstate highway system- witness the MSP bridge collapse) even in a "state of good repair". Given this, how do you expect them to design, engineer, procure land for, build, and especially properly maintain something to FRA 200mph standards?

 

[George Harris]

The third paragraph of the intro to the TGV article in Wikipedia states

TGV trainsets travel at up to 320 km/h (200 mph) in commercial use.

And a couple sentences after that, Wikipedia links to this article which states:

France's electrically-powered fast trains have been operating since 1981, daily reaching speeds of 320 kph over some 1,600 kilometres (1,000 miles) of track.

Mostly I'm posting this because I want to demonstrate that I'd made some halfway reasonable attempt at gathering accurate facts. (I hadn't bothered to check wikipedia's citation prior to this post, but the 320 km/h claim does seem to be accurate. Google's calculator seems to think that 320 km/hr is more like 198.8 MPH, but I don't think that 1.2 MPH discrepeancy has a meaningful effect on the passenger's experience.)

I'm not sure getting lost in splitting hairs over 300 vs 320 vs 350 km/h has much point, however. The basic problems involved in achieving any of these speeds are similar in needing to find a right of way that is very straight. Whether your minimum radius is 3.5 miles, 4 miles, 7 miles, or something even larger than that just in case faster speeds become possible later, you're stuck with the same basic problem that existing rights of way in the US have curves that are too tight. Any of these speeds would require trainsets faster than the existing Acela trainsets, and meeting FRA safety standards on such fast trainsets is something that has never been done, though I expect it's an entirely solveable problem.

The real problem to be solved is figuring out how to get the American public to decide that building these new high speed routes is worth the money.

I was not trying to either nit-pick or split hairs. My source happened to have been a couple of civil engineer type Frenchmen that worked on the system, but not on their newest line.

 

[AlanB]

There are a finite number of seats available on the Acela trainsets. I have to wonder if just buying two more business class cars for each existing Acela trainset and buying 10 or 20 more Acela trainsets would help Amtrak's market share more cost effectively. Or maybe even just some new electric locomotives so that Amtrak could add some trains with Amfleet Is that make only Acela stops. That would drive down Amtrak's ticket prices, which would make Amtrak attractive to more travelers. I know I've heard a coworker complaining about how expensive Acela has gotten lately, and there's a trip I'm contemplating later this year where I may choose the Regional in at least one direction due to Acela's high prices. That suggests to me that there are travelers who are more sensitive to price than any mode preference who will prefer the plane simply because Amtrak doesn't have cheap seats (because of stupidity in the equipment acquisition process, not because of underlying energy costs).

Amtrak's prices for Acela aren't due to a lack of seats or the fixed nature of the Acela. The prices are driven by the fact that people are paying for them, and Congress and the White House are breathing down Amtrak's neck to turn a profit.

Buying two more cars per train set probably would help a bit to improve Amtrak's marketshare, would certainly improve Amtrak's bottom line operationally as well as driving up ridership numbers. But I strongly suspect that it would have little impact on the prices being charged. There might be more low bucket seat available than there are now, but I doubt that you'd see much drop in the high bucket prices.

 

[Joel N. Weber II]

Our government can't keep our basic infrastructure (ex. the northeast corridor or even the interstate highway system- witness the MSP bridge collapse) even in a "state of good repair". Given this, how do you expect them to design, engineer, procure land for, build, and especially properly maintain something to FRA 200mph standards?

The FRA's standards are pretty strict. The rules are pretty clear about how when the standards for a particular track class aren't met, you have to run the trains slower (though I think there are a few cases where you have a month or two in which to meet the standards before the trains have to fully slow down). If a lot of track that was supposed to be 200 MPH isn't, there's a good chance that the trains will slow down for a little while, the news media will notice, and the politicians will decide to throw money at the problem.

Sometimes building new things from scratch for $X is also politically easier than spending one twentieth of $X fixing something.

And I have to wonder if the sense of satisfaction engineers get varies. Would you rather spend a year of your working life building a new 200 MPH track, or doing a bunch of things along the lines of fixing some track that has a speed restriction of 90 MPH so that it can operate at 135 MPH like it's supposed to (and when you're done, still not being able to do anything about that catenary wire that won't let the trains run at 150 MPH)? (I assume the old catenary is not considered to be not in good repair merely on account of not being constant tension, but I don't really know.)

The US Navy's submarine force found that when the officers who had done best in school got to pick their assignments, the ships that were in overhaul ended up getting the weaker officers pretty consistently. The Navy eventually noticed that having weaker officers on the ships that were getting overhauled was a bad thing, and changed how they handled assignments.

If indeed the civilian engineers working on the railroads are going to prefer working on the 200 MPH track, there is no way in our democracy to force the best engineers to work on the stupid annoying little problems on the NEC. (And I strongly suspect that good engineers / management can do a lot to stretch budget dollars in a safe manner, where weaker engineers / management will take a less efficient approach and get less done with the same money.)

Notice also how the way Amtrak most recently managed to buy new rolling stock at all was that it was ``high speed''. Nevermind that the top speed south of New York City is a whole 8% faster for the Acela than the Regional, and an 8% speed difference probably barely has any effect on how many people are willing to ride the train. (It's more complicated than that, of course, since the Acela doesn't even get to enjoy that 8% speed advantage for the whole route, but the Acela also skips a bunch of stops, and can probably accelerate faster. But Amtrak could have simply bought some more electric locomotives and made a habit of running them in pairs and rewritten the schedule if they wanted the Amfleet I coaches to enjoy those last two advantages.)

I'm sort of curious about the NEC state of good repair issue. I do think I recall reading about it being an issue in the bill that Congress has almost managed to pass. But when I've ridden the NEC, I haven't noticed it being broken, and I suspect that the NEC in its present state is a safer way to travel than the current Interstate highway system, or any improved version of the highway system we might see in the next few decades.

I'm also wondering if the highway bridge standards are as clear cut as the FRA track standards. For some reason, I suspect they may not be, but I'm not really sure. I think I've seen news articles saying that engineers inspected some particular highway bridge and decided it was safe enough in some case in the distant past. I'm wondering how much subjectivity gets involved there; I'm reminded of what Richard Feynman discovered about various estimates of the safety of the Challenger. (Various lower level people at NASA apparently had a pretty good idea of the actual safety. Management figured they couldn't possibly be sending actual living human beings into outer space in something that dangerous, because if it was that dangerous you wouldn't want to send actual living human beings, so obviously it was safer than the people who were working for them thought it was. Unfortunately, we discovered that management writing wishful numbers down on a piece of paper didn't make reality so.)

While I do think we need to be working to fix the highway bridges (as well as thinking about whether we could save money in the long run in meeting our total transportation needs by getting some of the load off the highway and onto the rails), a single highway bridge collapse really doesn't bother me much. I'm too lazy to go find the real numbers at the moment, but if there were 150 people who died in that collapse and about 150 million Interstate highway bridge users in the US (counting both drivers and passengers), it would seem that collapse only killed one out of every 1 million Interstate highway users. There are probably a bunch of other ways of dying on an Interstate highway that are a lot more likely.

 

[Joel N. Weber II]

Amtrak's prices for Acela aren't due to a lack of seats or the fixed nature of the Acela. The prices are driven by the fact that people are paying for them, and Congress and the White House are breathing down Amtrak's neck to turn a profit.
Buying two more cars per train set probably would help a bit to improve Amtrak's marketshare, would certainly improve Amtrak's bottom line operationally as well as driving up ridership numbers. But I strongly suspect that it would have little impact on the prices being charged. There might be more low bucket seat available than there are now, but I doubt that you'd see much drop in the high bucket prices.

I believe more seats would indeed increase the number of low bucket seats; the prices are supposed to be set so that however many people are willing to pay the high bucket prices will pay the high bucket prices, and lower bucket prices are available to people who book early to make sure all the seats get sold. Additional seats aren't going to increase the number of people willing to pay the high bucket, so they'll have to be sold to lower bucket customers.

(The system obviously doesn't manage to maximize revenue perfectly, because the people who are willing to pay the highest prices occasionally buy their tickets too early to do so.)

But when I'm looking at potential BOS-WAS trips in September where booking three months out can vary from under $100 for coach on a Regional to over $200 in Acela Business Class, I have to wonder if more Acela seats could increase the likelyhood that I'd choose to take the Acela instead of the Regional.

 

[Joel N. Weber II]

One other thing to consider is that the Northeast Corridor has grade crossings and moveable bridges. In some ways, grade crossings are harder to maintain than grade separated crossings (on the other hand, grade separated crossings do have bridges that need occasional work, and I'm not sure which factor ultimately outweighs the other).

A new high speed track that's specifically intended for passenger trains may be able to avoid having any moveable bridges, as such a track tends to avoid some of the issues that traditional freight railroads ran into with wanting to avoid having a steep track grade. That certainly will improve the reliability of the system. (The surface platforms at North Station in Boston are reached by a moveable bridge across the Charles river, and there has been at least one occasion in the last couple years when mechanical problems with that bridge have delayed a bunch of rush hour MBTA trains. If the North South Rail Link gets built, and if all of the trains start running through it, that maintenance issue might go away.)

Also, a dedicated high speed passenger track probably has to be built away from populated areas in order to get reasonable curves without huge amounts of land acquisition hassles. That is probably going to also mean that it will have far fewer switches where it connects to other branches of the system. And I suspect you may be able to get away with spacing passing tracks based upon the time the slowest train takes to reach them. If you're content to risk having trains have to wait 5 minutes to meet other trains, at 200 MPH you can space the passing sidings about 16 miles apart. (Maybe you'd regret that on days when the trains have to run slowly because of some mechanical problem, though.)

An interesting question is whether you can build a switch that will let a train run through it at 200 MPH regardless of which of the two directions the train is going (which you'd want if you build the track mostly single tracked, with double track in the sections where you expect the schedule will put the trains where they meet). I bet you can, but it probably doesn't tend to be done. But keeping high speed switches maintained is apparently pricy, I think somewhere I read that a switch that you can go through at 45 MPH (in the slower direction, I think) can cost something like $10k or $30k a year to maintain.

George Harris has also mentioned that concrete slab construction will cut maintenance costs, though I don't know if 200 MPH concrete slab track maintenance is cheaper than maintaining 135 MPH track that merely has concrete ties.

 

[Chafford1]

The third paragraph of the intro to the TGV article in Wikipedia states

TGV trainsets travel at up to 320 km/h (200 mph) in commercial use.

And a couple sentences after that, Wikipedia links to this article which states:

France's electrically-powered fast trains have been operating since 1981, daily reaching speeds of 320 kph over some 1,600 kilometres (1,000 miles) of track.

The line speed for French 'Lignes a Grand Vitesse' is 300km/h (186mph), apart from the LGV Est line where the line speed is 320 km/h (199mph). The French are looking into raising the maximum speed to 360km/h (225mph).

The recently opened Madrid-Barcelona line in Spain currently runs at 300km/h (186mph), but line speeds will be raised to 350km/h (217mph) later in 2008 or 2009.