Is Passenger Rail Doomed?

amm in ny wrote:

amtrakowitz wrote: [long quote of amm's post removed]

By that logic, SJ's X2000 (Sweden) should not be running at all, since the average population density of that country is 54 people per square mile and the X2000 has towns of between 14,000 and 50,000 inhabitants as some of its endpoint destinations.

Average population density would only be meaningful if Sweden's population -- and its rail network -- were uniformly distributed over its entire area.

A look at Wikipedia shows that most of Sweden's population is in the southern ⅓ or ¼, and most of it is on an axis running from Stockholm to Göteborg and on down to the southern tip. As is most of the rail network. In this area, the density is over 200 people per square km (500/mi²).
The article didn't specifically mention X2000, but the 200 km/hr. segments are (not surprisingly) only in this area.

Similar considerations apply to the NEC -- the relevant density is the area within, say, 50 -75 miles of the line, which leaves out most of New York, Pennsylvania, Virginia, and much of Massachussetts and Maryland. And this area includes the most densely populated parts of those states.

BTW, my comment wasn't intended to prove anything. It was intended to rebut the "Europe can do it, why can't the USA?" argument. Maybe the USA can make rail passenger service play a larger role than it does not, but comparisons with Europe don't prove it.

Wikipedia is hardly authoritative. The density of the rail network in Sweden is quite high in the sparsely-populated areas, especially when it comes to passenger transportation. My point about that was that just like the USA, Sweden's average population density, although not uniformly distributed (it can't be), serves more unpopulated areas with passenger rail (and even high-speed rail) than the USA does. The X2000 terminating at Arvika shows that high-speed rail is considered important enough to go to a town with 14,000 inhabitants on average, and this in a sparsely-populated province (Värmland, average population density 44 people per square mile, largest city Karlstad with a municipal population of around 85,000). Arvika is in a region with many lakes, and by road is only accessible by two-lane highway. As much could be noted for other X2000 destination cities: Uddevalla (pop. 31,212, not connected to Stockholm by motorway although there are local motorways); Falun (pop. 37,291, no motorways); Sundsvall (pop. 50,712, although with total of 145,000 in metro area; X2000's northernmost terminus); one summer destination is Strömstad (pop. 6,288).

amtrakowitz wrote:Also, that logic dictates that interstate highways should be restricted to the eastern and western seaboard and ought not crisscross the entire contiguous USA—cities ought to be connected with the lowest-possible-density roads between cities (preferably two-lane highways, with minimal grade and curvature easement) when these roads traverse low-density areas.

Hardly.

The ways roads are operated and used are so different from passenger railways that you really need to do the analysis from scratch.

That said, roads in truly remote areas (e.g., Northern Alaska) do tend to be two-lane highways with simple intersections, and speed limited mostly by the drivers' sense of what is safe.

WADR, that's evasive. Many rural areas in the USA would not be pressured by having no interstate highway passing through them. Most likely, there would be a greater benefit.

Sweden's X2000 destinations are not all connected by motorways, certainly not to Stockholm; however, Acela Express' destinations all have several interstates connecting them aside from the primary route (I-95). Further notable that the X2000's average speed on just about all of its routes is higher than the Acela Express, in spite of its top speed being held down to 127 mph. The only train that Amtrak operated whose overall average speed approached that of most X2000 trains is the nonstop NYP-WAS Metroliner (which achieved a 90-mph average speed).

amm in ny wrote:That said, roads in truly remote areas (e.g., Northern Alaska) do tend to be two-lane highways with simple intersections, and speed limited mostly by the drivers' sense of what is safe.

The only long road to northern Alaska I'm aware of is a one lane gravel road along the pipeline, like the dirt paths along rural railroads in the lower 48 states, used by maintenance vehicles maintaining the pipeline.
If you really need to get to northern Alaska, you will have to fly or go cross country.

electricron wrote:

amm in ny wrote:That said, roads in truly remote areas (e.g., Northern Alaska) do tend to be two-lane highways with simple intersections, and speed limited mostly by the drivers' sense of what is safe.

The only long road to northern Alaska I'm aware of is a one lane gravel road along the pipeline, like the dirt paths along rural railroads in the lower 48 states, used by maintenance vehicles maintaining the pipeline.
If you really need to get to northern Alaska, you will have to fly or walk.

This is consistent with the TV program that I learned about this road from.

I wasn't aware that it was restricted to pipeline maintenance vehicles, I got the impression that it was a public road. Of course, I imagine that not many people who aren't working for the pipeline -- or providing services for them -- would have any interest in driving up that way, anyway.

amm in ny wrote:I wasn't aware that it was restricted to pipeline maintenance vehicles, I got the impression that it was a public road. Of course, I imagine that not many people who aren't working for the pipeline -- or providing services for them -- would have any interest in driving up that way, anyway.

If it were a public road, it would have an Alaskan Highway number. Does it?
Is there a gas station on it for the public to use to refill their vehicles?

This Alaskan road, whether it's public or private, is an excellent example on where transportation links are built. First of all, it must have a purpose. No one has ever spent time or money building roads to nowhere. It may not even be built to move people at all, it could have been built just to move goods.

Rural interstate highways through the vast middle of America, where density levels reach single digits per square mile, weren't built for those living there in the middle, but for those living on either end of the super highways. The intercontinental railroads weren't built for those living in the middle, the indians didn't even want the iron horse, they were built to link the west coast to the east coast. These railroads were not great real estate developers, but they were able to eventually sell their land grants off to settlers.

It seems most new highway and rail projects today require land speculators and developers to gather enough political support for financing, not surprisingly much like over 150 years ago.....

Arguing that LA-SF is only a small percentage of passengers coming from LAX ignores the forest for one particular tree.

Imagine if someone in France said, in the 1970s "Oh, Paris-Lyon is only a few percent of the traffic coming from CDG - why bother with this newfangled TGV-thingy?"

The journey of a thousand miles starts with a single step.

djlong wrote:Arguing that LA-SF is only a small percentage of passengers coming from LAX ignores the forest for one particular tree.

Imagine if someone in France said, in the 1970s "Oh, Paris-Lyon is only a few percent of the traffic coming from CDG - why bother with this newfangled TGV-thingy?"

True, but Lyon is the second largest city in France, while Paris is the largest, and are only around 250 miles apart. You're missing my main point! The two largest cities in America are closer to 3,000 miles apart. Travel distances within America can be two, four, eight and up to ten time times further than within any nation in Europe. At the same or up to twice these distances, trains can compete. But beyond that, planes are far more attractive to travelers, whether on business or as tourists. Name one HSR train in the world outside of China that is even competitive with planes at distances beyond 500 miles. Look very hard, you won't find any.

During fiscal year 2010 long distance Amtrak trains carried 4.5 million passengers 2.8 billion passenger miles. That means the average passenger rode the slow long distance Amtrak trains an average around 622 miles. For comparison, long distance trains in Germany carry around 120 million passengers each year with an average train trip distance being around 179 miles. Note, these are average distances taken by riders of trains, not the average distances taken by auto drivers or plane flyers. Even with very low market share percentage, your average train rider in America rides the train almost 4 times further than your average Germany train rider. Why? Because they really need to ride 4 times further to get where they want to go......
Sources of data above:
http://www.google.com/url?sa=t&rct=j&q= ... GhKRhijPJQ
http://en.wikipedia.org/wiki/Rail_transport_in_Germany

Here's my point again, HSR trains should be competitive with all other modes of travel at 179 miles average distance in Europe, but not at the average 622 miles in America.

electricron wrote:Here's my point again, HSR trains should be competitive with all other modes of travel at 179 miles average distance in Europe, but not at the average 622 miles in America.

I don't see what you're getting at. Has anyone reasonably suggested high-speed rail from New York to Los Angeles? For that matter, anything more than 110 mph near-high-speed rail seems unlikely between Las Vegas/Phoenix and Dallas.

I have a feeling we're all arguing more-or-less the same point, that high-speed rail is competitive as a short/medium-distance form of transportation or as a concatenated corridor (New York-Albany-Buffalo-Cleveland-Toledo-Chicago comes to mind), but not as a true long-distance form of transportation competitive with airlines (New York-Los Angeles nonstop or one-stop.) I don't ever expect to be doing 150 mph through the Rockies, but 110 mph would be a good start between Seattle, Portland and Spokane, San Francisco and Reno, Los Angeles and Phoenix, and so on. Those journeys fit the European HSR model relatively well.

mtuandrew wrote:I have a feeling we're all arguing more-or-less the same point, that high-speed rail is competitive as a short/medium-distance form of transportation or as a concatenated corridor (New York-Albany-Buffalo-Cleveland-Toledo-Chicago comes to mind), but not as a true long-distance form of transportation competitive with airlines (New York-Los Angeles nonstop or one-stop.) I don't ever expect to be doing 150 mph through the Rockies, but 110 mph would be a good start between Seattle, Portland and Spokane, San Francisco and Reno, Los Angeles and Phoenix, and so on. Those journeys fit the European HSR model relatively well.

Get away from the coasts and look at the distances between American cities.
For example, 9th largest MSA being Atlanta, just to cover the Southeast quarter of the country.
Atlanta to Miami (8th largest MSA) = 665.1 miles
Atlanta to Washington D.C. (7th largest MSA) = 636.42 miles
Atlanta to Houston (6th largest MSA) = 791.26 miles
Atlanta to Philadelphia (5th largest MSA) = 774.47 miles
Atlanta to Dallas (4th largest MSA) = 780.5 miles
Atlanta to Chicago (3rd largest MSA) = 717.13 miles
Atlanta to Los Angeles (2nd largest MSA) = 2172.71 miles
Atlanta to New York City (largest MSA) = 879.51 miles
None of the higher populated metros (with a great chance most businessmen might wish to go) is closer than 600 miles from Atlanta. So, even if high speed rail is built to Atlanta, i.e. to the South, planes will still be the preferred choice for those traveling on business.
Can we safely assume the same for 8th rank Miami without going through the same process?

Washington D.C. deserves another closer look....
Washington D.C. to Atlanta = 636.33 miles
Washington D.C. to Miami = 1057.71 miles
Washington D.C. to Houston = 1410.16 miles
Washington D.C. to Philadelphia = 139.15 miles
Washington D.C. to Dallas = 1327.75 miles
Washington D.C. to Chicago = 699.39 miles
Washington D.C. to Los Angeles = 2672.4 miles
Washington D.C. to New York City = 227.35 miles
Wow, there are acceptable cities less than 250 miles. Is it a coincidence Acela trains go to both of them from D.C.? If you look closely, there's only one other city pair on this entire list that's closer than 250 miles apart. That's Dallas and Houston, which is 238.7 miles apart, which Amtrak doesn't even serve directly, and only serves indirectly (requiring an overnight transfer in San Antonio) only three days a week.
P.S. San Francisco MSA population is ranked 11th, while Boston MSA is ranked 10th. Boston is also serviced by Acela, and CHSR will hopefully service San Francisco.

Except for the cities on the Northeast corridor, only one top 10 MSAs city pairings is closer than 600 miles, and Amtrak doesn't even serve it daily. All the other major top 10 MSAs in America are further than 600 miles apart. And I think we should agree that a majority of the intercity traffic in America should be between the top 10 MSAs. Are you surprised that the average long distance Amtrak trip averages over 600 miles now?

I think that much the same argument could be made about Europe and eastern Asia.

First, electricon, it looks like you are using highway distance instead of great-circle distance (the geometrically shortest). I decided to compare the two for DC-LA Union Stations using Google Maps:
Great Circle: 2298 mi
Highway (I-70): 2671 mi

Since Google also has European highway distances, let's see what those turn out to be. Working from High-speed rail in Europe - Wikipedia:
Paris - Lyon: 289 mi
London - Paris: 282 mi / 309 mi
London - Marseille: 759 mi / 787 mi
London - Madrid: 1066 mi / 1381 mi
London - Rome: 1191 mi / 1239 mi
Berlin - Madrid: 1442 mi / 1644 mi (B - Hannover - Basel - Lyon - Avignon - M)
The second figure is for HSR estimates, found from approximations of the HSR routes by diverting into various towns along the way, like Lille, France.

Turning to the other end of Eurasia,
High-speed rail - Wikipedia
Tokyo - Osaka: 318 mi
Aomori - Kagoshima: 1198 mi / 1293 mi
Beijing - Tianjin: 86 mi
Beijing - Shanghai: 786 mi / 861 mi

lpetrich wrote:First, electricon, it looks like you are using highway distance instead of great-circle distance (the geometrically shortest). I decided to compare the two for DC-LA Union Stations using Google Maps: Great Circle: 2298 mi Highway (I-70): 2671 mi

I used highway distances as listed by Yahoo Maps. I sincerely believe highways distances more closely reflect rail distances than air distances.

While London to Madrid is close to 1400 miles, what is the HSR share compared to air for that city pair?
From http://www.traintoparis.com/eurostar.html Eurostar has 71% market share of London to Paris and 65% of London to Brussels travel routes.
From http://www.publications.parliament.uk/p ... 5vw154.htm35. The CCC believed that future rail market share on routes from London to mainland Europe would be largely dependent on integration of the European network. A number of European countries have, or plan to have, high speed rail networks. The analysis commissioned by the CCC concluded that with integration, high-speed rail could gain a market share of 30-60% on routes such as London to Amsterdam, Dusseldorf and Frankfurt, and see some increase possible on routes already well served by high-speed rail (eg London to Paris and Brussels). But, there is limited scope for significantly increased market share on longer distance routes, such as London to Berlin, Milan, or Madrid.
Less than 500 miles (30-60% market share)
London to Amsterdam is 334.02 miles
London to Dusseldorf is 354.9 miles
London to Frankfort is 476.31 miles
More than 600 miles (no gain)
London to Berlin is 680.6 miles
London to Milan is 782.66 miles
London to Madrid is 1381 miles

Again I repeat, outside the Northeast Corridor cities already serviced by Acela trains, only one top 10 MSA city pair in America has a distance less than 600 miles. Distance equals elapse time, and at a certain distance/time planes will continue to win market share over high speed rail trains. And in America planes will continue to be the choice of the majority of intercity travelers because the distances between major metros are so large.

electricron wrote: Eurostar has... 65% of London to Brussels travel routes.

I could be mistaken, but do you suppose this number may be inflated by the fact that some US > London flights then add on a stub flight LHR > BRU as a way to catch extra revenue. As in: hey look, we fly to Brussels too! (For some reason I"m thinking United does this, but I can't remember.)

I wonder if it is possible to tease out the London > Brussels passengers who did not originate their travel in London. That may deflate the non-rail numbers. Or it may be a very tiny number not worth teasing out.

Why does HSR have to compete with air travel? Why can't it compete with road travel with any gains over air travel being a happy coincidence?

I'm not disagreeing with you, Matt, but it is a valid metric, and one that Amtrak uses as comparison for NEC travel.

Jeff Smith wrote:I'm not disagreeing with you, Matt, but it is a valid metric, and one that Amtrak uses as comparison for NEC travel.

It's also a metric all HSR operators in the world use to justify state subsidies, not just Amtrak.

But I am tired of HSR supporters bringing up claims of successes from Asia and Europe to support widespread HSR lines all over the USA between city pairs that are more than 600 miles apart. HSR in the USA will be effective in only just a few city pairs, more regional than national in usage.
As the midwest region Chicago Hub plan is being planned and implemented, 79 to 110 mph operations are more likely because the costs are far less. FEC Miami to Orlando proposal is going to be more useful for more of the State of Florida than the Tampa to Orlando proposal two years ago, and implemented at half the total price as projected today.
As for the national long distance Amtrak network, it sure would be nice if they were capable of running at full Class IV passenger speeds of 79 mph everywhere. I doubt if half of the track milage Amtrak runs on does. I would rather see Federal dollars being used to eliminate slow orders and expanding services than building dreams of a national HSR network. Once most of the rails are up to 79 mph max speeds, then start a program upgrading the freight tracks to Class V and 90 mph max speeds.

Sizes of cities is only part of the story. Distances between cities also count in determining the number of trips. The farther apart two places are, the fewer the people who will want to travel between them. An often-used trip-forecasting model is the "gravity model" (Trip distribution - Wikipedia):
T(i,j) = N(i)*N(j)*f(travel time)

where T is the number of travelers, N is the populations at the two ends, and f is a declining function, sometimes modeled as f(t) = a*(t^b)*exp(-c*t)

However, I haven't been able to find the numbers that various trip-planning modelers have used, so I can't say how the size-vs-distance tradeoff works out.

As to HSR systems, I think that European systems now extend far enough to make possible some estimates of how many people are willing to ride such distances in trains as opposed to airplanes.
Amsterdam - Marseille: 826 mi
London - Marseille: 790 mi
Barcelona - Malaga: 731 mi
Turin - Naples: 569 mi