Why aren't overnight trains able to compete with flying?

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.
LSL typically runs 10 to 12 cars; the Silvers used to run *18* back in the day and the platforms are still long enough for it. Passenger trains can be, and have been, even longer than that without difficulty.
 
OK, conceded, very-long-haul in very large planes might be more efficient. I don't actually care, because the number of people who are going to consider trains on such three-day routes are... basically just the people who will not fly at all, like me.

I was thinking about stuff like NYC to Chicago or DC to Chicago. Or NYC to Miami -- at the longest. Or Buffalo to Chicago and Syracuse to Chicago -- the last of which is my typical trip. Chicago to Detroit, Detroit to NYC. Those are the ones where I did look up the numbers a while back. These single overnights can be time-competitive with flying, if you consider the overnight to be an alternative to sleeeping in a hotel room (the time spent sleeping doesn't really count against the travel time).

These are *much shorter trips*. Maximum 1300 miles, not 2700. A much larger percentage of the flight time is going up in the air and down again. Coasting is the efficient part of the flight; taking off and going up is the part which generates massive emissions. These routes are also, invariably, done in smaller planes. Think MD-80, not 737.

touche! Short haul flights, in my opinion, need to end! Though, I'd be curious where the cutoff is for distance/emission ratio.

MD-80's and 737's are roughly the same size, and have the same range/fuel/capacity constraints, so therefore would fulfill the same market.
Of course, the many variants of a 737 do make it complex, but 737's are a short-medium haul aircraft. The MD80 and 737 classical were introduced around roughly the same time as well. The only reason the 737 was further developed into a medium to semi-long haul aircraft was to meet the demand created by the 757 retirment void.

In terms of fuel burn, airplanes burn significantly less on decent, and it ends up being roughly canceled out in terms of ascent and descent.

No. 6 sleepers plus diner and ssl is not a long train to begin with. That’s a typical Amtrak western consist just swapping out 3 coaches for 3 sleepers.

That's all fine.
My point still stands. I don't believe a three day, all sleeper, P42 pulled train is more efficient than the new airliners on trans continental (or even Chicago-west coast) distance.
 
Last edited:
That's all fine.
My point still stands. I don't believe a three day, all sleeper, P42 pulled train is more efficient than the new airliners on trans continental (or even Chicago-west coast) distance.

Umm... your own math doesn’t seem to support that unless I missed something.
 
Just a couple of anecdotes.
My beautiful wife, known here as Whiterabbit, recently flew from DFW to Phoenix and back. Her flight out was delayed due to mechanical problems for an hour. For a 2.5 hour flight. That is appx 40% of the scheduled travel time as a delay. On her way back her flight was delayed a whopping 4 hours due to weather problems somewhere (it was clear there and here in DFW, so I cannot attest to what the weather problem was). That is appx 160% of the scheduled travel time.
So in those terms a train ride of say for instance, 30 hours of scheduled travel time, that was late on arrival at the endpoint of 12 hours would equivalent, and return trip that was late by 48 hours would be equivalent.

In actual hours it is all a lot less, but in perspective of scheduled travel time, most trains are pretty decent.

Just an anecdote, as always, YMMV.
 
I am trying to catch up on this thread but it took me til now to realize that although it is 800 miles from Penn Station to Chicago Union Station (by car) that it takes 20 hours for the LSL to make the trip. Holy Heck! I realize that going via Albany, Rochester and Buffalo adds a lot of miles but, man! I thought that it would be a 16 to 18 hour trip which is a bit long for a sleeper train but 20 hours is pretty much too long to attract business travelers. Even if the NY to Chicago sleeper had a nearly direct line, it would need to travel at better than 60 mph average for the 900 miles it would take on a shorter route to make it in a reasonable time give 79 mph max speeds. And 15 hours would mean that leaving NY at 7pm would pretty much guarantee that you arrive in Chicago after 10am, which is going to make business travel problematic.
Given our pokey max speeds in the US it looks like city pairs for overnight trains would have to be less than 700 miles apart to keep the total trip time under 12 hours.
Sorry for having stated what has been obvious for most of the rest of you, but I hadn't put the numbers together until now. we really need IC speeds of 125 mph on a substantial part of the Amtrak system to make most city pairs work for an overnight sleeper service. Europe has a lot of large'ish cities that are between 400 and 700 miles apart. The US does not. An American overnight train that averages 90 mph instead of 55-60 mph would mean that the 700 mile radius of acceptable city pairs would be more like 1000 miles which would really open up the map a good bit.
Again, sorry for stating the obvious but I didn't realize just how bad the LSL was for overnight business travel.

In principle, I agree. NYC-Chicago is a very poorly planned, and underutilized market. OH politics are partially the reason though.

If the transportation politics there looked a little more like VA, we’d have a much different situation, and maybe a more rail friendly state, with routes in between some of the major cities in OH.
 
The 20th Century Limited made the trip over essentially the same route as the Lake Shore in 16 hours. Of course, it made a lot fewer stops and those it did make were either receive only or discharge only.

The distance shown from GCT to La Salle St Station is 960.7 miles. It was doing an average of 60 mph inclusive of all stops and servicing.

Source is the January 1961 Official Guide.

Of course, the 20th Century owned the railroad, and much of the line was quadruple track east of Buffalo.
 
I beg to differ about fuel on descent . For a JT8D-15 at a cruise altitude depending on weight of aircraft fuel consumption rate is about 2100 #/hour. At idle descent fuel rate starts about 900 # linear until at about 10,000 feet between 1100# and 1200 # / hour. Usually some power will be needed for any of various reasons once going below 10,000 feet
 
According the Fred Frailey in Twilight of the Great Trains, SP kept close track of directly attributable above the rail cost. If a train was breaking even, they'd leave it alone. Once it went in the red, they'd go after it. They then used allocated indirect costs to make cost performance look worse to regulators.

Also in Twilight of the Great Trains passenger-friendly Santa Fe, in deciding whether or not to join Amtrak, estimated that passenger losses would wipe out all freight profits by 1975 if Santa Fe had to operate their entire then existing fleet, as the legislation called for.

Yes, the losses were real. Bear in mind, to discontinue a train they had to prove it was a "burden" on interstate commerce to the ICC. They could not just discontinue trains at will.
I suspect that many trains covered their direct costs and probably made a modest profit. What killed them was that some railroads really melted down in their service, and when the connections got bad, that killed patronage on other trains. It has to be a national system, or it doesn’t work.
 
These MPG equivalence numbers sound very extreme. I would love to see the calculations that produced these values. Also not being considered would be the needed increase in capacity of our electrical system, and the number of houses/apartments that would need upgrades in the capacity of their service entrances in order to have the capacity to charge the cars. Never forget the issues with materials, manufacturing, and disposal of the batteries. I am sure by now most of us have learned that batteries in our phones and computers have a finite life. Cars will not be any different, except the batteries will be MUCH larger.

When we go back to building large nuclear plants we might be getting more rational, but as long as we are generating electricity using current other-than-nuclear technology we won't be.

The reason we see no electrification on freight lines is that it makes no sense. We actually are electrified, it is just that the engine is carrying the power plant on its back. The weight on drivers is required for traction, particularly in territory with significant grades. The modern diesel is sufficiently efficient that it is questionable whether the efficiency of central plant generated BTU's produced would be enough to balance out transmission losses and additional system maintenance sufficiently to make it viable over the current diesel locomotive, which far exceeds that of steam locomotives or even the earlier diesel models.

Personally I agee that they environmentalist are probably in for a rude awaking in both environment damage from mining for batteries and other minerals And the wind generators around here are a total eyesore for those who have to contantly hear them swishing as they turn. However a Nuclear plant has a terrible impact should something go badly like they had in Japan and Russia. Too much risk for me even if they are efficent.
 
One of my feelings is that train travel in general is so out of the public consciousness and comfort that the thought of trains providing any reasonable alternative to driving/flying is just not in the question; most are surprised by the level of service provided on the NEC.
When I would travel on business to Japan or Europe, I'd have to argue with my travel office about why they were going to fly me from Tokyo to Osaka or Oslo to Trondheim when I could take the train just as quickly or quicker. They had a airline mindset. If you wanted to go any other way, you were probably up to something. They didn't even seem to have provisions to book train transportation. You had to do that on your own and get reimbursed later.
 
I beg to differ about fuel on descent . For a JT8D-15 at a cruise altitude depending on weight of aircraft fuel consumption rate is about 2100 #/hour. At idle descent fuel rate starts about 900 # linear until at about 10,000 feet between 1100# and 1200 # / hour. Usually some power will be needed for any of various reasons once going below 10,000 feet
It’s still pretty minimal compared to cruise and climb. A v2500 and a pw1100g are between 600-800lbs/hr in descent in an a321ceo/neo. Fuel burn on the takeoff roll up to 10,000 can be as high as 12,000lbs/hr per engine. Cruise depending on weight is between 2500-3200 lbs/hr/eng depending on ceo vs neo and weight.
 
There are so many routes that Amtrak could successfully market a 6pm departure and a 8am arrival that currently have heavy airline traffic they could capture from, but it would require reworking the train schedules and getting money from states, since nearly all the routes are under the threshold for LD Service.
  • CHI-MSP
    • Currently 19 flights per day (only checked O'Hare airport)
    • Currently a 7' 45" train ride. Could depart CHI 10pm, arrive into MSP 7am.
  • CHI-DTW
    • Currently 18 flights per day (only checked O'Hare airport)
    • Currently a 6' train ride. Could depart CHI 10:30pm, arrive into Detroit at 6am (or later with some more schedule padding).
  • CHI-KCY
    • Currently 10 flights per day (only checked O'Hare airport)
    • Currently a 7' train ride. Could depart CHI 11:00pm, arrive into Kansas City at 7am.
  • NYP-BUF
    • Currently 12 flights per day (looking at all 3 NYC airports)
    • Currently a 8' train ride. Could depart NYP at 10pm, arrive into Buffalo at 7am.
  • NYP-PGH(Pittsburgh)
    • Currently 15 flights per day (looking at all 3 NYC airports)
    • Currently a 9' train ride. Could depart NYP at 9pm, arrive into Pittsburgh at 7am.
  • SEA-EUG
    • Currently 10 flights per day
    • Currently a 7'18" train ride. Could depart SEA at 10pm, arrive into Eugene at 6am.
  • SEA-SPK
    • Currently 21 flights per day
    • Currently a 8' train ride. Could depart SEA 10:30pm, arrive into Spokane at 7am.
The list could go on and on. Granted some of these flights have connecting passengers, but there's plenty of O&D traffic at these airports. I could see a lot of business travelers looking to avoid 6am flights (which many companies booking software forces to be booked, as they're usually the lowest price). Going to Pittsburgh from NY for a 9am meeting? I'd much rather be arriving on Amtrak then flying down at 6am out of LaGuardia.
 
Regardless of how profitable some new routes could be - unless the freight railroads that own the tracks agree ... it doesn't even matter how good the idea is, how many riders it could attract or how willing Amtrak is - it won't happen
 
Regardless of how profitable some new routes could be - unless the freight railroads that own the tracks agree ... it doesn't even matter how good the idea is, how many riders it could attract or how willing Amtrak is - it won't happen
Amtrak seems to be doing a good job playing hardball with NS & CSX for the new corridor to Mobile. If they notch a win on this route, it could lead to more wins on overnight corridors like the above.
 
As mentioned before, overnight trains are not just for people going from city center to city center in either terminus. People like me could very easily be traveling from one suburb to another. It’s not like downtowns of major cities have a monopoly on office space. Someone could very easily be traveling from Fairfield to Ventura for business or Albuquerque to Littleton. Not having to get up at 0 dark 30 for a flight and take Uber across town would also be a selling point if Amtrak ever had a lie flat seat priced to be competitive on these sorts of trips. Which also brings up another point that Amtrak would have to sell to companies. The fact that you don’t have to pay for a hotel if you make them fly at the end of the day or make someone get up early as hell to avoid paying for a hotel on a one day trip.
 
There are so many routes that Amtrak could successfully market a 6pm departure and a 8am arrival that currently have heavy airline traffic they could capture from, but it would require reworking the train schedules and getting money from states, since nearly all the routes are under the threshold for LD Service.
  • CHI-MSP
    • Currently 19 flights per day (only checked O'Hare airport)
    • Currently a 7' 45" train ride. Could depart CHI 10pm, arrive into MSP 7am.
  • CHI-DTW
    • Currently 18 flights per day (only checked O'Hare airport)
    • Currently a 6' train ride. Could depart CHI 10:30pm, arrive into Detroit at 6am (or later with some more schedule padding).
  • CHI-KCY
    • Currently 10 flights per day (only checked O'Hare airport)
    • Currently a 7' train ride. Could depart CHI 11:00pm, arrive into Kansas City at 7am.
  • NYP-BUF
    • Currently 12 flights per day (looking at all 3 NYC airports)
    • Currently a 8' train ride. Could depart NYP at 10pm, arrive into Buffalo at 7am.
  • NYP-PGH(Pittsburgh)
    • Currently 15 flights per day (looking at all 3 NYC airports)
    • Currently a 9' train ride. Could depart NYP at 9pm, arrive into Pittsburgh at 7am.
  • SEA-EUG
    • Currently 10 flights per day
    • Currently a 7'18" train ride. Could depart SEA at 10pm, arrive into Eugene at 6am.
  • SEA-SPK
    • Currently 21 flights per day
    • Currently a 8' train ride. Could depart SEA 10:30pm, arrive into Spokane at 7am.
The list could go on and on. Granted some of these flights have connecting passengers, but there's plenty of O&D traffic at these airports. I could see a lot of business travelers looking to avoid 6am flights (which many companies booking software forces to be booked, as they're usually the lowest price). Going to Pittsburgh from NY for a 9am meeting? I'd much rather be arriving on Amtrak then flying down at 6am out of LaGuardia.
Even if every single one of these city pairs could support a well-patronized overnight train, the vast majority of people traveling between each city pair will probably be either driving, flying or using a day train, if available. Compare with Northeast Regional 65/66/67 -- The sleeper has a capacity of about 30 passengers, but usually carries a lesser number. The coaches can hold a couple of hundred passengers, but many (most) of them are not necessarily riding the entire distance, as it's used by commuters for early morning arrivals. Even if you ran the train with 5 sleepers, that's a maximum capacity of 150 people, or the load of one airplane. Clearly, although an overnight train can possibly be practical and even successful, it will always serve a niche market, unless there is some cataclysmic event that makes commercial aviation impossible. Thus, if you're running a taxpayer-supported passenger rail service, providing such an overnight train is going to be a lower priority than providing daytime corridor service that can also serve intermediate stations and thus serve a larger potential market.
 
if you're running a taxpayer-supported passenger rail service, providing such an overnight train is going to be a lower priority than providing daytime corridor service that can also serve intermediate stations and thus serve a larger potential market.

I guess the same reasoning could be used when planning additional highways or expansion of current ones ... if they are not being used mostly by overnight travelers, then, since they are tax-payer supported, perhaps they shouldn't be built or expanded.

Then again, the view of various tax-payer supported transportation endeavors seems to differ depending on whos lobbying influence holds the purse strings to those making the decisions.

For real benefit to the tax-payers, flying, driving and rail should all be supported with equal enthusiasm - and we all know that is not currently the case.
 
I have been monitoring this thread for a while. There are a few routes that are long enough to support overnight service also serve an almost continuous string of cities and thus have the potential to provide daytime corridor service, connecting multiple city pairs, as well. Boston-NYC-DC-Newport News is an example. Chicago-Cleveland-Buffalo-NYC has the potential to be another one. In addition to the Lake Shore Limited it already serves as a corridor east of Buffalo, with several daytime trains. Chicago-Cleveland has been proposed for corridor service. But to be truly successful, both of these corridors need to be faster, i.e. faster than driving, and have more frequent service (perhaps every two hours). An interesting blog post on this concept appears here: Thoughts on a New York - Chicago unified corridor . Chicago-Indianapolis-Nashville-Atlanta-Florida is another one. A problem that needs to be resolved is that these corridors are longer than 750 miles and, thus, by definition are long distance, requiring Amtrak to operate them without outside support. Legislation will be required to establish a special category for such routes.
 
A huge obstacle to running European-style night trains in the US is lack of a budget high density sleeping option like couchettes. Instead of ditching couchettes Nightjet has focused on ways to make them better, like improved bedding and attendants to provide room service. They even solved their biggest drawback, lack of privacy, with capsule hotel like pods for solo travelers in their next generation rolling stock. The new couchette cars will have a very respectable 40 pax capacity, most of which will be individual accommodation.
 
We need a practical Musk or Branson privatization of this. This vacuum tunnel Musk is working on is not the answer. Would it be possible to buy right of way to create private, modern train travel? That had all the latest bells and whistles (electric train, space saving “sleeper chairs”, great WiFi, free smart screen, good food, ability to bring your Tesla “Tesla owners ride at half price!”, fast and direct routes)
 
We need a practical Musk or Branson privatization of this. This vacuum tunnel Musk is working on is not the answer. Would it be possible to buy right of way to create private, modern train travel? That had all the latest bells and whistles (electric train, space saving “sleeper chairs”, great WiFi, free smart screen, good food, ability to bring your Tesla “Tesla owners ride at half price!”, fast and direct routes)

Somehow, throwing billions upon billions at space is for the good of all mankind... not that it doesn’t have its merits. But surely there are better endeavors for the good of many.

The answer more or less is yes. Someone with enough will, vision and $$ could make it happen. But it doesn’t mean it’s the best answer.

Also, privatization is not always the answer for providing a public service for the good of all.

Given that privatizing is incentivized by money, and not by providing a public good, they can pull the plug at any moment they deem appropriate (like during a pandemic). Anyone who would rely on that “public” transportation would be screwed.

Amtrak, through thick and thin, manages to consistently provide a service (no matter how bad and bare bones it gets).

An ideal for me, would be a mix of the two. Government operation competing against private operations (like Europe), ensuring the competition stays relevant. Amtrak would up it’s game, and others would be incentivized to keep upping their game.
 
Back
Top