NARP Whitepaper on Long Distance Service

[jis]

Thought this might be of general interest and might generate some lively discussion here....

http://www.narprail.org/resources/white-papers/2083-longdistancepaper

 

[Shanghai]

Very interesting and informative. I hope it will be read and considered by our national leaders

and not become a "political football."

 

[TimePeace]

Very cool. I see a lot of ways of looking at LD trains that had not occurred to me. I know I like them, but now I know that other people have some different reasons for riding, and liking them, than I do.

 

[Paulus]

As usual, no consideration for the fact that the long distance routes lock up an absurd amount of equipment which could be better used for more frequent corridor service. No discussion of the costs of providing sleeper service, merely of the revenue.

Connecting passengers generated 89% more revenue than this route would have if operated as an isolated, stand alone corridor.

Or: State supported corridors and the NEC are responsible for nearly half of all long distance revenue (unless they're counting the revenue on those corridors of course).

Consider the route between Chicago and Cleveland. This route currently has just two trains a day in each direction—both with unattractive late night or early morning arrival and departure times at Cleveland. More than 11 million Americans live within 25 miles of one of the nine stations on this 341-mile corridor. 24This market should have hourly departures with a transit time of less than three hours. But this level of service will require long stretches of track dedicated to passenger trains. There will be substantial lead times to design, finance and construct this needed infrastructure.

A relatively low cost and low risk method to ex- pand service in this market would be to work in partnership with the Norfolk Southern and CSX railroads to add daytime frequencies to the existing Chicago–New York Lake Shore Limited route with stops in Toledo, Cleveland, Buffalo and 14 other cit- ies. If scheduled correctly, the trains would make it possible to schedule a reasonable day trip to all cit- ies. Even with a low share of the Chicago–Cleveland market, this long distance service would still gener- ate economically viable volumes because it would serve 171 city pair markets instead of just 36.25

Every additional frequency on the Lake Shore Limited would require three additional sets of equipment. Looking here, that's 13 cars and two locomotives per set. Let's cut it down to just the New York section, Boston only gets overnight service, that'll bring us down to a single locomotive and 8 cars. That's a minimum capital investment of $75 million per frequency, not including the cost of spare cars and locomotives nor the infrastructure investment costs (which, yes, CSX and NSC are going to demand, and in large sums). More than one, perhaps two, additional frequencies and I'd lay money on it being cheaper to simply run Chicago-Cleveland. This is ignoring, of course, the fact that the longer the trip, the more likely it will be to be delayed.

As for this being low cost, low risk, and relatively quick, just how many decades has California been trying to bring back the Coast Daylight again? That's a simple second frequency of the Starlight from Oakland to Los Angeles and yet it is taking amazingly long to get it going.

A diesel-powered passenger train can move nearly two and one-half times more people per gallon than a typical automobile.

Amtrak really doesn't do well when it comes to fuel efficiency and long distance trains are probably no more fuel efficient than a typical automobile due to FRA inspired heavy locomotive weight and low seating density. Unfortunately Amtrak doesn't break it down any better than nationally. However, on a national level, Amtrak is only about as green as driving a Prius by oneself.

Because trains use fuel efficiently and do not have a significant fuel penalty for stops, the cost of train travel is not as heavily influenced by fuel prices as the cost of air and road travel.

Metrolink's fuel related fare increases would beg to differ.

Long distance trains are cost efficient—a finding that may surprise many. Despite years of neglect, underinvestment and retrenchment, Amtrak’s cost to move one passenger one mile (the accepted industry measure of efficiency) is roughly the same for both long-distance and in-state routes on corridors outside the Northeast. This parity is not obvious in Amtrak’s financial reports because these reports include state—but not federal—payments for service as revenue.

More due to a higher load factor on LD than on state (which has some incredibly low load factors for California routes, which heavily contribute to both passenger and seat mile figures) as well as government imposed subsidies by host railroads with below market MOW and access costs. From Amtrak's FY2012 figures:

NEC cost per passenger mile: $0.418

Corridor cost per passenger mile: $0.413

Long distance cost per passenger mile: $0.386

NEC cost per seat-mile: $0.216

Corridor cost per seat-mile: $0.180

Long distance cost per seat mile: $0.241

So seat for seat a long distance train is 33% more expensive to run than a corridor train and 15% more expensive to run than the NEC.

Breaking things out somewhat:

Acela

Passenger mile: $0.477

Seat mile: $0.298

Northeast Regional

Passenger mile $0.379

Seat mile: $0.183

The Acela is an expensive train to run due to the small number of seats (only 304), high level of service, and probably extravagant energy consumption. However, it has a fairly significant number of passengers paying quite a large sum, making for high revenue to offset said costs.

Capitol Corridor

Passenger mile: $0.665

Seat mile: $0.191

Pennsylvanian

Passenger mile: $0.318

Seat mile: $0.191

The Capitol Corridor has the single worst occupancy ratio in the entire national train system; note how it inflates the per passenger mile cost compared to how it is with the Pennsylvanian, which is fairly healthy in terms of occupancy ratio. Since the second worst one is the Surfliner, the third is the Empire, #6 is the San Joaquin with the Keystone just edging above it, which represents the five routes with the highest frequencies and 38% of the corridor passenger miles, it's a bit unsurprising that per passenger-mile costs will look a touch higher than long distance trains. A certain degree of low occupancy is probably an unavoidable feature of higher frequencies, but quite frankly, marketing is abysmally low as well (California's current marketing contract calls for only $9 million over three years). However, note that these trains, despite low occupancy, generally recover more of their costs than do the long distance trains, even without state subsidies.

A comprehensive national system with more routes and greater frequencies will require a higher level of public support. But since many of these costs are fixed, expanded service would increase efficiency and lower the public cost per passenger mile.

Fixed overhead really isn't that big of a deal for the long distance trains; the killer for them is the wage costs which will remain disproportionately high for them because they have disproportionately large crews.

Congress could lower this cost even further by funding the purchase of modern, high performance trains to replace Amtrak’s aging long distance fleet and to provide the capacity needed to add extra cars to existing trains and to launch new service. New equipment costs less to maintain.

Eh, this is mealy mouthed. Congress has to fund the purchase of modern, high performance trains order to provide the expanded service that they want in the first place. Though high performance and long distance sleeper train really don't belong in the same sentence.

The fact that NARP includes a picture of a Chinese HSR sleeper is absolutely ridiculous. They make sense given the length of the Chinese high speed rail network; they do not make sense in any American context.

Frequently sold-out trains indicate that the demand exists to justify greater capacity.

Frequently sold-out trains also indicate that demand exists for higher priced trains.

Edit: Not sure how much this changes things, but the short distance corridors also get hit with the costs of moving all the equipment via Hoosier State, which inflates cost per seat mile and cost per passenger mile.

 

[leemell]

The Coast Daylight as is currently composed would not go to Oakland, but rather downtown San Francisco on the other side of the bay.. Not exactly a duplicate of the Coast Starlight.

As for this being low cost, low risk, and relatively quick, just how many decades has California been trying to bring back the Coast Daylight again? That's a simple second frequency of the Starlight from Oakland to Los Angeles and yet it is taking amazingly long to get it going.

 

[Paulus]

The Coast Daylight as is currently composed would not go to Oakland, but rather downtown San Francisco on the other side of the bay.. Not exactly a duplicate of the Coast Starlight.

I've seen both Fourth & King station and Oakland in Daylight material.

Edit: In any case, the change is over state owned trackage, it doesn't affect the freight trackage that's the hold up.

 

[Trogdor]

In my opinion, I'd have the basic LD system (don't really need multiple frequencies for long-distance trains), with corridors overlaid on it to provide higher frequency service at lower costs over the sections of the route that need it.

An example, given current railroad infrastructure (but ignoring the fact that the freight railroad owners would demand tons more investment), is the Lake Shore Limited route.

That route would do fine on its own, combined with the Empire corridor service, plus a few CLE-BUF trains (maybe 3-4 per day), a few CLE-TOL-DET trains, and a couple extra CLE-CHI trains (on top of the Capitol Limited). The corridor trains are really cheap to operate, because they can be done with minimal staff, and scheduled times can be optimized for the market that they serve. This model would make the Lake Shore cheaper to operate, because T&E crews could be more efficiently scheduled. With multiple frequencies, you could do better to have crew change points approximately every four hours, with a couple hour layover and then have them take another train back where they came from. This would cut down on crew hotel costs, held-away pay, etc., plus it would lower the spare ratio of crew required, because you would have a smaller extra board-regular assignment ratio.

No need to have three Lake Shore frequencies to do that, though maybe a second frequency would be handy, to get folks from out west into New York in the morning, and allow them to leave late evening.

 

[henryj]

Trogdor has the best solution, at least in my opinion. NARP is generally known to be a bunch of nostalgic foamers. Their support of LD trains is nice, but misguided. The LD train is just a land cruise/vacation train. It may provide for some transportation needs in certain areas, but for the most part it is just an anachronism. Particularly for a train such as the Lake Shore. No business man would use this train to get between NY and Chicago unless he was a secret rail fan. No one is going to voluntarily put them selves in a cramped little room with a public shower down the hall in a closet and get jostled all night long when they can fly to their destination in three hours and stay in a nice hotel with a restaurant and bar and a real bath and shower. Trains like the Lake Shore serve greyhound type passengers and sleeper passenger that are afraid to fly are or too old. Amtrak's western LD routes remain popular because of the scenery and people consider them part of the vacation. Trains like the Eagle serve a multitude of small communities that would otherwise be isolated. Amtrak's Florida service could be successful if they would spruce it up a little. As it is, they are just another bland Amtrak train except for the Auto train. The Crescent is a good train as far as Atlanta, but beyond it should just be a coach train. Missing links are Chicago to Florida thru Atlanta and Texas to Colorado and Portland. The real future for US passenger rail is short distance and fast corridor trains that can compete with the automobile and the plane.

 

[Anderson]

1) Interesting...it seems that LD load factors (CPSM/CPPM) are somewhere around 62%. On the NEC it's about 51%. On other corridors, it's around 43-44%. The first figure surprises me a bit (although not much anymore, considering how slammed the Silvers are coming out of DC half of the time). The second doesn't surprise me, since while you've got the Acela packed at peak hours, you also have partly-empty Regionals running late at night as well to balance things out...but the NEC is anchored at its ends by two major cities (BOS/WAS) and most of the not-through trains terminate at the largest (NYP). The third isn't that surprising...you've got a couple of state routes that have a fairly empty "tail" (as I understand it, a lot of the Midwest trains fill up as they approach CHI, the Vermonter tends to be fairly empty at SAB, etc.) as well as a real stinker in the mix (the Hoosier State leaps to mind, with <90 passengers per day). And, as noted, you've got your frequent, unreserved corridor trains with lower load factors (Capitol Corridor, Surfliner) as well, and they're not even doing that badly.

2) Yes, the long-distance trains are more expensive per seat-mile to operate. However, I'd point out that a number of them run with a second food service car (the diner) and a baggage car, so we know where at least some of the cost is. Some more would invariably come from held-away time and what-have-you for crew. At the same time, however, those LD trains also have more room for upsold accommodations (and indeed, likely demand for it as well), not to mention demand for longer consists.
-To highlight this, a 9-car Regional has IIRC 566 seats (7*72 seat coaches, 1*62 seat BC) and one cafe. The Silver Meteor tends to run with far less seats at maximum length (5*59 seat coaches for 295 coach seats, plus around 90 hypothetical slots in the sleeper, for 385 "seats") but has a second food car.
--The Meteor is actually one of the higher-capacity LD trains (a California Zephyr running with 3 coaches, 2 sleepers, and a transdorm at the moment has 252 coach seats, 96 hypothetical sleeper slots, for about 348 "seats"). Please consider that in both cases, the sleeper is likely to run with about a quarter to a third fewer riders because of single-occupancy use of roomettes (and that on the single-level trains, a certain amount of sleeper space presently goes to crew and not revenue use).
--To wit, the problem with the LD trains' cost isn't that they're more expensive than the Regionals as such, but rather that the costs of the OBS aren't distributed among as many seats. If you could add another two coaches and a sleeper to a Silver, for a quick example, you'd bring the number of slots available on-board in like with that 9-car Regional. Doing this across the LD system (even if some of the cars were only run for part of a train's run a la the CHI-MSP or CHI-STL cars) would likely bring the CPSM down substantially.

3) Moving on to the number of sets needed per frequency, that's an aggravatingly variable figure. For example, the Capitol Limited needs three sets at the moment (for once-daily service). You could probably add a frequency with only a single additional set if you time both trains out right. Likewise, extending the Palmetto back to Miami would require 2, 3, or 4 sets depending on how all three Florida trains were timed (and/or routed). Out west, the CHI-West Coast trains tend to need five or six sets...depending on their timing/turning in CHI and on the West Coast. Moving back east, the Cardinal needs two sets for 3x weekly operation. It would only need a single additional set for daily operation, and I can't even guess what would be needed for a hypothetical second frequency on the route.

Edit: 4) Henry chimed in while I was away. I disagree with what he's saying, to at least some extent. There are definitely markets to be had for "long distance" or overnight train service. Florida is one such market (though as he noted, the trains could use some beefing up). I'm not going to dismiss the NEC-CHI market, especially given the cost of flying these days, and this is going to go double for some of the secondary/intermediate markets that force transfers when flying and turn those trips into /very/ long (and expensive) days.

Basically, at least as I see it, if you're burning a full day from end to end flying (getting to the airport by 8 AM and not getting out of the parking lot at your destination until around 6 PM), you're burning at least one night in a hotel that you wouldn't otherwise at the same time and you've got multiple hours of "dead time" either waiting in the terminal or with takeoff/landing where you can't even use the internet. I don't find it to be hyperbole to say that a friend who did that today wouldn't have broken even on the experience had he been able to take a pair of overnight trains (late evening departure from OKC for CHI on the Texas Chief, followed by the Charles Nelson Riley) and get into Newport News by 10 AM.

But setting aside that (admittedly long) trip, ignore NYP-CHI as your sole market and remember that the LSL also covers ALB/BUF-CHI and CLE/TOL-NYP as well, and I am going to doubt that those intermediate markets are going to be sitting as happy as Newark-O'Hare might be. And of course, there's PGH (on the old Broadway run) to look at as well.

 

[Paulus]

--To wit, the problem with the LD trains' cost isn't that they're more expensive than the Regionals as such, but rather that the costs of the OBS aren't distributed among as many seats. If you could add another two coaches and a sleeper to a Silver, for a quick example, you'd bring the number of slots available on-board in like with that 9-car Regional. Doing this across the LD system (even if some of the cars were only run for part of a train's run a la the CHI-MSP or CHI-STL cars) would likely bring the CPSM down substantially.

To an extent, but bear in mind that this may well require an additional locomotive or diner which would raise the prices right back up again. Also, take a look at Palmetto CPSM vs Silver Service: Meteor is 38% more expensive and the Star 54% despite the fact that they're really about equivalent in terms of the number of seats and mostly run over the same route.

Basically, at least as I see it, if you're burning a full day from end to end flying (getting to the airport by 8 AM and not getting out of the parking lot at your destination until around 6 PM), you're burning at least one night in a hotel that you wouldn't otherwise at the same time and you've got multiple hours of "dead time" either waiting in the terminal or with takeoff/landing where you can't even use the internet. I don't find it to be hyperbole to say that a friend who did that today wouldn't have broken even on the experience had he been able to take a pair of overnight trains (late evening departure from OKC for CHI on the Texas Chief, followed by the Charles Nelson Riley) and get into Newport News by 10 AM.

If you're burning a full day flying, you're either in layover hell or a transcontinental flight. Now, there might be other reasons to burn the day (I'd always block out the day from any major plans simply due to a tendency towards air sickness), but to, in, and from the airport shouldn't take that long. Perhaps I've been lucky with air travel however.

 

[Anderson]

--To wit, the problem with the LD trains' cost isn't that they're more expensive than the Regionals as such, but rather that the costs of the OBS aren't distributed among as many seats. If you could add another two coaches and a sleeper to a Silver, for a quick example, you'd bring the number of slots available on-board in like with that 9-car Regional. Doing this across the LD system (even if some of the cars were only run for part of a train's run a la the CHI-MSP or CHI-STL cars) would likely bring the CPSM down substantially.

To an extent, but bear in mind that this may well require an additional locomotive or diner which would raise the prices right back up again. Also, take a look at Palmetto CPSM vs Silver Service: Meteor is 38% more expensive and the Star 54% despite the fact that they're really about equivalent in terms of the number of seats and mostly run over the same route.

Actually, the Meteor/Star difference is illustrative here: The Meteor often runs with 5 coaches and 3 sleepers, while the Star tends to run with 4 and 2. Likely not by coincidence, the Meteor seems to recover just over 100% of direct costs while the Star falls short on that front.

I do know what you're saying on the equipment front; however, a fully utilized 48-seat diner should be able to host somewhere between 96 and 144 passengers at dinner (2-3 full seatings). While I noted the 90 hypothetical slots in the sleeper (30/car), you're more likely to end up somewhere around 72 right now (assuming 3 Viewliner I sleepers plus a bag/dorm) since those roomettes will tend to end up with a mix of lone travelers and pairs (I generally assume half-and-half), and about 68 with Viewliner IIs (due to the elimination of a revenue room for the toilet).

We've been around the mulberry bush a few times in here, but it seems that on a single-level train with a decent crew, you ought to be able to handle a fourth sleeper; it's more the crush of coach passengers that is at issue. Of course, you've got three constraining factors in the diner:

-Seating space, which could be addressed by using space in the cafe, or through some sort of at-seat meal service option.

-Food storage space, which could be addressed by cold storage space in part of a baggage car.

-Kitchen/cooking space, which is the only thing with the potential to demand a second diner, but which could also be ameliorated with food contracts in the vein of the Builder (or, with the Meteor and Crescent, with an adapted version of the Acela First Class meal service that's picked up at WAS/NYP SB), or through switching to a diner+diner/club setup instead of a diner+cafe setup.

As to locomotives, please keep in mind that an additional locomotive doesn't equal additional crew. It will add to fuel and maintenance, yes, but you could run sixteen locomotives with a single operating crew.

 

[Paulus]

Something that could probably help out the long distance trains and food service is to

1. Have the diners open longer

2. Update Arrow to allow pre-purchase of meals or pre-reservation of meal times taking advantage of 1.

Edit instead of double post: Palmetto is tied for 2nd for highest per-mile fare yield amongst long distance trains (with Crescent at 20.1¢, Auto Train holds lead at 32¢) despite a lack of sleepers. This strongly suggests to me that coach fares (and by extension sleepers, since they pay basic rail fare plus accommodation) could rise 50-100% without significantly impacting ridership.

 

[OBS]

Basically, at least as I see it, if you're burning a full day from end to end flying (getting to the airport by 8 AM and not getting out of the parking lot at your destination until around 6 PM), you're burning at least one night in a hotel that you wouldn't otherwise at the same time and you've got multiple hours of "dead time" either waiting in the terminal or with takeoff/landing where you can't even use the internet. I don't find it to be hyperbole to say that a friend who did that today wouldn't have broken even on the experience had he been able to take a pair of overnight trains (late evening departure from OKC for CHI on the Texas Chief, followed by the Charles Nelson Riley) and get into Newport News by 10 AM.

But setting aside that (admittedly long) trip, ignore NYP-CHI as your sole market and remember that the LSL also covers ALB/BUF-CHI and CLE/TOL-NYP as well, and I am going to doubt that those intermediate markets are going to be sitting as happy as Newark-O'Hare might be. And of course, there's PGH (on the old Broadway run) to look at as well.

Anderson thanks for mentioning a favorite actor of mine, Charles Nelson Riley (sp), but I assume you meant the James Whitcomb Riley...LOL

 

[jis]

I don't find it to be hyperbole to say that a friend who did that today wouldn't have broken even on the experience had he been able to take a pair of overnight trains (late evening departure from OKC for CHI on the Texas Chief, followed by the Charles Nelson Riley) and get into Newport News by 10 AM.

Anyone who would not mind spending two nights and a day and a half doing a journey that by other means would take 8 to 10 hours, well I am sure they'd spring for it. But I don't see that as a common phenomenon unless the fare differences get to be truly enormous. It involves two train rides each of which is longer than the entire journey by the other despised mode, by a factor of 2 or 3, and a lengthy layover in between.
I suspect if there were a direct train from OKC to NPN it could do the 1300+ miles in what the SWC takes from Chicago to ABQ and then some, since the track quality between OKC and NPN ain't BNSF's.

 

[Anderson]

OBS: You guessed right...ask Charlie about where that line comes from.

jis: A good part of my hangup is that the cost of travel includes that night in the hotel room before/after the trip. Probably the two best examples of this growing up were (1) When we had to stay at a hotel at BWI the night before a flight to Arizona because it left so early the next day and (2) When we had to do a two-transfer saga to avoid having to fly out of Norfolk at like 7 AM on the way to Des Moines (and got in on about the last flight of the evening into DSM), since that would've involved getting up at about 4 AM. I know this is me speaking, but "hotel time" on one (or both) ends of a flight goes into my equation as well.

Alan: Might want to check the formatting there. The board's code gremlin has struck again.

 

[AlanB]

As usual, no consideration for the fact that the long distance routes lock up an absurd amount of equipment which could be better used for more frequent corridor service. No discussion of the costs of providing sleeper service, merely of the revenue.

As a NARP White Paper from 2008 showed, sleepers cover their incremental costs. So much so that in fact, at that time the subsidy per passenger mile for all on the train was $0.1870. If one removed sleepers and their revenue, then coach passengers would have been subsidized at a rate of $0.1888 per passenger mile.

Seeing as how sleeper prices have risen dramatically since then, far faster than inflation would have affected the costs, I can only assume that they saw no reason to rehash those numbers and that things have actually improved since then with the excess sleeper revenue covering still more of the costs to move the coach passengers.

 

[AlanB]

A diesel-powered passenger train can move nearly two and one-half times more people per gallon than a typical automobile.

Amtrak really doesn't do well when it comes to fuel efficiency and long distance trains are probably no more fuel efficient than a typical automobile due to FRA inspired heavy locomotive weight and low seating density. Unfortunately Amtrak doesn't break it down any better than nationally. However, on a national level, Amtrak is only about as green as driving a Prius by oneself.

According the US Department of Energy, on average Amtrak consumes 2,271 BTU's per passenger mile. And while Amtrak might not break it down, it is clear that it has to be either the NEC or the LD's or both that keep Amtrak's numbers down, since commuter trains consume 2,897 BTU's per pax/mile.

The average car consumes 3,447 BTU's per pax/mile. Only the motorcycle beats Amtrak, or any rail transportation for that matter.

 

[CHamilton]

OBS: You guessed right...ask Charlie about where that line comes from.

Several of us used to commute on MARC back in the 70s, and we invariably got delayed by the James Whitcomb Riley. So it became a running joke to call the train the Charles Nelson Reilly instead, since he was all over the talk shows and game shows at the time. Pretty lame joke, actually, but I've heard worse around here

 

[AlanB]

Consider the route between Chicago and Cleveland. This route currently has just two trains a day in each direction—both with unattractive late night or early morning arrival and departure times at Cleveland. More than 11 million Americans live within 25 miles of one of the nine stations on this 341-mile corridor. 24

This market should have hourly departures with a transit time of less than three hours. But this level of service will require long stretches of track dedicated to passenger trains. There will be substantial lead times to design, finance and construct this needed infrastructure.

A relatively low cost and low risk method to ex- pand service in this market would be to work in partnership with the Norfolk Southern and CSX railroads to add daytime frequencies to the existing Chicago–New York Lake Shore Limited route with stops in Toledo, Cleveland, Buffalo and 14 other cit- ies. If scheduled correctly, the trains would make it possible to schedule a reasonable day trip to all cit- ies. Even with a low share of the Chicago–Cleveland market, this long distance service would still gener- ate economically viable volumes because it would serve 171 city pair markets instead of just 36.25

Every additional frequency on the Lake Shore Limited would require three additional sets of equipment. Looking here, that's 13 cars and two locomotives per set. Let's cut it down to just the New York section, Boston only gets overnight service, that'll bring us down to a single locomotive and 8 cars. That's a minimum capital investment of $75 million per frequency, not including the cost of spare cars and locomotives nor the infrastructure investment costs (which, yes, CSX and NSC are going to demand, and in large sums). More than one, perhaps two, additional frequencies and I'd lay money on it being cheaper to simply run Chicago-Cleveland. This is ignoring, of course, the fact that the longer the trip, the more likely it will be to be delayed.

That is not a correct assumption that the LSL would need 3 more sets for each frequency. Back when the Palmetto was the Silver Palm and Amtrak was running 3 trains to Miami, they weren't tying up 12 trainsets. At the moment I no longer recall if it was 10 or 11, but still they were able to rotate sets through in such a way such that they did not need 4 sets for each train like they do today.

 

[OBS]

OBS: You guessed right...ask Charlie about where that line comes from.

Several of us used to commute on MARC back in the 70s, and we invariably got delayed by the James Whitcomb Riley. So it became a running joke to call the train the Charles Nelson Reilly instead, since he was all over the talk shows and game shows at the time. Pretty lame joke, actually, but I've heard worse around here

Thanks for the info!

 

[jis]

The more important point I think is that the expected clientale of a train that does as quick a run as possible overnight and another that is a day train would be substantially different so as to require different kinds of accommodation. At least in countries where they run serious LD passenger service they have almost exclusively sleeper service on overnight trains with very little sitting accommodation and almost exclusive Coach and Excutive sitting accommodation with very little if any Sleeping accommodation on day trains.

A train, the main purpose of which, is to provide daytime service from New York to Toledo or Buffalo to Chicago would be much more efficient with predominantly Coach accommodation run like the Palmetto, than one with all the bells, whisltes and expenses of a fullfledged sleeper/diner train.

OK here are some random thoughts......

If I were trying just to get the New York - Chicago market straightened out, I'd probably run the following:

1. LSL more or less leace as is westbound, change eastbound to depart 3 hours earlier and run 3 hours earlier. 3 consists full blown Sleeper/Diner/Lounge/Coach + 2 P42 set

2. A day train departing NYP at 6:30 - 7am getting into Toledo around 9pm and if it runs through to Chicago arriving there at after midnight, but there is really no reason for it to do so. In the recverse direction leave Toledo again around 7am and get into new York around 9pm. No diner and sleeper just coach and BC and Dinette. 2 consists 6 Coaches, 1 BC 1 Dinette + 1 P42 set. With a little adjustment it could possibly be extended to the Detroit area.

3. A day train Chicago to Albany depart Chicago 6:30 to 7am arrive Albany around 11pm. In the reverse direction depart at around 7:30 am arrive Chicago by 11pm. 2 consists 6 Coaches, 1BC 1 Dinette + 1 P42 set

4. Extend a suitable Keystone (with different equipment of course to Toledo/Detroit getting there around 8 to 10pm, and heading back an hour behind the New York train.

5. Pennsylvanian to Cap through cars as are planned now

This immediately gives you massive coverage of the population centers in all routes between the East Coast and Chicago for the cost of 6 daytime trains, and gives you freedom to add any number of sleepers you want to the night trains that the engines will bear. 36 Coaches, 6 BCs 6 Dinettes. Shouldn't Pennsylvania, New York, Ohio and Indiana be able to bear the capital cost amortized over 20 years? As for trackage charges, even if CSX and NS charge an enhanced trackage charge it seems to me that a strong case could be made for such a service expansion. It would be way more useful for more people than a few more overnight trains.

 

[Nathanael]

Trogdor has the best solution, at least in my opinion. NARP is generally known to be a bunch of nostalgic foamers. Their support of LD trains is nice, but misguided. The LD train is just a land cruise/vacation train. It may provide for some transportation needs in certain areas, but for the most part it is just an anachronism. Particularly for a train such as the Lake Shore. No business man would use this train to get between NY and Chicago unless he was a secret rail fan. No one is going to voluntarily put them selves in a cramped little room with a public shower down the hall in a closet and get jostled all night long when they can fly to their destination in three hours and stay in a nice hotel with a restaurant and bar and a real bath and shower.

You're basically wrong.
You're right about NY-Chicago, but that's an irrelevance.

Look at Buffalo-Chicago. Businessmen actually use the LSL for this. Routinely. (I know Buffalo doesn't have a lot of businesses left, but still.) Buffalo-Chicago flight? Eats up an entire day, and you still have to get from O'Hare to the Loop and back. Alternatively, you're up all night on the red-eye, if there is one.

Overnight on the LSL? No wasted time.

I'm glad that the report was highlighted with a photo of a person who takes Syracuse-Chicago for business trips. I think using the LSL as their example train was smart; its potential is larger and clearer than, say, the Cardinal or the Sunset Limited.

Something that could probably help out the long distance trains and food service is to1. Have the diners open longer

2. Update Arrow to allow pre-purchase of meals or pre-reservation of meal times taking advantage of 1.

Now, this, this is a BRILLIANT idea. Smart, Paulus. Is there a way we can get this idea through to Amtrak? The diners are a loss leader; this would *both* utilize them more heavily *and* make them operate more quickly *and* make them more attractive and convenient -- a win-win-win result.

Every additional frequency on the Lake Shore Limited would require three additional sets of equipment.

So run the trains faster, and then it won't. Improvements from Chicago to Cleveland -- a valid corridor route on its own -- could allow for running each of the Capitol Limited and the Lake Shore Limited with only two trainsets, eventually. Obviously the more frequencies per day, the more efficiencies of scale you get, as someone else pointed out.

For me, everything keeps coming back to that. Run the trains faster. Make them longer. Then run more. Do so in the areas where long-distance and corridor routes are the same, and it benefits both of them.

This is actually where I have previously disagreed with NARP; they, for a long time, put a much larger emphasis on new routes than on expanding service on existing routes.

This report appears to signal a change in views, putting a higher priority on creating faster, higher-frequency-service corridors, while still emphasizing network connectivity. I like it.

If I were trying just to get the New York - Chicago market straightened out, I'd probably run the following:

I like your proposal, jis. In order to get decent results, the proposed exclusive pasenger tracks from Chicago Union through Gary are needed, because that has been wrecking on-time performance. There's also various stuff needed on the route in New York. NS would probably be OK with it once the passenger trains were cleared off the critical section near Chicago; the tracks from there to Cleveland have a certain amount of redundancy. CSX, which has been underinvesting in the Empire Corridor, probably would not. Ohio and Indiana are unlikely to put in any money, but thankfully the sections which need the most trackwork are in Illinois and New York...

 

[Anderson]

Some thoughts:
(1) I like the idea of slipping a train or two around to cover TOL-DET. Yeah, I know the track conditions there...but I still think it would be a good idea to offer some EB coverage from the Detroit region that doesn't force you through CHI first.

(2) I think a second train on the LSL route would be good, with the "counterpart" train leaving NYP later (say, 1830-1900) and getting into CHI accordingly, and leaving CHI earlier (say, 1600 or so) and getting into NYP by noon. To be fair, what I might end up doing is running one of the pair via Detroit (I'm thinking the current EB LSL plus the late-out-of-NYP train), eating the time penalty, and marketing that pair for the NYP/ALB-CLE/TOL/DET markets.

(3) I'd do the Cap-Pennsylvanian through cars as well, with an eye towards restoring the Broadway Limited (even if under a different name) as time goes by.

(4) As I noted with the remarks on the Cap elsewhere, I'd aim for 2x daily coverage on this "complex" as well. Really, the only place without workable service at that stage would be Erie, PA. Buffalo could probably get a set-out sleeper on one train, and I think that would probably sell (not to mention that logistically, it would just pop on/off behind any BOS equipment). Pittsburgh is in a similar boat, but you've likely got at least one train in each of the three directions (to CHI, WAS, and NYP) that is at a decent hour that a set-out wouldn't be needed.

-It should be noted that you'd basically have one train on each line (the early-west and late-east ones) aimed at connecting with the western LD network and another (the late-west and early-east) aimed at connecting with either corridor trains or just serving the NEC-CHI and the major intermediate markets.

(5) After this, I'd focus on daylight service where demand indicates a need. It's possible that a BUF-CHI daylight train would work well for covering CLE/TOL-CHI.

(6) Just as a note, you'd need one additional set on the Cap (since you should be able to get a pair of same-day turns in the equation). You would probably need either two or three sets for the LSL, but that's a slightly more complex matter of timing (especially with the rush hour stuff).

 

[Paulus]

As usual, no consideration for the fact that the long distance routes lock up an absurd amount of equipment which could be better used for more frequent corridor service. No discussion of the costs of providing sleeper service, merely of the revenue.

As a NARP White Paper from 2008 showed, sleepers cover their incremental costs. So much so that in fact, at that time the subsidy per passenger mile for all on the train was $0.1870. If one removed sleepers and their revenue, then coach passengers would have been subsidized at a rate of $0.1888 per passenger mile.

Seeing as how sleeper prices have risen dramatically since then, far faster than inflation would have affected the costs, I can only assume that they saw no reason to rehash those numbers and that things have actually improved since then with the excess sleeper revenue covering still more of the costs to move the coach passengers.

On the other hand, hasn't Amtrak had new union contracts since then, with above rate of inflation cost increases (though to be fair, it had been several years without one)?

Sleeper revenue for FY08 was $151,922,419; revenue in FY12 was $181,487,996, annual growth rate of 4.87%. Meanwhile long distance costs increased from $1,020.8 million to $1,158.1 million, an annual growth rate of 3.38%.

A diesel-powered passenger train can move nearly two and one-half times more people per gallon than a typical automobile.

Amtrak really doesn't do well when it comes to fuel efficiency and long distance trains are probably no more fuel efficient than a typical automobile due to FRA inspired heavy locomotive weight and low seating density. Unfortunately Amtrak doesn't break it down any better than nationally. However, on a national level, Amtrak is only about as green as driving a Prius by oneself.

According the US Department of Energy, on average Amtrak consumes 2,271 BTU's per passenger mile. And while Amtrak might not break it down, it is clear that it has to be either the NEC or the LD's or both that keep Amtrak's numbers down, since commuter trains consume 2,897 BTU's per pax/mile.

The average car consumes 3,447 BTU's per pax/mile. Only the motorcycle beats Amtrak, or any rail transportation for that matter.

It does beat the average car, but not by too much. 50 passenger-miles per gallon equivalent for Amtrak vs 39 for commuters and and 33 for the average car. But, like I said, toss someone in a Prius, and they're just as green as Amtrak. Amtrak does have a per seat mile breakdown between electric and diesel down on page 36. Electric comes out to .146 kWh per seat-mile, diesel comes out to 163 seat-miles per GGE. For comparison, a BR Class 222 Meridian (478 seats) achieves 240 miles per gallon (272 mpGGE) and a Class 373 Eurostar (750 seats) consumes 0.0885 kWh per seat-mile. Neither of these was selected for purposes of best fuel or energy economy.

 

[henryj]

No, I am not wrong. We are railfans here and think differently from 'normal' travelers. No businessman or woman is going to use the LSL to get from NY to Chi period. Those days are gone forever. NARP in their glorious support for LD trains is stuck in the 1950's and dreams of recreating those days. They should spend some time in Europe. Fast, frequent corridor trains are the future. Europe still has a few overnight sleeper trains, but very few, and they lose all kinds of money. You can sit in any big train station in Europe and watch ICE's and regional trains coming through all day long, sometime several at one time. But you will rarely see a LD train or a sleeper train. Most routes boast hourly or every two hours service. Trogdor's idea is where it's at. One overnight sleeper train and many fast corridor train on parts of the route that can support it. Like I said, Amtrak's LD routes are pretty much set. I only see two missing links, the Chicago to Florida through Atlanta and Texas to Colorado and perhaps Portland. These trains could be made more profitable by cutting back on coach attendants and upping coach fares at least above Greyhound. Sleeper's pretty much pay for themselves. It will be interesting to see how Amtrak eventually handles the LD train thing.

Trogdor has the best solution, at least in my opinion. NARP is generally known to be a bunch of nostalgic foamers. Their support of LD trains is nice, but misguided. The LD train is just a land cruise/vacation train. It may provide for some transportation needs in certain areas, but for the most part it is just an anachronism. Particularly for a train such as the Lake Shore. No business man would use this train to get between NY and Chicago unless he was a secret rail fan. No one is going to voluntarily put them selves in a cramped little room with a public shower down the hall in a closet and get jostled all night long when they can fly to their destination in three hours and stay in a nice hotel with a restaurant and bar and a real bath and shower.

You're basically wrong.