Viewliner II - Part 1 - Initial Production and Delivery

[neroden]

As I read through this thread (and related ones about equipment), I find myself wondering about how much equipment Amtrak needs to haul baggage. Although the current crop of cars is aged, is it worth upgrading them instead of rolling stock that will accommodate more passengers?

I'm not giving an opinion here, I'm just wondering.

There are 64 baggage cars, the oldest dating from the late 1940s, and the newest from (IIRC) 1959. 35 of them were converted from coaches and were never really laid out properly for baggage; many of these also have structural damage because the undercarriage wasn't designed for baggage. The other 29, which were designed as baggage cars, are all shorter than standard Amtrak cars, and they are mostly not really well arranged for modern baggage either. (Any interior shelving disappeared long ago). The interior floors on all of them are in bad shape. The doors on practically *all* of them are damaged and are not reliable; dust and snow gets in. The underfloor gear is ancient and worn out and well overdue for total replacement. They have top speed limits which prevent them from running at full speed on the NEC and the Empire Corridor, so they are the limiting factor on top train speed on all the single-level trains. They are out of service so often that Amtrak has substituted cafe cars on occasion.

It is not worth spending a penny on any of the Heritage baggage cars. They will probably be cut up for scrap.

 

[WoodyinNYC]

Thanks for your thoughtful response.

The 25 Viewliner II sleepers in the current CAF order will be added to the 50 Viewliners now in the fleet, for a 50% increase in roomettes.

I had no idea that there were more sleepers on order. I'm an ignoramus.

Plus the 10 bag-dorms (letting crew move out of the sleepers), will open up

passenger roomettes for 10 "half of" sleepers, making a 60%

increase in roomettes from now

Hmm. That too. Seems like we're talking a serious amount of increased passenger room.

Well, we're talking about a $300 million order for a lousy 130 cars,

with 65 of them full baggage cars. Then 25 diners, which will replace

Heritage diners one for one. Remember, "Heritage" means inherited

from the trains that ran pre-Amtrak, pre-1971. so the diners are terribly

ld too. But worse, the 25 diners came from like 25 different trains, well,

maybe a slight exaggeration, but there are no economies of scale buying

spare parts or training maintenance crews, because they are mostly

one-of-a-kind cars. Then 10 baggage crew dorms. Half the car is for

baggage, half for the crew. The bag dorms "create" space to sell,

because if there are no dorm cars, the crew sleeps in roomettes

in sleepers.

Finally comes the order for 25 sleepers, which will increase inventory

considerably for the premium service, which often sells out holidays

and weekends. But each new car will have 11 roomettes, 2 bedrooms,

and 1 ADA bedroom, so a maximum of about 27 passengers. If they

sell every room on the sleepers, that's 25 X 27 = 675 additional riders

per day, or 246,000 a year. Not much against Amtrak's ridership of

31 million.

The current order does NOT include any additional coaches. Today's

single-level Amfleets typically carry 59 or 60 passengers. Filling every

seat of a new coach would add about 1,500 riders a day, or almost

550,000 a year. Adding even 25 coaches, while replacing the existing

fleet, could add 13 million seats. Now that would be a capacity increase.

The CAF order for Viewliners does include an option for another 70 cars --

diners, sleepers, bag dorms, and baggage cars -- for 200 total.

Almost everyone on these boards hopes to see the option order placed,

but timing is everything.

It would much cheaper to buy additional cars now. Buying in batches

of 5 diners or 10 sleepers after the assembly line shuts down would

mean costly, almost handcrafted cars, so better to do it while CAF is

still building the 130 or 200 cars. Meanwhile a large order of 600 or so

coaches over five or six years, as is suggested in the Fleet Renewal Plan,

would bring economies of scale to the manufacturing process resulting

in high-volume lower prices.

But of course, Amtrak is being squeezed by Congress, and must very

carefully plan and prioritize its spending. The next big thing will be

the order for 28 Acela II train sets. Once those bids come in, and

once Congress passes a new transportation appropriation, Amtrak

will have a better idea of its cash flow, total borrowing, etc before

it can commit to ordering any more cars for its existing trains.

 

[neroden]

The Heritage baggage and diner cars are rated for 110 mph speeds, so they don't present a speed constraint except on the NEC.

I dug up some information, and I find that the 80 ft. baggage cars (converted coaches) are good for 110 mph, but have special speed restrictions at certain locations (probably curves or grades) --- while nobody outside Amtrak seems to know what the speed restrictions are for the 60 ft. baggage cars.

It appears that *individual cars may have their own speed limits*, lower than 110.

http://cs.trains.com/trn/f/743/t/218616.aspx

These things are the opposite of a uniform fleet; practically every Heritage baggage car is a unique snowflake. I can't imagine how much that costs in maintenance.

 

[neroden]

Finally comes the order for 25 sleepers, which will increase inventory

considerably for the premium service, which often sells out holidays

and weekends.

And random Wednesdays in November! The LSL sells out of rooms *all the time*, except in the weak months of January and February. The Cardinal sells out even more often.

But each new car will have 11 roomettes, 2 bedrooms,

and 1 ADA bedroom, so a maximum of about 27 passengers. So if they

sell every room on the sleepers, that's 25 X 27 = 675 additional riders

per day, or 246,000 a year. Not much against Amtrak's ridership of

31 million.

Yeah, but it's a much bigger deal in terms of revenue. At average yields of $700 per bedroom and $350 per roomette, this is (700 * 3 + 350 * 11) = $5950/day or $2.17 million/year.
These are realistic yields for the Lake Shore Limited (except probably in January and February).

The current order does NOT include any additional coaches. Today's

single-level Amfleets typically carry 59 or 60 passengers. Filling every seat

would add about 1,500 riders a day, or almost 550,000 a year.

Yeah, but the coaches aren't filling every seat yet (well, except on the Cardinal).
There are still tickets available. At plausible loads and yields, an added coach is not as profitable as an added sleeper. With a $101 yield (typical for this month's prices), a completely full coach (which is unlikely) realizes $2.17 million in revenue.

Just for fun, I looked up today's (Friday's) NY-Chicago availability. The Cardinal is completely sold out. The LSL has one accessible bedroom available for $1102, and that's it. Coach seats remain available for $101.

Checking Amsnag, rooms on the LSL remain at near-sellout conditions for most of November. This is despite the OTP disaster recently. The lowest bucket in the next 30 days for a room on the LSL or Cardinal is $213, which occurs on exactly two days. And November isn't typically a peak month, Thanksgiving aside.

(The Capitol Limited, and Superliner trains in general, do not have these routine sellout conditions.)

This is why Amtrak needs those new Viewliner sleeping cars ASAP. They're leaving money on the table. Even assuming pessimistic running cost guesses, each sleeping car will pay for itself in 3 years. They'll last for decades. The commercial case for them is extremely strong; given that the Viewliner trains are being operated at all, the case for the new Viewliner sleeping cars is nearly as strong as the case for the Acela IIs.

(Just to preemptively knock down one common but stupid argument: the sleeping cars will look worse financially if you attribute all the dining car costs to them; but that's grossly inappropriate, given that about half the dining car patronage on the LSL is coach passengers.)

----

The dining cars and baggage cars, unfortunately, are just a "we have to replace these before they fall apart" situation, and will not realize money immediately, apart from the maintenance savings. Hopefully that will be substantial.

 

[sitzplatz17]

Just spotted two of the baggage cars outside Union Station in DC. I assume those are the two that are out and about for testing?

 

[OBS]

They have been there for about 2 weeks, not sure if there for training or what...

 

[Viewliner]

61001 was in Philly early this afternoon, coupled to an Amfleet I coach.

 

[Acela150]

Cars are on the pick up track in Elmira.

 

[jis]

One of them is reported to be 61006.

Sent from my iPhone using Amtrak Forum

 

[Steve4031]

I know this was written some where but I could not find it. Im planning to take gf to nyc on 48/49 at the end of june. It would be awesome if we rode new viewliners sleepers. However its not life and death. I also know it's impossible to predict with certainty. But an educated guess would be appreciated.

 

[WoodyinNYC]

I know this was written some where but I could not find it. Im planning to take gf to nyc on 48/49 at the end of june. It would be awesome if we rode new viewliners sleepers. However its not life and death. I also know it's impossible to predict with certainty. But an educated guess would be appreciated.

Sorry. The first 70 will be baggage cars, 10 bag-dorms, then 25 diners.

No sleepers until 2016, or if delays continue, later, LOL.

 

[Steve4031]

Ok. Thank you. This was helpful.

 

[Dollars_and_sense]

Well, we're talking about a $300 million order for a lousy 130 cars,

with 65 of them full baggage cars.

Yeah, but it's a much bigger deal in terms of revenue. At average yields of $700 per bedroom and $350 per roomette, this is (700 * 3 + 350 * 11) = $5950/day or $2.17 million/year.

So, basically, Amtrak is spending $300M to generate $2.17M/year in additional revenue. So, these new cars will need to provide service for way over 100 years to just break even.

Yea, I realize that Amtrak needed new cars that don't directly produce any revenue, but I doubt the "bean counters" factor such into their accounting books.

 

[Ziv]

Dollars and sense, I may be wrong but the guesstimate was that it was possible for EACH sleeper to generate $2.17Mn per year, so that would be more than $50Mn a year at full occupancy, plus the $3-$4Mn the bag dorms will add to the bottom line. And the fact is that eliminating the old cars add value in a lot of different ways. Obviously the sleepers won't be at 100% occupancy for the year, but they will be close.

This equipment purchase will pay for itself in a very short time, especially when you consider the fact that these cars will be in the fleet for 30-40 years.

Well, we're talking about a $300 million order for a lousy 130 cars,
with 65 of them full baggage cars.

Yeah, but it's a much bigger deal in terms of revenue. At average yields of $700 per bedroom and $350 per roomette, this is (700 * 3 + 350 * 11) = $5950/day or $2.17 million/year.

So, basically, Amtrak is spending $300M to generate $2.17M/year in additional revenue. So, these new cars will need to provide service for way over 100 years to just break even.

Yea, I realize that Amtrak needed new cars that don't directly produce any revenue, but I doubt the "bean counters" factor such into their accounting books.

 

[Ryan]

Well, we're talking about a $300 million order for a lousy 130 cars,

with 65 of them full baggage cars.

Yeah, but it's a much bigger deal in terms of revenue. At average yields of $700 per bedroom and $350 per roomette, this is (700 * 3 + 350 * 11) = $5950/day or $2.17 million/year.

So, basically, Amtrak is spending $300M to generate $2.17M/year in additional revenue. So, these new cars will need to provide service for way over 100 years to just break even.

Not really, since they're buying more than one car.

 

[jis]

Yeah, but it's a much bigger deal in terms of revenue. At average yields of $700 per bedroom and $350 per roomette, this is (700 * 3 + 350 * 11) = $5950/day or $2.17 million/year.

So, basically, Amtrak is spending $300M to generate $2.17M/year in additional revenue. So, these new cars will need to provide service for way over 100 years to just break even.

That $2.7 million is per car per year. If you take the trouble to do the multiplication by 25, you land up with something north of $50 million per year, which changes your bottom line to something under 6 years. Not bad at all.
Getting the arithmetic straight helps arrive at the right conclusions.

 

[afigg]

I know this was written some where but I could not find it. Im planning to take gf to nyc on 48/49 at the end of june. It would be awesome if we rode new viewliners sleepers. However its not life and death. I also know it's impossible to predict with certainty. But an educated guess would be appreciated.

Sorry. The first 70 will be baggage cars, 10 bag-dorms, then 25 diners.

No sleepers until 2016, or if delays continue, later, LOL.

I don't recall reading that the change in production was going to be all baggage cars first, then bag-dorms, then diners, etc. Only that the baggage cars were going to move to production while the production of the other 3 car types was to start later. In the October, 2013 press briefing, after the 2 sets of test cars, CAF was going to build the Viewliners on 2 production lines, one for baggage cars, the other sequentially to produce diners, bag-dorms, sleepers in that order. Since Amtrak badly needs the new diner cars as much as the baggage cars, the diner cars should see production before the bag-dorms and sleepers.

Getting back to the original question, the odds of being on a new sleeper car on the LSL next June are very small. The odds of the LSL having a Viewliner diner car, either the 8400 or a new one, are better. But given the years of delay, the odds are may be poor even on encountering a new diner car.

 

[afigg]

That $2.7 million is per car per year. If you take the trouble to do the multiplication by 25, you land up with something north of $50 million per year, which changes your bottom line to something under 6 years. Not bad at all.

Getting the arithmetic straight helps arrive at the right conclusions.

Neroden's number was $2.17 million per sleeper car per year, but that is still higher than what Amtrak gets.
In FY2013, the total sleeper revenue from the LSL, Silver Star, Silver Meteor, Crescent, Cardinal was $37.6 million. That was spread over 39 sleeper cars in regular service rotation. Of course, without crew-dorm cars, a lot of roomettes are taken up by OBS and lost to revenue sales. Still, less than $1 million per sleeper car.

From the OIG report, the plan is to deploy a total of 60 sleepers and 5 bag-dorm cars in service rotation. So some back of the envelope get in the ballpark additional revenue projection.

The new sleepers will have 11 roomettes instead of 12, so the 25 new sleepers won't quite equal 25 Vw Is in capacity. Amtrak can make up some of the difference with 2 attendants splitting 3 sleeper cars between them. The OBS staff size won't increase, so going from 2 sleepers to 3 for a train not getting a bag-dorm is more than a 50% increase in roomettes.

The 5 bag-dorms will provide 45 roomettes in capacity for the LSL and Cardinal. For the LSL, they will free up 8-9 roomettes currently used by OBS. Until the Cardinal gets a full service diner, extra bag-dorm roomettes are likely to be sold for revenue service. So the 45 roomettes are equal to 4 Viewliner II sleepers in roomettes, but not in higher revenue bedrooms. Hand waving here, I'm going to say 5 bag-dorm cars roughly equals 3 current sleeper cars in revenue capacity.

So, ball-park the expanded fleet as roughly 63 sleeper cars in capacity compared to the baseline of 39 sleepers for about a 62% increase in revenue capacity. That may be low because I have not compensated for the capacity expansion with a fixed sized OBS. I expect someone else has calculated that. I'm ignoring the effect of bucket prices here as Amtrak may sell more rooms at lower buckets than high buckets with increased capacity but this is back of the envelope computation here.

So 62% of FY13 $37.6 million is $23.3 million for the additional sleeper revenue that Amtrak could/should get from the Viewliner IIs. This does not include increased operating costs, the expected reduction in maintenance costs, nor the benefits of improved equipment reliability (one hopes).

 

[Just-Thinking-51]

Per Passenger Train Journal the sleeping cars are going to name for "rivers", and the viewliners 1 sleeping cars will be renamed to "rivers".

Dinning cars will get Eastern state capital names.

whiteriverproductions.com

 

[jis]

And the name Trenton will be reserved for future use, since it is the only eastern capital in alphabetical order that does not have a car assigned to it at present.

 

[neroden]

So, basically, Amtrak is...

No. 2.17M/year PER SLEEPING CAR.
Multiply by 25 cars to get about $100M/year in revenue. (Substantially less in profit after operations costs, of course.)

Call me optimistic, but I think still think this estimate will be roughly correct by the time the new Viewliners come into operation, assuming OTP recovers.

The current sleeping cars have a huge chunk of space taken up by staff (over half of a car on each train), so they're performing much worse than an added sleeping car should. I have no idea how the OBS staff wangled solo compartments in their union contact; this is an example of Amtrak having a much cushier union contract than the pre-Amtrak railroads.

In FY2013, the total sleeper revenue from the LSL, Silver Star, Silver Meteor, Crescent, Cardinal was $37.6 million. That was spread over 39 sleeper cars in regular service rotation. Of course, without crew-dorm cars, a lot of roomettes are taken up by OBS and lost to revenue sales. Still, less than $1 million per sleeper car.

Roughly half a car is taken up by staff on each train; more on some, less on others. Something like a quarter of the sleeper space is used on the staff, who have really excellent union contracts. So, um, spread that revenue over roughly *30* cars, and you get around $1.3 million per car. The Crescent has poor utilization south of Atlanta; it isn't profitable to add cars there; so the revenue number for cars assigned elsewhere is better than that.
FY2013 was a very bad year for OTP, though (and so was FY2014). I've been trying to use FY2012 numbers.

The average car in the CAF order costs $2.3 million/car. Sleepers probably cost a bit more -- call it $2.6 million to be conservative. Suppose you assume it costs $800K/year to operate & maintain (which is a wild-ass-guess and probably MUCH MUCH too high -- I think it's probably more like $300K/year). Even if you go with a $1.3 million/year estimate for revenue, you're STILL breaking even in 6 years. If you use more optimistic estimates for revenue, expenses, and the capital cost of the cars, each sleeping car might pay for itself in 2 years!

(Of course, the revenue won't materialize if the "freight" railroads continue to get away with their lawbreaking practices of sabotaging Amtrak. But that's another issue.)

It's worth recalling that Amtrak originally intended to order 100 Viewliner sleeping cars, and only ordered 50 due to cash deprivation. The additional 25 (total of 75) don't even get Amtrak back to the level Amtrak originally planned to have in 1990, when demand was much lower than today.

It should be *expected* that additional Viewliner sleeping cars will have incredibly fast payback times: I really would bet on 2 years rather than 6 years, but call it 6 years to be conservative.

The expenditures on baggage cars and dining cars don't pay for themselves, but they needed to be done anyway.

It is, indeed, very hard to calculate the added revenue from having checked baggage service vs. not having it. (And with the Heritage baggage cars collapsing, "not having checked baggage" is the alternative to the new baggage cars.) So it's not straightforward to figure out how long it'll take the baggage cars to pay for themselves. Amtrak seems to think that the value of checked baggage service (in terms of attracting passengers) is significant and outweighs the costs -- Amtrak has recently added checked baggage service to a number of stations and believes that this was a positive to the bottom line. But I can't put any numbers on it.
For the dining cars it's somewhat easier to compute, as we have an example of a previous period when Amtrak trashed dining car service at the behest of Congress and promptly lost 13% of *all* ticket revenue on the affected trains. I have not sat down to do an estimate based on this. Based on 2014 revenue from the trains running Heritage dining cars, this would be about $18 million/year. (And these use about 18 dining cars, including shop count.) Unfortunately the dining cars are expensive to operate -- I have no idea how expensive, really -- in a worst-case scenario, it might actually be higher than $1 million/year, in which case the dining cars simply wouldn't pay for themselves. At a more plausible $700K/year in running/maintenance costs, they do pay for themselves, but it takes around 8 years. This is a much worse result than for the sleeping cars.

It seems to me that the payback on the sleepers is much, much clearer than for the baggage or dining cars. (Provided you can assign them to trains where the latent demand exists, of course -- not to the Crescent south of Atlanta.)

Of course, this may be another reason why Amtrak management is pushing the baggage cars and dining cars to the front of the queue. Intangible railroad know-how means management knows they need to replace these, but it's much harder to put solid numbers on the benefits, so when penny-pinching micromanagers show up, the baggage car and dining car orders are harder to defend. (Still, every year without the extra sleeping cars is revenue foregone, and so this political decision is costing Amtrak money.)

A sleeping car order can, if necessary, be defended with pretty-easy-to-generate numbers, even in a hostile environment. Paulus guessestimated $200 million in increased revenue from Acela II; at 28 trainsets and $40 million/trainset, the capital costs are $1.12 billion; that's a 6-year payback time.

Right now, Viewliner sleeping cars are probably a better investment than Acela II. (This is due to the extreme shortage of Viewliner sleeping cars, of course; at some point, after ordering enough Amtrak will have the "right number" and it will become unwise to order more.)

 

[Paulus]

For the dining cars it's somewhat easier to compute, as we have an example of a previous period when Amtrak trashed dining car service at the behest of Congress and promptly lost 13% of *all* ticket revenue on the affected trains.

13.6% of ridership, though I'd love to see what routes they were and the ridership figures. Would also be good to see the cost figures: A ridership drop may still well have been a financially beneficial move given.

Unfortunately the dining cars are expensive to operate -- I have no idea how expensive, really -- in a worst-case scenario, it might actually be higher than $1 million/year, in which case the dining cars simply wouldn't pay for themselves. At a more plausible $700K/year in running/maintenance costs, they do pay for themselves, but it takes around 8 years.

Are you including labor costs in running or just fuel, switching, and maintenance?

The average car in the CAF order costs $2.3 million/car. Sleepers probably cost a bit more -- call it $2.6 million to be conservative. Suppose you assume it costs $800K/year to operate & maintain (which is a wild-ass-guess and probably MUCH MUCH too high -- I think it's probably more like $300K/year). Even if you go with a $1.3 million/year estimate for revenue, you're STILL breaking even in 6 years. If you use more optimistic estimates for revenue, expenses, and the capital cost of the cars, each sleeping car might pay for itself in 2 years!

Couple things

1) Annual revenue on Viewliner sleepers in use was around the $950K range on the dailies. A 37% increase in revenue for an unstaffed Viewliner, while it will have more rooms available, seems to me optimistic, especially since such a major increase in the number of rooms should depress the revenue per passenger.

2) Not all sleeper revenue is incremental revenue. While I figure that linens and whatnot are going to be a fairly marginal cost, you need to account for the F&B allocation of the sleeper ticket which will drop down the revenue somewhat

3) You'll have a few that don't make any money all year due to acting as spares or some maintenance hold (not any individual one, but a certain percentage at any given time equates to the equivalent).

Interestingly, while doing the numbers, it looks like coaches, on a per-car basis, brought in significantly more money per day and per year than do the sleepers, even before accounting for any sleeper expenses. Belated edit: Doesn't account for amount of space taken up by crew.

 

[neroden]

I reran the numbers again using a different set of principles, and I get a more pessimistic 600K/year in profit for each new sleeper. It's *still* an extremely good return on investment, as good as the Acela IIs.

Interestingly, while doing the numbers, it looks like coaches, on a per-car basis, brought in significantly more money per day and per year than do the sleepers, even before accounting for any sleeper expenses. Belated edit: Doesn't account for amount of space taken up by crew.

That's almost certainly because you did the numbers wrong.

Roughly *half a sleeper on each train* is wasted for crew space and shouldn't be considered available sleeper space. Get your denominator right and try again.

There are a ridiculous number of rooms taken up by staff, including the coach attendants and cafe attendants, but perhaps most expensively, the dining car staff. (Hence my suggestion not so long ago to find some way to get the dining car staff off the train at night, by only hooking the dining car up in the daytime.)

I've seen historic "crew cars", and crew got bunk beds (in space terms, pretty much equivalent to dual occupancy in Viewliner roomettes). I have no idea how they wangled single occupancy in their current union contracts, but it's a very cushy contract clause which is costing Amtrak a lot of money. Maybe it had something to do with the switch to mixed-gender crews and sex panics in the 1950s/1960s?

Anyway, I guesstimate the foregone revenue as about $80,000/year/roomette on the Lake Shore Limited. So add $80K per year to the salaries and benefits of each OBS employee to get a true costs estimate. (Would have been $40K with double occupancy).

This means that it's really costing something like $400K/year just to haul a five-person dining car staff around overnight! This is very substantial. It's probably not accounted for in the F&B numbers, either. Since the dining car isn't serving overnight, if the staff could be dropped at a hotel in the evening and picked up in the morning, would it be cheaper? Yes, it would, and it would be a lot cheaper. A hotel room in most cities (such as Buffalo, NY) costs far less than a roomette.

---

Every time I look at this, the dining cars look more and more expensive under current methods of operation. I believe dining car costs are being (mis)attributed to the high-value sleeping cars in order to keep the dining cars around. (The coach attendants also get free roomettes, of course, which is also not accounted to the coaches.) And yet I don't think there's anything essential -- in fact, I don't think there's anything customer-facing -- about the aspects of the operation which are so extremely expensive.

The cafe cars are no substitute, and I think it's really valuable to be able to get fresh-cooked eggs or oatmeal in the morning. But why does this require hauling 5 employees by single-occupancy roomette overnight? It shouldn't! They aren't cooking or serving overnight!

This may partly be *another* issue which is due to the trains not running on time, and it may partly be due to running too-few trains on the routes. If the train runs on time, you can pick up your dining car crew in the early morning and have them serve breakfast; if there are many trains per day on the route, you can then send 'em back home on a later train. You want sleeper & coach attendants on board at night because they need to actually *work* at night, but the same is not true of the dining car staff.

But when the train *doesn't run on time*, you need to keep the staff on board because you have no idea where the train will be at any given time. And when there isn't another train, you can't return the staff the same day, either. And this has huge overhead costs.

When the Empire Builder was unable to keep schedule, a number of us discussed the idea of simply keeping three engineers and three conductors on board in rotating shifts -- which would eliminate crew time-out problems. However (apart from nonsensical FRA rules) it would also require providing at least four sleeper berths. The situation with the OBS crew is like that, only worse.

The inability to keep time creates problems beyond the obvious ones: it prevents more efficient methods of operation from working.

 

[Paulus]

Interestingly, while doing the numbers, it looks like coaches, on a per-car basis, brought in significantly more money per day and per year than do the sleepers, even before accounting for any sleeper expenses. Belated edit: Doesn't account for amount of space taken up by crew.

That's almost certainly because you did the numbers wrong.
Roughly *half a sleeper on each train* is wasted for crew space and shouldn't be considered available sleeper space. Get your denominator right and try again.

'Tis the same, except for the Cardinal unless staffing is lower and only a quarter of Cardinal sleeper is unavailable, then coach again has higher revenue per car. LSL comes close, it's only a $35K difference with the lower capacity, but it also the worst performing coaches (annually) behind the thrice weeklies and the Capitol Limited (it also has the best performing sleepers).

The Superliners actually do worse, in terms of revenue per car, than do the Amfleet/Viewliner trains.

Yes, it would, and it would be a lot cheaper. A hotel room in most cities (such as Buffalo, NY) costs far less than a roomette.

Costs even less with a steady contract for it (or even less if you do what Amtrak ought to do and build hotels at their own stations for a degree of vertical integration).

 

[neroden]

I'm not remotely surprised that the Superliners do worse than the single-level trains. All the ticket prices are lower, and the coach-sleeper differential is also lower. And frankly most of them are not filling up most of the year (coach or sleeper).

FWIW, even if you find lower revenue per car (and I'm still extremely suspicious you've got your calculations wrong), much of that is going to be an artifact of short-haul passengers being charged more. For some reason, Amtrak charges significantly less for NYP-CHI (in the same class of service) than it does for NYP-SYR + SYR-CHI. I've never figured out why, and this should probably be changed. (The same is true on a smaller scale, for NYP-ALB / ALB-SYR, for instance. And it's even more extreme on the western trains.) I don't think this actually reflects people's price-sensitivities, so it seems really weird, and I don't know why Amtrak continues to do it -- surely the money of a short-haul passenger is as good as the money of a long-haul passenger? Why give such large discounts for the long-haul? But anyway, the short-haul passengers are disproportionately in coach; and this doesn't mean anything when you're deciding whether to add more cars, *unless* there's a large number of unserved short-haul passengers.

It is still almost certainly more valuable to add a sleeping car to a train like the LSL than to add a coach car, since most of the time the coaches aren't full (short-haul demand is satisfied), and the sleepers are full (long-haul demand isn't satisfied), even though sleeper prices are going up faster than coach prices. If the coaches were consistently filling up and the sleepers weren't, that would of course imply that adding a coach was more valuable.

I may, of course, be biased by looking at the LSL (my local train),... where not only does it have the best performing sleepers and the worst-performing coaches, but in fact the amount of room space occupied by staff is *more* than half of a car (about 60% actually, more when extra coaches are added as happens sometimes). Furthermore, the averages conceal the fact that the NY sleepers do much better than the Boston sleeper (which is usually the last to fill up).

I haven't bothered to look carefully into the numbers on the Florida trains, or their rates of selling out. Maybe their coaches are filling up faster than on the LSL, and their sleepers less quickly.

The key thing is to match the consist to the market. If the passengers you're turning away at the ticket office due to sellouts are travelling "sleeper" numbers of hours, you put sleepers on the train; if they are travelling "coach" numbers of hours, you put coaches on the train. Of course this calls for somewhat more rational relationships between fares and distances.

As a result of an artificially small Viewliner order back in 1990 which did not cover normal replacement, combined with the removal of Heritage dorm cars without replacement, I believe that the market which would be served by additional Viewliner sleeping cars is very strong. People have, off the top of their heads, named several trains which could have sleepers attached to them which would probably attract passengers who are not currently travelling on those trains. (The Pennsy-CL through cars, and #66/#67 among others.)

The same is *not* true for Superliner sleeping cars, where there seem to be enough to attach them to all trains where they might be appropriate, and arguably an excess (Amtrak has scraped up extras on short notice repeatedly).

----

I had one funny thought about fares recently. In some parts of the country, people seem to think the sleeper fares are very high, perhaps because they are comparing them to local hotels. Even though the sleeper fares are much higher on the routes to NYC than they are elsewhere, they seem quite reasonable if you're comparing them to *New York City* hotel prices. (They don't seem that high by Chicago prices, either.) Different expectations regarding hotel prices may partly account for the variation in how much Amtrak can charge for sleepers on different routes. This would imply that sleeping cars to cheap-hotel areas would do poorly.