Article: "What's Wrong With Amtrak"?

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.
Joined
Jul 25, 2015
Messages
2,060
Location
Philadelphia Area
https://ntbraymer.wordpress.com/2017/04/14/whats-wrong-with-amtrak-2/

Andrew Selden in the article calls the LD system its most successful branch and seems to be critical of the NEC and the amount of money invested in it. Admittedly I have ridden LD trains far more than NEC trains (and I live along the NEC) because to me a I can drive a 2-3 hour trip but not a longer trip. But I also recognize the value of the NEC to Amtrak. Selden seems to pit the NEC vs. the LD trains and we all know you hate it when I pit trains vs. each other.

I responded to the article itself but it says my comment is awaiting moderation which means you won't be able to see it until it is (or won't if it's rejected).
 
Selden uses an interesting way of evaluating rail transportation performance in his letter to secretary Elaine Chao. It is hoped that the points that he made were understood and that adequate funding will be there for Amtrak to operate the system efficiently. The argument for passenger rail really must be made to the House and Senate appropriation committees who will pass the budget, reconcile differences and vote on a final version. Secretary Chao, and the White house will provide inputs but the budget is really in congress hands.Congress can chose to accept or reject these recommendations as they please and the president can sign or veto the budget and send it back for changes. If the president veto's the budget government shuts down. I cannot believe that politicians want this to happen..
 
The author is a lot smarter than me. However, I had some of the same thoughts, but could not put them into words with all the facts and figures. It is stunning that the NEC only generates a load factor of 50%. It seems that long distance and intercity segments are definitely the areas for future investment to help ensure success.

I have always been troubled by Amtrak's lack of investment in long distance equipment. Can you imagine a major airline, like Southwest or American, having an average fleet age of over 30 years? In contrast, they receive new planes almost every month.
 
Last edited by a moderator:
Can you imagine a major airline, like Southwest or American, having an average fleet age of over 30 years? In contrast, they receive new planes almost every month.
You can't compare apples to oranges.

Airplanes are made primarily of aluminum and have pressurize the passenger cabin which adds a stress and aging factor that differs greatly from train passenger cars.

Airlines keep track of how many flight cycles (takeoffs & landings) and flight hours a plane has rather than keeping track of time.

But basically airlines upgrade the equipment when when it is no longer economically viable to keep any particular airplane in service.

Even then, those airplanes often get sold to smaller airlines and get used for many more years before the air frame can no longer be certified as safe to fly.

While a major airline may replace a handful of airplanes monthly, a majority of their airplanes are more than 10 years old with some of them approaching being 20 years old.
 
You can't compare apples to oranges.

Airplanes are made primarily of aluminum and have pressurize the passenger cabin which adds a stress and aging factor that differs greatly from train passenger cars.

Airlines keep track of how many flight cycles (takeoffs & landings) and flight hours a plane has rather than keeping track of time.
Its not that simple for airplanes. Best way for us rails is to remember FRA allows ~1500 hours of operation or 15 years whichever occurs first for steam engines.

Airplanes can have different parts of aircraft be time limited, cycle limited, pressurization cycles. engine time, flight time, etc. Good example is landing fear for some aircraft have a hard 5 years between overhauls no matter how many landings.. What is worse is that some parts in engines have different requirements including calendar, cycles, operating times, etc. location of operations, etc.
 
You can't compare apples to oranges.

Airplanes are made primarily of aluminum and have pressurize the passenger cabin which adds a stress and aging factor that differs greatly from train passenger cars.

Airlines keep track of how many flight cycles (takeoffs & landings) and flight hours a plane has rather than keeping track of time.
Its not that simple for airplanes. Best way for us rails is to remember FRA allows ~1500 hours of operation or 15 years whichever occurs first for steam engines.
Airplanes can have different parts of aircraft be time limited, cycle limited, pressurization cycles. engine time, flight time, etc. Good example is landing fear for some aircraft have a hard 5 years between overhauls no matter how many landings.. What is worse is that some parts in engines have different requirements including calendar, cycles, operating times, etc. location of operations, etc.
I thought it was 1472 service days before a boiler inspection for a steam locomotive
 
You can't compare apples to oranges.

Airplanes are made primarily of aluminum and have pressurize the passenger cabin which adds a stress and aging factor that differs greatly from train passenger cars.

Airlines keep track of how many flight cycles (takeoffs & landings) and flight hours a plane has rather than keeping track of time.
Its not that simple for airplanes.
My point exactly, though I didn't go into as much detail regarding how the airplane/train equipment comparison made was invalid.
The bottom line is that passenger train cars are rightly expected to stay in service longer than commercial airplanes.
 
I don't agree with a few of their metrics.

Passenger miles mean one CHI-SEA passenger is more valuable than eight NYP-WAS passengers. Remember in order for that passenger to travel 2000 miles means the train has to run 2000 miles. If you're going to consider passenger miles, do PM/TM. In terms of revenue, the one EB passenger might pay more in fare than eight NYP-WAS regional passengers (I'd think eight Acela passengers would combine for more revenue than one EB passenger) but he/she costs much more.to transport.

For load factor, the LD trains have higher load factors more because of less supply than more demand and Amtrak doesn't have as much control over more LD service. Forget a train from Philly to Chicago, try getting NS to run a second train from Philly to Pittsburgh. Then there's the CAF order that was delayed which hurt Amtrak's LD inventory.

Finally, market share? A small part of a large market or a large part of a small market?

Amtrak invests the most money on the NEC. They own the tracks and can run at higher speeds. Would you rather run 20 faster trains or 20 slower trains? Would you rather run 20 trains where you're not handicapped by freight or 20 trains where you are? Maybe there is too big an imbalance but I can't blame Amtrak to spend money on an investment they control more. You take more care of a house you own vs. an apartment your rent or a hotel room you spend a few nights in.
 
Last edited by a moderator:
Another thing to consider for discussing the NEC is that the Acela and NE Regionals should be considered different. One is for faster service aimed at business travel whose companies are willing to pay higher fares for faster and more convenient service while the Regionals are alternate travel more towards the regular average Joe. I'd be willing to bet that the load factor on Acela's is way higher than the NE Regionals and Acela's load factor compares more favorably to the LD's than the regionals.

I'm guessing Mr. Selden would be against the Avelia Liberty investment: http://media.amtrak.com/wp-content/uploads/2016/08/NextGen-Fact-Sheet_Proof5.pdf.

If they are as advertised, you would increase capacity (perhaps the ability to lower fares and still increase revenue and increase the access to the regular passenger to take the route as opposed to the NER's) along the route, have faster and more frequent service. If Amtrak believes what Selden says then why would they spend money to increase capacity and frequency to a service that can't fill the seats now (that probably is true ... for the NER's). I'm trying to figure out the downside other than the obvious cost expense. Could you buy 28 Viewliner and/or Superliner sets with the same money and if so how much would that increase service when a train like the Empire Builder requires 5 (or is it 6?) per week now (forget the fact you have to fight the host railroads)?

I'm guessing many of you feel Selden is a train god but I do question some of his arguments. That must put me on the outside ... again. I'm used to it though and what's a newsgroup if everyone agreed with everyone else)?
 
Passenger, passenger mile, train mile and load factors are all interesting facts but as the old saying goes "Show me the money." The only thing that matters is what is the total cost to run a train for a mile and what is the total amount of all the fares that the passengers paid for that same mile. What is the cost recovery? If the fares do not even cover the cost of the train then it is a triple loss. If the fares cover the cost of the train but not the overhead than it is a double loss. If the fares cover the train and the overhead but not the track charge then it is a single loss. If the fares cover all three then the train breaks even. That is not good enough. If one wishes to replace worn out equipment then the train needs to have a total operational profit. So are LD trains triple losers, double losers or single losers? Are NE Corridor trains single losers, breakeven or profit makers?

Does anyone even know?
 
The only thing that matters is what is the total cost to run a train for a mile and what is the total amount of all the fares that the passengers paid for that same mile.
On what basis do you consider train miles (loss per train mile) the only metric which really matters? What are you including in "total cost" (why not track usage?) and how do you appropriately divide expenses (and sometimes revenue) between long-distance, state-supported, and Corridor services?

Amtrak financial numbers are far more complex than you give them credit.
 
The only thing that matters is what is the total cost to run a train for a mile and what is the total amount of all the fares that the passengers paid for that same mile.
On what basis do you consider train miles (loss per train mile) the only metric which really matters? What are you including in "total cost" (why not track usage?) and how do you appropriately divide expenses (and sometimes revenue) between long-distance, state-supported, and Corridor services?

Amtrak financial numbers are far more complex than you give them credit.
My answer:

The equipment and labor for a specific train (single frequency) have to be allocated to that specific train.

The fees to use any tracks unique to specific routes used are allocated to that train (example, Capitol Limited between PGH and WAS, CSX's track)

The maintenance costs of any stations unique to specific routes used are allocated to that train (ex. Harper's Ferry to the CL).

It's harder to split tracks that are used by multiple trains (ex. NEC, CHI-Carbondale, CLE-CHI) or splitting the maintenance costs of stations shared by several trains.
 
Another thing to consider for discussing the NEC is that the Acela and NE Regionals should be considered different. One is for faster service aimed at business travel whose companies are willing to pay higher fares for faster and more convenient service while the Regionals are alternate travel more towards the regular average Joe. I'd be willing to bet that the load factor on Acela's is way higher than the NE Regionals and Acela's load factor compares more favorably to the LD's than the regionals.

I'm guessing Mr. Selden would be against the Avelia Liberty investment: http://media.amtrak.com/wp-content/uploads/2016/08/NextGen-Fact-Sheet_Proof5.pdf.

If they are as advertised, you would increase capacity (perhaps the ability to lower fares and still increase revenue and increase the access to the regular passenger to take the route as opposed to the NER's) along the route, have faster and more frequent service. If Amtrak believes what Selden says then why would they spend money to increase capacity and frequency to a service that can't fill the seats now (that probably is true ... for the NER's). I'm trying to figure out the downside other than the obvious cost expense. Could you buy 28 Viewliner and/or Superliner sets with the same money and if so how much would that increase service when a train like the Empire Builder requires 5 (or is it 6?) per week now (forget the fact you have to fight the host railroads)?

I'm guessing many of you feel Selden is a train god but I do question some of his arguments. That must put me on the outside ... again. I'm used to it though and what's a newsgroup if everyone agreed with everyone else)?
Load factor is a dubious metric for comparing the business lines. Yes, the NEC's load factors are lower...but at the same time that's because Amtrak offers a more comprehensive timetable on the NEC (not to mention, they also basically have to run the whole train WAS-NYP even if a bunch of seats are empty south of PHL...early/late Acelas, I'm looking at you).

As to the cost of the Acela IIs ($2.45bn, presuming that the whole package associated with them is considered part of the purchase), that would fund about a thousand new single-level railcars. Using a trainset model for this (even if they would be individual cars), let's say you want your "standard" LD train to be 12 cars (bag, 4 sleepers, diner, cafe, 5 coaches)[1] and two locomotives. Per Amtrak's estimates from a few years ago that's going to be $2.5m/car and $5-7m/locomotive (so $40-44m/set). That's 54-61 sets.

Right now the LD system uses something like 52-53 sets of equipment plus spares (I forget how many the Sunset Limited uses), so in theory you'd be able to increase LD service by somewhere around 100%, maybe a hair more if you were then able to use those frequencies to improve equipment turn rates (e.g. reducing the amount of equipment that sits idle for close or more than to 24 hours somewhere). My guess is that you'd be able to roughly get a 110-115% increase in service.

To be fair, there's not a strict apples-to-apples comparison here given the nature of some of the Western LD consists but you'd suddenly have a ton of additional LD capacity. Still, you'd probably be looking at ridership on the LD system coming close to doubling...maybe more, maybe less depending on the schedules that you opted for as a result (and any knock-on effects to other schedules and through-traffic). Still, that's probably somewhere in the ballpark of another 4m riders give or take.

Edit: If a mass order like that knocked the unit price of the cars down by 10% that's $37-41m/set that's another five sets. At $2m even (20% down) that's $34-38m/set and 64-72 sets (10-11 additional).

[1] Obviously there would be variations: The Silver Star presently runs with 7-8 cars and a single locomotive while replacing the Builder with single-level equipment while preserving present capacity would probably involve something like a 15-20 car train (I'm thinking 6-7 sleepers, 7-9 coaches, and 2-3 food service cars). You could also stipulate that on some routes with multiple long-haul trains there would be one or two "A trains" with a diner, second cafe/lounge, and more sleepers and one or two "B trains" with just an enhanced cafe and fewer sleepers. Playing with the Builder as an example, a Western Star/Mainstreeter running with more limited amenities on an alternate schedule would arguably cover many of the same riders...but should induce additional riders as well through the added convenience of a second daily train. In this respect, two 12-14 car trains might well perform better than a single 20-car train in terms of attracting ridership, even if one of them is obviously the "lesser" of the two.
 
Last edited by a moderator:
If you produce enough new cars to double LD service, it seems like negotiating with the freights for new slots is going to be a sizeable expense on top of the car order. Not a problem on the NEC because Amtrak owns the line.

Also, doesn't Amtrak need to replace a lot of the existing LD cars within the next decade or so? I thought that was largely what the latest five year plan budget was contemplating, with the 1000+ car order from 2019-2025. Not that I'm really expecting Congress to come up with the money for that.
 
Selden has always been good at cooking ... er massaging the numbers to support his somewhat off the wall position on LD trains. And of course we have to suffer to this nonsense ad infinitum through NARP from time to time too.
 
If you produce enough new cars to double LD service, it seems like negotiating with the freights for new slots is going to be a sizeable expense on top of the car order. Not a problem on the NEC because Amtrak owns the line.

Also, doesn't Amtrak need to replace a lot of the existing LD cars within the next decade or so? I thought that was largely what the latest five year plan budget was contemplating, with the 1000+ car order from 2019-2025. Not that I'm really expecting Congress to come up with the money for that.
Definitely agreed. My response was to Philly Amtrak Fan wanting an estimate using Viewliner sets for LD service.

Being a bit more realistic, such an order would probably be coach-heavy with an eye towards replacing and supplementing the existing Amfleet supply (said fleet approaching 40 years old and the cars would probably be approaching 50 years old by the time they were getting cycled out) and providing equipment for more service on the "extended NEC" (e.g. Virginia, New York, Pennsylvania, Vermont, and other states in the region), SEHSR, etc. as well as supplementing LD equipment. The real question is whether Amtrak would be inclined to swap any trains from Superliner to Viewliner under these circumstances to allow heavier use of the Superliners on the three daily transcons and/or the Starlight. We probably would get some new trains out of the deal, albeit not on the scale mentioned above (I suspect there would be not-insignificant pressure form the region for a North Coast Hiawatha and/or Pioneer in the package, as well as a Daily Sunset, Daily Cardinal, and a few of our other regular guests on this show...a second LSL might also show up, for example, as might a Broadway) but it wouldn't be a truly massive LD expansion.
 
Selden has always been good at cooking ... er massaging the numbers to support his somewhat off the wall position on LD trains. And of course we have to suffer to this nonsense ad infinitum through NARP from time to time too.
Isn't that what most people do when they want to support their argument? It is unfortunate that it will remain us vs them.

PhillyAmtrakfan, I have to say it has been agreeable to read your posts (on all boards) as of late. Even though we don't necessarily agree with some of your stances, the lack of venom in your posts makes it easier see what I do agree with. That being said, I do not agree with some of the metrics and your summary is golden. I would also add that the load factor on the NEC is always going to drop close to the originating and final terminals. Local travel is not excepted and local travel would hinder longer travel, particularly if some of the middle stations are "soft."
 
Or simply plow back the $500 million over the rail surplus into funding NEC maintenance. If the Commuter Agencies cough up a bit more then some of that surplus can be plowed back into mew equipment instead.
There's no free lunch. The $500 million still represents real money which has been accounted for already in Amtrak's annual budget. You can't come out ahead by redirecting Acela revenues into additional infrastructure work; You'll just need an additional half billion to make up the difference somewhere else.

The states, through the respective Departments of Transportation and the various commuter agencies, would balk at paying a greater share of Northeast Corridor maintenance and improvements, even though they often soak up greater capacity than Amtrak does on its own railroad. In theory, Amtrak could function fine with only one tube under the Hudson River in service; it's NJ Transit that couldn't, and needs more capacity in Penn Station. I'll agree the commuter agencies would fairly be paying more (much more, in fact), but politically the idea is a non-starter. It would be seen as just another way of shifting costs onto the states (again.....), absent a reasoned federal transportation policy. And, honestly, they would have a point.
 
Or simply plow back the $500 million over the rail surplus into funding NEC maintenance. If the Commuter Agencies cough up a bit more then some of that surplus can be plowed back into mew equipment instead.
There's no free lunch. The $500 million still represents real money which has been accounted for already in Amtrak's annual budget. You can't come out ahead by redirecting Acela revenues into additional infrastructure work; You'll just need an additional half billion to make up the difference somewhere else.

The states, through the respective Departments of Transportation and the various commuter agencies, would balk at paying a greater share of Northeast Corridor maintenance and improvements, even though they often soak up greater capacity than Amtrak does on its own railroad. In theory, Amtrak could function fine with only one tube under the Hudson River in service; it's NJ Transit that couldn't, and needs more capacity in Penn Station. I'll agree the commuter agencies would fairly be paying more (much more, in fact), but politically the idea is a non-starter. It would be seen as just another way of shifting costs onto the states (again.....), absent a reasoned federal transportation policy. And, honestly, they would have a point.
Well, and let's imagine that Amtrak did go ahead and pass the bill along. What would their reaction be if NJT and SEPTA turned around and said "You know what? We want to bid out several slots to a third-party operator to try and offset our costs..." Amtrak would (understandably if not justifiably) have a whole litter of calves at the concept.

To be fair, there's a block of NEC maintenance stuff that the states don't need for their trains (basically, most of the Acela-related improvements) and a large chunk of that does need to come from Amtrak...but at the same time I do think that should be allowed to cross-subsidize the rest of the system, if at a reduced rate (I do get the states not wanting to pick up the check for a bunch of 160 MPH improvements that they could never hope to use for their trains).
 
But the there also is a whole slew of tracks that needs to be kept operational mostly for the use of Commuter agencies, and just the operation of Amtrak trains do not require them.

The way to handle it is to manage tariff through the use of slots of various quality priced at appropriate levels.

As things stand, it is the NEC Commission, not Amtrak, that is tasked with figuring out who should pay how much, and it is quite likely that both Amtrak and Commuter agencies will land up paying more to cover the day to day maintenance to maintain SOGR. Capital improvements are the things that will be subsidized.
 
The only thing that matters is what is the total cost to run a train for a mile
Bzzzt. Wrong. You've failed to distinguish between fixed costs and variable costs. This matters. Look it up if you don't know the difference.
Which do you mean by "total cost"? I can tell you that practically every Amtrak train covers its variable costs. They nearly all generate a "gross profit".

The fixed costs, however, are most of the costs. Amtrak needs to run *more trains* in order to cover the fixed costs.
 
As to the cost of the Acela IIs ($2.45bn, presuming that the whole package associated with them is considered part of the purchase), that would fund about a thousand new single-level railcars. Using a trainset model for this (even if they would be individual cars), let's say you want your "standard" LD train to be 12 cars (bag, 4 sleepers, diner, cafe, 5 coaches)[1] and two locomotives. Per Amtrak's estimates from a few years ago that's going to be $2.5m/car and $5-7m/locomotive (so $40-44m/set). That's 54-61 sets.

Right now the LD system uses something like 52-53 sets of equipment plus spares (I forget how many the Sunset Limited uses), so in theory you'd be able to increase LD service by somewhere around 100%, maybe a hair more if you were then able to use those frequencies to improve equipment turn rates (e.g. reducing the amount of equipment that sits idle for close or more than to 24 hours somewhere). My guess is that you'd be able to roughly get a 110-115% increase in service.

To be fair, there's not a strict apples-to-apples comparison here given the nature of some of the Western LD consists but you'd suddenly have a ton of additional LD capacity. Still, you'd probably be looking at ridership on the LD system coming close to doubling...maybe more, maybe less depending on the schedules that you opted for as a result (and any knock-on effects to other schedules and through-traffic). Still, that's probably somewhere in the ballpark of another 4m riders give or take.

Edit: If a mass order like that knocked the unit price of the cars down by 10% that's $37-41m/set that's another five sets. At $2m even (20% down) that's $34-38m/set and 64-72 sets (10-11 additional).

[1] Obviously there would be variations: The Silver Star presently runs with 7-8 cars and a single locomotive while replacing the Builder with single-level equipment while preserving present capacity would probably involve something like a 15-20 car train (I'm thinking 6-7 sleepers, 7-9 coaches, and 2-3 food service cars). You could also stipulate that on some routes with multiple long-haul trains there would be one or two "A trains" with a diner, second cafe/lounge, and more sleepers and one or two "B trains" with just an enhanced cafe and fewer sleepers. Playing with the Builder as an example, a Western Star/Mainstreeter running with more limited amenities on an alternate schedule would arguably cover many of the same riders...but should induce additional riders as well through the added convenience of a second daily train. In this respect, two 12-14 car trains might well perform better than a single 20-car train in terms of attracting ridership, even if one of them is obviously the "lesser" of the two.
I like the way you think. Here's an elaboration. And I'm not trying to compete with the Avelia Liberty order here; I'm trying to make a straight-up pitch for more cars.

How much would it cost, and how many cars would we need, to simply do this on the eastern services? So:

-- LSL, and my "TWO A DAY" proposal running on the opposite schedule (6 trainsets?)

-- Broadway Limited -- TWO A DAY (6 trainsets?)

-- One a day Cardinal (worst case, 4 trainsets?)

-- One a day CL (3 trainsets?)

-- Two a day Silvers (as now -- 8 trainsets?)

-- Two a day Crescent -- one to New Orleans, one to Fort Worth (9 trainsets?)

That's about 36 sets of equipment. Amtrak is trying to make its platforms in the East suitable for 12-car trains consistently, so 12-car is a good standard to work with. So call it 1.44 to 1.58 billion dollars... except that a large order should be able to cut that down a bit. We do already have 50 perfectly good Viewliners and 50 more on order, not counting the baggage cars, so you can reduce that by $250 million. The order should still be large enough.

I think the exceedingly important thing to do is, however, to start buying the tracks. The freights have caused trouble with almost all attempts to get more service. Build South of the Lake and you relieve NS's biggest national headache, and this sort of proposal becomes a lot easier (note that I just proposed running 4 new passenger trains per day through that section from Porter to Chicago).
 
Last edited by a moderator:
Back
Top