Boardman Testimony 5 Mar 2013

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.
Status
Not open for further replies.

Anderson

Engineer
Joined
Nov 16, 2010
Messages
10,432
Location
Virginia
Does anyone know what was actually said beyond the PR here? Likewise, where on earth did the numbers in that PDF/PPT come from? That graphic shows the Silver Service combined slightly in the black, which is /totally/ at odds with last year's PIP, which showed the three at a combined CR of 85%. Likewise, it shows the Auto Train at break-even while last year's PIP showed the AT with over 100% DCR. Any ideas on what's up or how to find out what's up?
 
His viewgraph presentation is showing direct costs. not overhead. Since the route performance table in the monthly reports don't separate overhead from direct costs, we don't see the breakdown that Boardman was presenting. As you noted, some of the Performance Improvement Plans show more details in the breakdown of the total cost components, but those are for earlier FYs and are not that complete.

My takeaway is that the eastern LD trains and the 2 one night trip Superliner trains are either above or pretty close to breaking even on recouping direct operating costs. If all of the eastern LD trains can be made to run in the black with regards to covering direct costs, that could make it easier to restore a Three Rivers/Broadway Limited service. If there is enough equipment available (and already paid for) and the service could be added with only a minimal increase in shared overhead costs, Amtrak could add the train with no impact on the annual operating subsidy.

BTW, we had a thread started yesterday on the Boardman presentation about Amtrak's good cost recovery numbers. Maybe we should fold them into 1 thread.
 
Ok, digging back through my PIPs and MPRs, it seems that the FY2011 PIPs used FY2010's pre-OPEB numbers for their total allocated cost figures. At that point, Silver Service numbers were:
Meteor: $37.6m revenue, $76.6m allocated expenses
Star: $32.7m revenue, $78.3m allocated expenses
Palmetto: $16.2m revenue, $30.0m allocated expenses
Total: $86.5m revenue, $184.9m allocated expenses
I find revenues coming in a bit lower in the PIP than I do in the performance report (about $0.3m for the Crescent). The expenses are off by $0.1m, which is frankly a rounding error in this context. Also, going by the PIP's stated direct cost recovery numbers, I get direct costs of $100m.



Now, that's FY10. For FY12, we get:

Meteor: $42.6m revenue, $79.5m allocated expenses (revenue +5.0m, expenses +2.9m, margin +2.1m)
Star: $38.7m revenue, $82.8m allocated expenses (revenue +6.0m, expenses +4.5m, margin +1.5m)
Palmetto: $18.4m revenue, $29.3m allocated expenses (revenue +2.2m, expenses -0.7m, margin +2.9m)
Total: $99.7m revenue, $191.6m allocated expenses (revenue +13.2m, expenses +6.7m, margin +6.5m)
Honestly? With allocated costs on the Palmetto down (likely due to off-season disruptions due to construction), I can believe that overall direct costs were probably flat...and flat is all you really need to get the Silvers to break-even.
-I'm going to put an aside in here, but this argues for the Silvers producing a gross profit with at least some frequency over time, and backs up the old assertion that SCL considered sticking it out in Amtrak considering how massive those trains were.


LSL FY10: $29.3m revenue, $63.6m allocated expenses
Here, the revenue is off by a bit, but the expenses also list as being a bit higher in the PIP. PIP-indicated direct costs come to $38.6m.
LSL FY12: $35.0m revenue, $66.6m allocated expenses (revenue +5.7m, expenses +3.0m, margin +2.7m)
-If direct costs basically held the line between FY10 and FY12, this would produce a $2-3m loss in line with the graph.

Crescent FY10: $30.6m revenue, $70.8m allocated expenses
Crescent FY12: $34.9m revenue, $75.5m allocated expenses (revenue +4.3m, expenses +4.7, margin -0.4)

Going by what I'm seeing, it looks like the direct cost/revenue situation is improving on at least four of the five eastern trains covered here given that the allocated cost margins actually improved. The Crescent is probably improving ever-so-slightly, if I had to guess; I'm going to tentatively blame the situation ATL-NOL, which is likely killing its cost allocations (since it seems to be absorbing the same direct cost "damage" down there that it is further north).

At this point, I have to wonder...if Amtrak can get a bit more coach space on these trains and add the additional sleeper space the Viewliner IIs promise, how much of a contribution are these trains going to be able to make to the system, financially-speaking? For that matter, how much would adding FEC service help out?

Edit: Another thought: In a discussion with a friend, I once described Amtrak as being three different companies:
-The NEC, which made a lot of money.

-The eastern parts of the network, which could at least be pushed to break-even.

-And the western trains, which are a money pit.

This seems to be sort-of true: If nothing else, the Western trains are doing a lot to "soak" overhead that would have to be reallocated (and I'm going to object to the idea that the overhead is mis-attributed right away, partly based on the Builder's situation). With that said, I have to wonder what would happen if you simply ripped the overhead off of everything and looked at gross profits on a route-by-route basis.
 
Last edited by a moderator:
Backtracking from revenues and expenses via eyeballing the graph and adjusting from 2012 revenues, Palmetto is implied at a direct cost of ~$23 per train mile.

Silver Meteor is implied $41

Auto Train at $112

Lake Shore Limited is $51.42

City of New Orleans $39.60

Capitol Limited $48.47

Silver Star $43.09

Cardinal $40.24

Crescent $42.17

Empire Builder $44.16

Texas Eagle $29.77

Coast Starlight $62.47

Sunset Limited $57.03

Southwest Chief $45.48

California Zephyr $46.74

The Palmetto, Auto Train, Texas Eagle, Coast Starlight, and Sunset Limited have deviant costs. The Auto Train's is to be expected due to its nature and the Texas Eagle is likely an artifact from the through cars. If it weren't for the Southwest Chief being normal, I'd blame the Starlight and Sunset Limited. I'd really love to know what's going on with the Starlight; that's a massive difference from the apparent norm.

Initially I wanted to point to the lack of dining car and sleepers on the Palmetto as the reason that it's half the cost of other long distance trains, however, the Capitol Corridor has train expenses in the $43-45 per train mile range (Appendix C). This makes me believe that there's something rotten about all these numbers. I flat out don't trust the Palmetto's numbers and I suspect that the direct costs, in the supreme irony for all rail fan conspiracy theories, hide the actual costs of long distance trains; I do not believe that they include the OBS expenses (that being the only explanation I can think of for how a short corridor train can have equal costs to a train twice the length and three times the crew).
 
Does anyone know what was actually said beyond the PR here? Likewise, where on earth did the numbers in that PDF/PPT come from? That graphic shows the Silver Service combined slightly in the black, which is /totally/ at odds with last year's PIP, which showed the three at a combined CR of 85%. Likewise, it shows the Auto Train at break-even while last year's PIP showed the AT with over 100% DCR. Any ideas on what's up or how to find out what's up?
All the numbers Amtrak usually provides (in monthly reports, PIPs, etc.) are the next-to-useless "fully allocated" costs, which dump overhead various places. And because railroads are operations with enormous overhead, there's a lot of overhead to dump.

I believe, though I cannot be sure, that the numbers Boardman is presenting here (called "direct" costs) are something approximating the "avoidable" costs (though probably without station costs; who knows what they're doing with those). In short, if the Silver Meteor was cancelled, Amtrak would need a larger federal subsidy. I can easily believe that. Give the Lake Shore Limited a few more sleepers and watch it go into profit too. :)

Accounting is an art, not a science. You want different types of computations for different purposes, and we really don't know exactly which ones Boardman is using. One thing which makes it hard to compare anything from this year to previous years: Amtrak recently redid its entire accounting system, as ordered by PRIIA, so that it could get agreement on how to allocate costs to states.

Anderson's summary is pretty good.

The Eastern long-distance trains are pretty close to covering their direct costs.

- The Crescent needs cut-off cars at Atlanta (as mentioned in the PIP) so that it's not eating the same costs for the half-empty trip from Atlanta to New Orleans as it is from Atlanta to New York.

- The Cardinal needs to go daily and needs to be longer. And needs a faster route from Indianapolis to Chicago, of course.

- I don't know what to do about the Silver Star; the backtracking to Tampa has to be hurting it.

- The Capitol Limited appears to have hit a ridership/revenue wall, in that not that many people go between Chicago and DC; hooking it to a revived Broadway Limited, loosely what was proposed in the PIP, would probably be an improvement (more people go between Philadelphia and Chicago).

- The City of New Orleans probably has a similar problem to the Crescent (ridership dropping off as one goes south), but I'm not sure where you'd put the cut-off cars. Memphis?

The Western trains cost a lot to operate and suffer from going through low-ridership areas, but even there there is room for examination of detail.

In particular, the California Zephyr performs badly financially because of the very long, slow mountain crossings between Denver and Salt Lake and between Salt Lake and Sacramento; this also includes the sections with lowest ridership. It requires more trainsets than any of the other transcontinentals.

I would love to see a nominal breakdown of costs and revenue between Denver-Chicago and Denver-California. I would bet Denver-Chicago, which on its own could be operated with two trainsets -- and they could be quite long, not having to go up mountains -- is doing much better than Denver-California. Before Iowa's legislature got taken over by crazy anti-rail nuts who refused free money for rail, Iowa was proposing a 79mph or 110mph Iowa Interstate (old Rock Island) corridor from Omaha to Chicago, hitting most of the major cities in Iowa. Reroute the Denver-Chicago route over that, and you could have a very successful "Denver Zephyr" which would probably do as well or better than the City of New Orleans.

Similarly, get a corridor from Chicago *through Madison* to Minneapolis, run the Empire Builder on it, and watch performance improve. The Empire Builder does better than the other transcontinentals due to the heavy online traffic from small towns, but skipping Madison hurts it a lot.

The Coast Starlight is probably also suffering from "mountain" expenses due to the mountain crossing between California and Oregon. It also has a poor revenue seating percentage in its consist, due to the "Arcade Car" (though they're planning to fix that) and the PPC. Also abnormally high maintenance costs due to the PPC. Although there's enough demand to lengthen the train, the mountain crossings mean that this would likely require additional engines. At the moment I see no way of improving it, and it's important to maintain the links which it provides; perhaps after various California rail projects are built it will become more obvious what to do.

The Southwest Chief suffers from running what is really a very short train over a very long distance. The lack of intermediate traffic generators is what keeps the train short. Reroute through Amarillo and Wichita, please! If you also extend the Heartland Flyer to Wichita, you'll start to get network effects (Albuquerque to Fort Worth, etc.)

The Sunset Limited -- well, again, less-than-daily is no good, but also missing Phoenix is no good, as we all know.

---

I'll say something else about the numbers in this presentation: they "feel right" to me as avoidable-cost numbers, which the usual numbers do not (though of course the usual numbers aren't avoidable-cost numbers). I've never believed numbers which make the Lake Shore Limited look like a poor performer, given the large number of cars, high occupancy, high prices, and relatively good speed. With these numbers, there's no relationship between two trains' numbers where my reaction is "No way, that one can't be performing worse financially than that one, that's absurd". With the usual numbers, that IS my reaction.

One thing to consider, Paulus: the Capitol Corridor is probably paying more to Union Pacific than the Palmetto is paying to CSX, for reasons of history (Capitol Corridor was "new service", Palmetto wasn't).
 
Eastern trains:
-I think you've hit it on the head with the Crescent and the Cardinal. The former is probably more likely than the latter...all of those issues between the VA/WV border and IND/CHI are going to be a nightmare to work through.
-With the Silver Star, an FEC section would likely be a major boon. As it stands, you've likely got virtually no business from anywhere "up north" to anywhere past Tampa on that train, just a lot of intrastate business. Offering a section to MIA via the FEC line would whack a couple of hours off of the JAX-MIA time (SB would depart JAX at 7:15 AM and arrive MIA around 1:30-2:00 PM; NB would probably depart MIA around 4 PM). If AAF's plans come to full fruition, this might even trim down further (110 MPH running was never in the timetable back in the 50s).
-I agrew with the Cap. The other thing I would /seriously/ consider would at least be a through sleeper between the Cap and the Silver-of-your-choosing.
-I also agree with Memphis cut-off cars for the CONO. This was done at Thanksgiving a few years ago, IIRC, and the market seems to be big enough even with messy times there. The other option would be a Carbondale cutoff, since Memphis seems to send a lot of folks both ways (MEM-NOL is actually bigger than MEM-CHI, and is the biggest market on the train), and since Carbondale-north is likely boosted by the IL-sponsored trains on that end of things. Either option works, though MEM would at least add some optional capacity for both cities.

Western trains:
-CZ: I'd like to see the situation in five segments:
(1) CHI-DEN
(2) DEN-Glenwood Springs
(3) Glenwood Springs-SLC
(4) SLC-RNO
(5) RNO-EMY
Looking back to the original CZ, my guess is that ridership on 1 and 5 is very strong. It's probably a bit lower Denver-Glenwood Springs, but not too bad...lower still out to SLC...and in the gutter RNO-SLC. This sort of pans out from Amtrak's early years: A Denver Zephyr would split off seasonally, even with the Overland Route routing.
--What you probably need here is a substantial cutoff set at Denver and another at Reno, with some seasonal variability in each. I'm thinking 2 sleepers and 2 coaches at Denver, 3 coaches at RNO, and a CCC that runs through but gets a staffing cutback between DEN and RNO. Amtrak more or less hinted at this in the PIP for the CZ, and I'd like to see it aggressively worked up.
--If a reroute can't be managed in IA, it would at least make sense to aggressively Thruway the heck out of Iowa. At the very least, DSM needs a connection; you could probably add a few cities to bus combos and get some serious business there as well.
--The other option, quite frankly, would be to revive the Desert Wind and have it fill the "Denver Zephyr" need east of DEN. From what I can tell, it's PRIIA-allowable (as a CHI-LAX train); the main issue would be justifying even more train capacity west of DEN. A pallative here might be if you could force better times for SLC on the second train...midnight WB and 0300 EB are not the best hours to sell people on, after all. SLC really ought to be a bigger city for the Zephyr than it is.

-EB: Really, I don't think there's much improvement here without additional frequencies and/or better reliability MSP-CHI. CHI-MSP can hurl off all the business it wants, but when winter trains are regularly hours behind on the return, that kills your business market. What you probably need is a corridor train to carry the MSP-CHI cars back in some sort of deal, since I suspect you'll get more eastbound business on the corridor train and westbound busines on the Builder.

-CS: Here what may be needed is also cutoff cars somewhere in the southern end. LAX-EMY/SAC cars would probably do well, and I could see a "full" BC car being run here if the BC experiment succeeds. I don't think there's quite as much business north of SAC, so at least some "mountain costs" could be ducked.
--The impact of the Coast Daylight will be interesting. Total ridership should jump, but it's going to be interesting since the Daylight will be a clear backup train to the Starlight's premium nature.

-SL: This is a train that simply doesn't work well. It covers a lot of ground, and there are a respectable number of workable markets, but it's a route that sorely needs a second train to make more pairings work well and give everyone decent-hour service. The other consideration would be more aggressive Thruway offerings (Phoenix, etc. come to mind, and I can't help but wonder, should the timing ever be passable, if Vegas might not be worth considering as well). Still, it's a particularly messy item, especially west of SAS.

-The TE is another train that has some possibilities: A Houston section is one, and it would be nice if you could improve some of the intrastate times in TX as well.
 
I'll need to look at the video sometime soon. As to the energy use, at least occasionally it gyrates due to fuel hedge valuation.
 
I'll need to look at the video sometime soon. As to the energy use, at least occasionally it gyrates due to fuel hedge valuation.
It is a ridiculously high energy consumption however. There are other Amtrak documents which confirm that as well (including an average kWh per-seat mile figure that's 2.5 times that of a TGV).
 
It is a ridiculously high energy consumption however. There are other Amtrak documents which confirm that as well (including an average kWh per-seat mile figure that's 2.5 times that of a TGV).
If I were to wager a guess, I'd say part of it has to do with the fact that the trains are heavier and lower capacity, to start with.
 
--What you probably need here is a substantial cutoff set at Denver and another at Reno, with some seasonal variability in each. I'm thinking 2 sleepers and 2 coaches at Denver, 3 coaches at RNO, and a CCC that runs through but gets a staffing cutback between DEN and RNO. Amtrak more or less hinted at this in the PIP for the CZ, and I'd like to see it aggressively worked up.
Cutoff coaches at Denver would be quite easy starting in early 2014 when work at Denver Union Station is finished. The new Denver station will effectively have two platforming tracks for Amtrak alone (unless someone revives Ski Train), plus an auxiliary track (the current "temporary station" platform track), plus a headshunt. (Of course, it would be nice if someone would revive Ski Train, but there would still be enough room.)

Cutoff coaches at Reno.... would require trackwork. They'd have to be physically cut off somewhere east of the station (there's no way to store rolling stock in the trench), and it would probably require construction of a new siding, and I'm not sure where you'd put the siding. It would at the least require paying UP for use of a siding. This therefore seems less likely. It is perhaps more likely to see one Capitol Corridor trip extended to Reno, perhaps seasonally, to relieve the demand on that section. There is already talk of this.
 
Cutoff coaches at Reno.... would require trackwork. They'd have to be physically cut off somewhere east of the station (there's no way to store rolling stock in the trench), and it would probably require construction of a new siding, and I'm not sure where you'd put the siding. It would at the least require paying UP for use of a siding. This therefore seems less likely. It is perhaps more likely to see one Capitol Corridor trip extended to Reno, perhaps seasonally, to relieve the demand on that section. There is already talk of this.
Sparks, NV. That's where the Zephyr's PRIIA report suggested the cutoffs
 
Cutoff coaches at Reno.... would require trackwork. They'd have to be physically cut off somewhere east of the station (there's no way to store rolling stock in the trench), and it would probably require construction of a new siding, and I'm not sure where you'd put the siding. It would at the least require paying UP for use of a siding. This therefore seems less likely. It is perhaps more likely to see one Capitol Corridor trip extended to Reno, perhaps seasonally, to relieve the demand on that section. There is already talk of this.
Question: how does the Reno party train cope with these issues?
 
Cutoff coaches at Reno.... would require trackwork. They'd have to be physically cut off somewhere east of the station (there's no way to store rolling stock in the trench), and it would probably require construction of a new siding, and I'm not sure where you'd put the siding. It would at the least require paying UP for use of a siding. This therefore seems less likely. It is perhaps more likely to see one Capitol Corridor trip extended to Reno, perhaps seasonally, to relieve the demand on that section. There is already talk of this.
Sparks, NV. That's where the Zephyr's PRIIA report suggested the cutoffs
I'm honestly thinking both options might come to pass. EB, the Zephyr is great for this. WB, it can suffer from usual LD train end-of-run issues with timekeeping and food stocks. What I would expect, honestly, would be a CC train extended year-round plus cutoffs during peak season (especially since I suspect that the CC being added would augment demand somewhat due to adding a second frequency there). Amtrak has said there's probably enough demand for a whole train on that section (that they spoke of three coaches for the section says something), so adding a frequency that also has to do double duty as a standard CC train would likely result in a very crowded CC train by the time it gets to SAC.
 
I agree with this thought below. I remember riding the old Rio Grande Zephyr before it was shut down. The train was jam packed up to Glenwood Springs, at which point about 7/8th of the passengers got off the train. The balance of the train emptied out largely in Grand Junction. I remember several trips where I could choose which dome car I wanted to ride in alone in to Salt Lake City. Amtrak currently limits the number of seats available DEN - GS, so if more coach seating could open up, it could be really competitive. Particularly with I-70 turning into a parking lot every weekend from Denver through to Vail, the train would be a nice option. Only problem would be UP, negotiating with them for anything different could be a challenge.

Western trains:
-CZ: I'd like to see the situation in five segments:
(1) CHI-DEN
(2) DEN-Glenwood Springs
(3) Glenwood Springs-SLC
(4) SLC-RNO
(5) RNO-EMY
 
Status
Not open for further replies.
Back
Top