Amfleet Refresh

Not a particularly good comparison. Delta didn't have to pre-fabricate parts to keep 757's that have been out-of-production for 13 years in full state-of-repair with viable upgrade paths intact for future overhauls yet-to-come. Everything they fly still has vendor service-and-support and a robust maintenance supply chain that's guaranteed for a couple more decades. It costs AMTK a fortune to do that on the defunct Budd design for the Amfleets. Livery updates for customer service bona fides?...sure, that's easy, because it doesn't touch any car innards which are still wearing away underneath. Another 20-year life-extension midlife overhaul? They long ago crunched the numbers on that, and the results were brutal for what supply chains have to be replaced in-total to modernize those cars for the long-haul at any rational dollars-and-sense.

The more correct analogy would be "Why not keep flying 727's?", which were last produced in 1984...did go through multiple modernization overhauls straight through the end of the 20th century to keep up with the times...and could've kept going on the rebuild circuit except that the supply chain eventually ran out and any future modernizations could only be done in-house without vendor support. They got dropped en masse from airline rosters in the late-90's/early-00's once that reality set in. Right now the largest non-cargo operator of 727's in the world is the Mexican Air Force, who have just 5 of them in-service. Delta retired the last of their 727 fleet in 2004, when Anderson was Northwest's CEO in the middle of their merger transition into Delta (the merger also being the official end of Northwest's 727 fleet, which wasn't re-badged for Delta). He's got firsthand experience with exactly these kinds of fleet sustainability decisions.

The livery refresh was going to happen in some form or another because the practical fastest timetable for Amfleet replacements is a wee too far for milking the current decaying livery another 7-10 years without customer backlash. I doubt anything Anderson brought to the table substantially informed that decision, other than maybe stating preference to pay up-front for a preemptive 'package' update instead of doing reactive band-aiding (i.e. pretty much a marketing decision rather than a fleet management-driven decision). They already knew they had to do something, and had their final choices narrowed down to Door #1, Door #2, Door #3 long before he got hired. But that's not the same as pursuing a strategy to lean more on old stuff that's tarted up to look new...because there's no decision to make there. If they issued the Amfleet-replacement RFP tomorrow and knew in advance that it was going to be a PRIAA-fied Siemens Brightline, it'll still be 2025 before the last Amfleet I gets punted to reserve service, because it simply takes that long to produce, test, and accept 600 cars while pushing all the up-front paper. The current livery wasn't going to last until '25 without becoming reaaaallly dodgy, so the cosmetic refresh program was always a fait accompli.

Mr. F-Line, all I can add to your thoughtful analysis of the Delta business decision is that regarding B-727's, three Officers are needed; the third being the Engineer who "once upon a time" was not qualified as a pilot.

The DC-9 and its latter day variants have always been two Officers.

Finally, I hear you loud and clear regarding sources of structural parts. One can only hope that Amtrak can find enough qualified Machinists in both the Indianapolis and Wilmington areas to fabricate the parts needed to keep 40 y/o cars rolling. Hopefully the IAMAW sponsors good apprentice programs in those regions.

It's easy to forget that on "A-Day" all of those years ago all of the major carbuilders were still in business, still had drawings and in many cases specialized patterns to produce parts. The big failure in Amtrak car procurement is not that the designs aren't successful or long-lived (by almost any measure they are) but that the rail car industry became horribly distorted by the "big order" mentality. Amfleet cars were the last Budd orders (except for the V-1 shells), Superliner 1's the last by Pullman, V-1s the only by Amerail, and Superliner 2s the last complete cars at Bombardier's Barre Vermont plant (Acelas were finished in Plattsburgh NY). All of the companies survived a while after the last big order, but none of them thrived - at least not on long-haul cars.

The first result of this is that each order has more or less had to pay for the cost of its own factory from outfitting to retirement, massively inflating the per-piece cost. Skilled workers are lost after each order, and must be trained anew for the next one, at a different plant. "Buy local" provisions only make this worse.

The second result is that the fleet for any given service never gets a stable average age. Replacing five percent of the fleet each year would result in an average age of only 10 years. Instead there are new cars with unproven systems and old cars that are overdue for replacement. Assuming a fleet of 2000 cars to make the math easy only 100 cars per year are necessary. Split that across two builders and that's only one car a week that has to be delivered by each. With a continuous stream of cars even the auxiliaries work out better - new parts can be adapted into production with the cost of retrofitting older cars lowered by volume production instead of custom fitting.

It's unlikely that this pattern will change, given the political value of large orders.

Perhaps but I think Amtrak is trying to recognize this with their fleet-planning for the Single Level fleet. They don't want one huge batch, but a more consistent, year to year order.

I will partly disagree with a comment above. While the "plan" probably has state-funded cars coming first, I can see scenarios were a few states might be able to pony up funds before the Feds to get the ball rolling.

The plan is 600 cars, but honestly, I think 700-800 is a better long-term plan.

Of course, I can't see Amtrak replacing cars on a 1:1 basis right away.
Get 100 cars... get rid of your worst 50 cars.
Get the next 100, get rid of another 50 cars.
Repeat.
Ultimately you probably want to get rid of all the Amfleets, but for awhile, even if you only increase the total number by 50-100 that's a huge win. (For NYS for example, one goal has been hourly service to NYC from Albany. They almost have this. Another 10 cars I think would allow them to achieve this.)

The Metroliner cab cars, while essentially Amfleet shells (with smoother riding outboard bearing trucks), have a few more years under their belt. I assume as long as they're good for routine 125 mph ops, there's no real concern over structural fatigue on the younger Amfleets.

Allouette wrote:It's easy to forget that on "A-Day" all of those years ago all of the major carbuilders were still in business, still had drawings and in many cases specialized patterns to produce parts. The big failure in Amtrak car procurement is not that the designs aren't successful or long-lived (by almost any measure they are) but that the rail car industry became horribly distorted by the "big order" mentality. Amfleet cars were the last Budd orders (except for the V-1 shells), Superliner 1's the last by Pullman, V-1s the only by Amerail, and Superliner 2s the last complete cars at Bombardier's Barre Vermont plant (Acelas were finished in Plattsburgh NY). All of the companies survived a while after the last big order, but none of them thrived - at least not on long-haul cars.

The first result of this is that each order has more or less had to pay for the cost of its own factory from outfitting to retirement, massively inflating the per-piece cost. Skilled workers are lost after each order, and must be trained anew for the next one, at a different plant. "Buy local" provisions only make this worse.

The second result is that the fleet for any given service never gets a stable average age. Replacing five percent of the fleet each year would result in an average age of only 10 years. Instead there are new cars with unproven systems and old cars that are overdue for replacement. Assuming a fleet of 2000 cars to make the math easy only 100 cars per year are necessary. Split that across two builders and that's only one car a week that has to be delivered by each. With a continuous stream of cars even the auxiliaries work out better - new parts can be adapted into production with the cost of retrofitting older cars lowered by volume production instead of custom fitting.

It's unlikely that this pattern will change, given the political value of large orders.

While I'll agree pacing out both the ordering and building of new cars keeps the average age of railcars down, it doesn't keep the costs down. The Boeing 737 plant in Seattle ships 47 jets per month, an average of nearly 12 airliners per week. That's just one of their jet models from just one of their plants. If CAF was producing railcars at the pace Boeing produces jets from just one plant, they could have completed the 130 car Viewliner 2 order in 11 weeks. Please do not suggest construction jet liners is easier than building railcars....but I'll admit, Boeing has spent a small fortune over the years to make that plant so efficient.
Budd'S Red Lion plant built 6,834 lightweight stainless steel railcars over 40 years during its heyday, average 171 railcars per year, 14 railcars per month, 3.5 railcars per week. To be generous, most of that production was over 30 years, not 40 - so actual production runs average higher numbers. Source of data = http://www.trainweb.org/phillynrhs/BuddCarOrders.html" onclick="window.open(this.href);return false;

electricron wrote:
While I'll agree pacing out both the ordering and building of new cars keeps the average age of railcars down, it doesn't keep the costs down. The Boeing 737 plant in Seattle ships 47 jets per month, an average of nearly 12 airliners per week. That's just one of their jet models from just one of their plants. If CAF was producing railcars at the pace Boeing produces jets from just one plant, they could have completed the 130 car Viewliner 2 order in 11 weeks. Please do not suggest construction jet liners is easier than building railcars....

I will.
In part because there's a larger on-going market for 737s. This means a deeper supply chain, a trained body of workers and more. If we had a market for 9,000 Amfleet type cars, (even if you consider going back 30 years), I'm sure producing 300/year on average wouldn't be a problem.

The issue isn't necessarily "ease" but the infrastructure and the efficiency built into the system.
And not Amtrak related, but train related and to tie this together:

^ Ladies and gents, Amtrak's new single-level cars

I see new efficiency in airport-mainline connections. Rather than some remote walk from the train station just pull the train cars right up to the terminal gates!

I like the "cat's whiskers" frame on the front--that could save a bundle of money one day (provided the car doesn't get turned in transit somewhere).

It seems to me that almost any Amtrak order of passenger cars will become an orphan as soon as the last one is built, because no one else is continuing to buy that particular kind. So, all those worries about Amtrak having to build their own spare parts date back to whenever the original order of spares was used up.

Greg Moore wrote:And not Amtrak related, but train related and to tie this together:

Didn't a couple of those take a swim in the Clark Fork a few years back?

HEy with that cage up front, they probably beat Nippon Sharyo's crush resistance numbers, too.

All kidding aside, the Amfleet parts argument doesnt sway me much. Consider how many Amfleet were built: 642. It was 642 when Budd was around, and 642 now. There were no other takers for Amfleet. So regardless of Budd or Amtrak sourcing the parts, its not any more or less of an orphan fleet than it once was. Where once Budd made or bought and resold parts, now Beech Grove does. There is no change in the economy of scale.

As for Delta, I'm on their fleet every week pretty much. It's the nicest in the air of the American carriers. American's fleet could be from the eastern bloc circa 1990 and UAL's is a bit better but nowhere near delta.

Tadman wrote: Consider how many Amfleet were built: 642. It was 642 when Budd was around, and 642 now.

Less than 642 by now. Counting wrecks and write-offs, the Amfleet I total is around 455. Amfleet II is 120 (of 125) coaches, plus the 25 diner lite lounges.

Depending on how we count "Amfleet" carbodies, also the original Metroliner cars and a relative handful of SPV-2000 railcars. A few Metroliners survive as cab cars, of course, but are the (demotored) SPV-2000's still around?