Amfleet I Replacement

[jis]

However infra improvements that enable additional frequency usually trump merely a few minutes saved here and there.

 

[WoodyinNYC]

However infra improvements that enable additional frequency usually trump merely a few minutes saved here and there.

That seems to be the idea on the Cascades route Portland-Seattle, where the $850 million (or is it more like a Billion) current Stimulus-funds upgrades aim to increase frequencies by 50% Cascades only or 40% over Cascades plus the Coast Starlight (that is, 4 will become 6, or 5 will become 7). Meanwhile the on time performance is to improve, they promise, but only 5 or 10 minutes tops will be cut from the timetable. So basically this round is all about frequency and capacity.

OTOH, both the Chicago corridors seem to concentrate on faster times, about 40 minutes faster from St Louis and about 50 minutes faster from Detroit. There will be about a 30% increase in capacity from going over to bi-level equipment, but NO promised or announced increases in frequencies. Well, maybe they had little choice. More frequencies thru the South of the Lake corridor (without another Billion or two in upgrades SOTL) could bring the Norfolk Southern main line to a halt, as seen before. LOL. And of course, for both the Wolverines and the Lincoln Service, this is only the first round of needed, planned, and as-yet-unfunded upgrades.

But I'm not recalling a specific infrastructure project that expanded capacity without shaving run times, or one that cut times without at least potentially (like the Michigan trains) increasing capacity.

That study to add one-a-day corridor service St Paul-Chicago is weak on infrastructure and speed increases, but seems weakest in lacking enuff frequencies, disappointing from either point of view,

Anyway, I resolve any contradictions by favoring all the upgrades, LOL, Congress permitting.

 

[PVD]

Improvements have very different impacts based on type and location. Boston/NYP/WAS the train gets a very high percentage of the passenger load right now. Shaving small amounts of time probably would have very little impact. 2-4-6 NYP-ALB-SYR-BUF with an additional run West of ALB might take a fair number of cars off the road, at least to ALB. Improving corridor running times between key city pairs in the Midwest and perhaps in VA might have the same effect. Pre-clearance might help the Adirondack pick up some added MTL traffic, but it is still really slow North of Albany.

 

[neroden]

I'm assuming that overhead should grow more slowly than train miles or route miles increase.

Why, yes, yes, it should. In practice, we've seen ballooning overhead lately... I think a lot of that is due to the extremely high expense of replacing Amtrak's 1950s-era and 1970s-era IT systems, which shows up in overhead. If Amtrak ever finishes that process (ARROW is written in mainframe assembly language -- seriously?!?) the overhead from that should drop back to normal.

But I'm not sure what Amtrak includes in this overhead.

Everything you mentioned. EVERYTHING. Debt service, back shops, reservations system, corporate HQ, everything. This is the craziness of "fully allocated costs".

Some folks at headquarters are working on the pending order for Acela IIs, and the long delayed order for Viewliners. Are their costs allocated to the Coast Starlight and other trains by route miles?

I can't be sure ... but it certainly looks like it.

​I keep ending up where I left off. More Amtrak would spread overhead over a larger base, bring the ratio into balance, and reduce the allocated overhead per route mile or per train-mile.

Yep.

 

[neroden]

That seems to be the idea on the Cascades route Portland-Seattle, where the $850 million (or is it more like a Billion) current Stimulus-funds upgrades aim to increase frequencies by 50% Cascades only or 40% over Cascades plus the Coast Starlight (that is, 4 will become 6, or 5 will become 7). Meanwhile the on time performance is to improve, they promise, but only 5 or 10 minutes tops will be cut from the timetable. So basically this round is all about frequency and capacity.

The on-time performance is a big deal. The promise is that it will go up from 75% to 95%.

Riders are very sensitive to on-time performance. Any project which improves OTP significantly has huge payback in terms of ridership and revenue. This is why buying the tracks from the freight operator is likely to give the single best payback, even though it costs a lot and makes no physical improvement.

 

[afigg]

That seems to be the idea on the Cascades route Portland-Seattle, where the $850 million (or is it more like a Billion) current Stimulus-funds upgrades aim to increase frequencies by 50% Cascades only or 40% over Cascades plus the Coast Starlight (that is, 4 will become 6, or 5 will become 7). Meanwhile the on time performance is to improve, they promise, but only 5 or 10 minutes tops will be cut from the timetable. So basically this round is all about frequency and capacity.

OTOH, both the Chicago corridors seem to concentrate on faster times, about 40 minutes faster from St Louis and about 50 minutes faster from Detroit. There will be about a 30% increase in capacity from going over to bi-level equipment, but NO promised or announced increases in frequencies. Well, maybe they had little choice. More frequencies thru the South of the Lake corridor (without another Billion or two in upgrades SOTL) could bring the Norfolk Southern main line to a halt, as seen before. LOL. And of course, for both the Wolverines and the Lincoln Service, this is only the first round of needed, planned, and as-yet-unfunded upgrades.

But I'm not recalling a specific infrastructure project that expanded capacity without shaving run times, or one that cut times without at least potentially (like the Michigan trains) increasing capacity.

I think the difference in the emphasis of the improvements for the corridors you discuss is driven by what the problems are. The Cascades corridor has decent Seattle to Portland trip times, but historically poor reliability. The CHI-STL and CHI-DET corridors both have circa 5.5 hour trip times which hold down ridership and, particularly for the CHI-STL corridor, poor reliability due to single ended sidings, freight traffic, poor track conditions.
By cutting the trip times for 2 Midwest corridors to circa 4:30 to 4:45 AND improving reliability (well, whenever the track work is finally done), ridership will increase. The lack of additional frequency for the CHI-STL corridor I think is mostly on IL DOT, which added a 4th Lincoln service train some years back, but does not appear that they really focused on getting UP to agree to a 5th frequency in return for close to a billion plus in track, signal, and crossing upgrades. If the annual HSIPR funding had not dried up completely after the 2010 election, I expect we would IL would have gotten follow-on funding to double track more of the corridor and gotten the rights to run more daily Lincoln service trains from UP.

NC DOT has had a different emphasis on the Piedmont corridor with projects to separate and close over 20 grade crossings on the corridor. Given the persistence of grade crossing collisions in NC, especially involving trucks, that emphasis is not unreasonable.

There is, of course, considerable overlap in projects or upgrades to improve reliability versus to reduce trip times. Adding a 2nd or 3rd track, more or longer sidings, track rebuilds, cross-overs, new signals, longer station platforms to improve reliability will also usually trim padded trip times.

 

[Anderson]

(1) Frequency improvements and speed improvements aren't mutually exclusive. An improvement which allows a train to pass another train at a meet while moving is one example of this (you can save time for avoiding the stop, which presently has to be added to the timetable, as well as any situation-specific padding needed for reliability as a result).

(2) Reliability is a big deal, and I think that is no small part of the planned improvements on the Cascades. Ridership is probably going to be better on a train that takes, say, 2:30 with 95% reliability than on one that takes 2:15 with 40% reliability and intermittent major delays. Put another way, we don't complain too much about how long an LD train takes...but we sure aren't happy when it comes in six hours late. On most corridors the tolerances are tighter (e.g. you want arrival within 5-10 minutes of the stated time).

In the case of Virginia, this is why the goal is "90-90-90". 90 minutes WAS-RVR is a big deal, and connected with that is lifting the MAS on the route to 90 MPH (it is presently 70 MPH). However, the third leg (reliable 90% OTP) is important because if someone has a 10 AM meeting in DC they ought to be able to board the 8 AM train out of Richmond (on the faster timetable) and get there comfortably. If they have to board an earlier train and plan out more time than it would take to reliably drive (a very vague idea in NOVA, to be sure, with that being no small part of why VA's trains are making money).

(3) This isn't to say that 5 minutes here and 3 minutes there don't add up. At some point the incremental improvements add up to material improvements...either in getting travel time under highway times by a reasonable margin, getting city pairs under the magic three hour line, or other key/headline-grabbing points that help ridership.

(4) Finally, on frequencies...the problem is that each frequency on some routes is another couple hundred million dollars. Setting up a 4x daily corridor CHI-MSP, for example, could easily burn through well over a billion dollars.

 

[jis]

I think trip time minimization is the more important goal. Increasing MAS from 70 to 90 is one of the derived goals from it and indeed may have much lower priority if a bunch of 20 mph zones (curves, crossovers or interlocking) are available to raise to 50 mph somehow, or a passing track to increase reliability, for equal or close to equal cost.

 

[west point]

Trip time minimizing will have an important secondary benefit Using present timetables imagine a WASH - BOS train now scheduled 0315 - 1105 arrival ( 7:50 en route ).. If that trip time can be reduced to 6 hours then then the train set can cover the 1310 departure instead of the 1520 departure. Arrival WASH 1845 instead of 1945 ( present schedule 2305 for the 1520 ). Then cover the 2105 departure to NYP instead of terminating in WASH.

That allows the same number of rolling stock cars to be used for more miles a day increasing capacity. During slack times gives more time for car PM to increase reliability.

 

[Anderson]

I think trip time minimization is the more important goal. Increasing MAS from 70 to 90 is one of the derived goals from it and indeed may have much lower priority if a bunch of 20 mph zones (curves, crossovers or interlocking) are available to raise to 50 mph somehow, or a passing track to increase reliability, for equal or close to equal cost.

My understanding is that you simply can't get to 90 minutes WAS-RVR without a good chunk of the line being upgraded to 90 MPH. You have about a 110 mile run; doing that in 90 minutes requires an average speed of over 70 MPH (110 miles/90 MPH gives an average speed of about 73 MPH). 90 MPH is also derived from the fact that CSX has indicated that they'll work with 90 MPH on the RF&P but aren't inclined to cooperate with a higher MAS.

 

[neroden]

I think trip time minimization is the more important goal.

More important than top speed, but...
Based on psych studies, minimization of the 5th percentile (or 10th percentile, or 1st percentile) trip time is the appropriate goal. People seem to judge 'how fast' a train or bus is by the worst case they can remember, so if a train is an hour late 5% of the time, people will think of it as being an hour late all the time.

This means OTP is rather more important than anything else, up to a certain point.

Given the current situation with regards to OTP, this means that South of the Lake is by far the most important unfunded project Amtrak has on its future agenda. More important than new Hudson tunnels, since the one-at-a-time closure of the old tunnels mostly hurts NJ Transit rather than Amtrak.

Currently the catastrophic OTP messes from Chicago to Porter are severely damaging the viability of all the Michigan lines, and the LSL and CL. Englewood Flyover should have helped, but apparently NS mismanagement has eaten up all the gains from it -- although it's not as bad as when NS decided to melt down their entire system by ordering dispatchers to obey an incompetent automated program *only last winter*, it's still really bad. Amtrak needs its own route from the east into Chicago.

 

[WoodyinNYC]

Trip time minimizing will have an important secondary benefit Using present timetables imagine a WASH - BOS train now scheduled 0315 - 1105 arrival ( 7:50 en route ).. If that trip time can be reduced to 6 hours then then the train set can cover the 1310 departure instead of the 1520 departure. Arrival WASH 1845 instead of 1945 ( present schedule 2305 for the 1520 ). Then cover the 2105 departure to NYP instead of terminating in WASH.

That allows the same number of rolling stock cars to be used for more miles a day increasing capacity. During slack times gives more time for car PM to increase reliability.

Thanks for that good n clear explanation. But you may be more misoveroptimistic :giggle: than I am, to think times WASH-BOS will drop to 6 hours without many good years and many Billions gone by.

A second round of upgrades on the St Louis- and Detroit-Chicago routes could have the Lincoln Service and Wolverines combining trains for more efficiency. Or maybe we'll see one or two run thru starting late 2017.

 

[WoodyinNYC]

I'm assuming that overhead should grow more slowly than train miles or route miles increase.

Why, yes, yes, it should. ...

But I'm not sure what Amtrak includes in this overhead.

Everything you mentioned. EVERYTHING. Debt service, back shops, reservations system, corporate HQ, everything. This is the craziness of "fully allocated costs".

​I keep ending up where I left off. More Amtrak would spread overhead over a larger base, bring the ratio into balance, and reduce the allocated overhead per route mile or per train-mile.

Yep.

But now I'm confusing myself about Amtrak overhead allocated to the state-supported corridor trains. Did the post-PRIIA formulas limit all that, or what?

Because, of course, most of the sizable growth we can count on is from the three prime project corridors -- the Cascades, the Wolverines, the Lincoln Service. I figure 200,000 more passengers a year after the trip time cut on the Wolverines and on the Lincoln Services, and maybe 300,000 or more on the Cascades. If those 700,000 new passengers start to help pay for Amtrak's reservations system, Boardman's salary, etc that would help to reduce the overhead allocations on all the other passengers/train miles/route miles/whatever. But if the 700,000 new riders won't pick up a chunk of the overhead, well, damn.

Otherwise, on Amtrak's own accounts, the NEC services should continue to grow and presumably reduce the overhead allocated to other routes.

But big growth in the LD line is gonna be hard. I mean, 25 (or 30 equivalent) new sleepers could add $30 or even $60 million to the bottom line, but would we notice a change in revenues or passenger totals? Where and how will 95 or so refurbished Horizons boost revenue and ridership? To get more LD passengers, we're looking at a daily Eagle/Sunset plus Sunset Shuttle, roughly 125,000 more riders, and about the same from a daily Cardinal, probably more. But then what?

Maybe replacing the Amfleet I's and having longer trains could get as much added revenue and riders as the two new daily trains could get. But then what?

Wait. Stop. How did I get back on topic? :giggle:

 

[Anderson]

Shush! We were having a perfectly good off-topic ramble and then you came along! :giggle:

Ok, joking aside there are a few points to pass around:
(1) I've generally assumed that ridership on the daily trains would be somewhere around this formula:
Daily ridership=(3x weekly ridership*7/3)*1.05

Basically, I'm assuming that a daily train gives you a 5% boost in ridership per-train (about what I think was observed with the frequency cuts back in the 90s). That would bump the Cardinal and Sunset by about 150-160k each, or 300k-ish total.

(2) There appears to be about...I'd guess another 250-300k riders that have been lost due to lousy OTP which could easily come back...albeit needing a few years of increased reliability to induce.

(3) Each batch of 25 sleepers should be able to add somewhere in the range of 75-100k riders (depending on exactly how the rooms are being used, how demand shifts around, etc.). The 75k total is based on the Viewliner trains having about 150k riders last year and having about a 50% capacity boost; the 100k total is based on me making some guesses at load factors and the like and then reducing my estimates a bit to account for timing on routes/non-turnover of sleeper spaces (e.g. NYP-MIA riders taking a slot on two days). I'm ignoring the bag-dorms in this calculation, FWIW.

This would get somewhere around 700k riders onto the LD trains without adding anything not presently ordered. After this, there are a few other additions that would likely add significant ridership. One would be a JAX-MIA section on the Silvers, which would probably add about 100k riders per train. Adding north-of-Atlanta capacity on the Crescent, through coaches from the Pennsylvanian to the Capitol Limited, tweaking the LSL's schedule, possibly north-of-RGH capacity for the Star...all of these would probably add 25-50k per move, and again most or all of this could be done with existing equipment and some shuffling with the Horizons and Amfleet IIs. At that point there's a good chance you've managed to add about a million riders per year to the LD system, albeit with the cost of stretching equipment to the brink (you'd probably have some wiggle room with Superliner coaches to add cut-out equipment on, for example, the Zephyr or to attempt some sort of business class experiments on other routes...but that's about it). At this point I think "What's next?" is a bit like the joke about the rabbi who keeps asking it of the newly ordained priest...

Beyond that I think you could make a case for another 50 single-level sleepers (as well as some bag-dorms to complement baggage cars on some of the bigger LD trains) to deal with converting the CONO/Cap (at least on the sleeper side of things) to single-level and add more capacity on the eastern trains. You could probably stretch some more capacity out of a partial Amfleet reshuffle as well (e.g. adding an Amfleet I to each LD train for short-distance riders as new equipment is ordered), and there's the option of reviving the Silver Palm in some form as well.

Edit: To address the Horizon point...

I don't recall the mix of Horizon equipment (cafes vs. coaches), but 95 cars of some mix of the two is enough to equip 10-12 LD consists with 5-6 coaches and a cafe. Put another way, you could re-equip the Silver Service (Silver Meteor, Silver Star, and Palmetto) with 6 coaches and a cafe each for the Meteor and Star, run a ten-car Palmetto (8 coaches, a cafe, and a BC car), and have enough equipment for 20% spares. Or you could do the Meteor, Star, and a daily Cardinal with 6+cafe configurations and be in more or less the same position. With the Meteor and Star you'd be freeing up 32-36 Amfleet IIs; with the Cardinal you'd free up another 6 (I forget the numbers on the Palmetto since the consist there is a mix). That's enough to lengthen the LSL and Crescent, add a few through cars between the Cap and Pennsylvanian and possibly do full re-equips on one or two of the longer state trains.

 

[west point]

The Senate bill proposal to add some day time LD trips over routes with just night time trips

brings up several possibilities.

Western

1. SAN - LAX - SFO connections at SJC to Oakland.

2. PHX - LAX With the I-10 bridge collapse if this route was in service then a real service would be possible say 2 or 3 daily RTs.

3. DEN - CHI

4. MSP - CHI

5. MEM - CHI

Eastern routes

11. ATL - NYP

12. NYP - CHI various routes available

13. BOS - ALB - CLE

14. CLE - CHI

15. WASH - RVM - RGH - CLT - BHM Could be the new bi-level equipment

All these routes have current service just needs a large amount of equipment. Will reduce overhead.

Other overnight routes ?

 

[jis]

The length of the running time on many of these routes makes a daytime train over the entire route impossible. However, daytime trains over part of the route is possible.

 

[PerRock]

Just a little footnote about the AAF Siemens order.

The cars AAF are getting aren't new untested designs. Siemens is doing the same thing they did for the ACS-64, Velaro (Acela IIs/Cali HSR), and Diesel locos (bets it'll be called the ACR-##?) which is to take their already successful european designs and convert them for use here in the states.

The AAF coaches will be a modified version of the Viaggio coaches used by RZD, RIC/UIC, CD, and OeBB (CD & OeBB are branded RailJet). Just like the ACS-64 is a modified version of the ES-64u4, and the new diesels are modified versions of the EuroRunner platform. Having ridden the OeBB cars I can tell you they are quite comfortable & very modern feeling.

peter

 

[MikefromCrete]

Who the heck are RZD, RIC/UIC, CD and OeBB? This is like using station codes instead of the real names of places. It seems cool, but is unnecessarily confusing. Really, I've never heard of these carriers. I assume they are European railroads, but who knows?

 

[jis]

Who the heck are RZD, RIC/UIC, CD and OeBB? This is like using station codes instead of the real names of places. It seems cool, but is unnecessarily confusing. Really, I've never heard of these carriers. I assume they are European railroads, but who knows?

I typed RZD into google and it said Russian Railways. I suppose others could do so too if they really wanted to know instead of just moaning about it?

 

[PerRock]

RZD is the Russian State Railway, CD is the Cezch Republic's and OeBB (also seen anglicized as 'OBB' and technically 'ÖBB') is the Austrian Railway. RIC/UIC is essentially a leasing company in Europe for passenger cars (think TTX). All but RIC/UIC are pretty much the Amtrak's of some European countries.

peter

 

[jis]

Like Amtrak, well actually at a more advanced stage, CD and OBB are facing some significant challenges from private train operators, as is DB the German Federal Railway and PKP, the Polish Govt. Railway.

The reason that through international service ceased between Dresden and Wroclow is that the Wroclow region banned PKP before their chosen alternative operator managed to come up with an arrangement with DB. DB almost lost its operating franchise for the Nurenberg S-Bahn, and was saved by an obscure technicality. Around Berlin there are already a number of alternative operators holding franchise on various specific local routes.

Surprisingly to me, all these alternate operators seem to somehow come in with new equipment for their service.

 

[WoodyinNYC]

RZD is the Russian State Railway, CD is the Cezch Republic's and OeBB (also seen anglicized as 'OBB' and technically 'ÖBB') is the Austrian Railway. RIC/UIC is essentially a leasing company in Europe for passenger cars (think TTX). All but RIC/UIC are pretty much the Amtrak's of some European countries.

peter

It was a dirty job but somebody had to do it.

Thanks, Peter.

 

[MikefromCrete]

Who the heck are RZD, RIC/UIC, CD and OeBB? This is like using station codes instead of the real names of places. It seems cool, but is unnecessarily confusing. Really, I've never heard of these carriers. I assume they are European railroads, but who knows?

I typed RZD into google and it said Russian Railways. I suppose others could do so too if they really wanted to know instead of just moaning about it?

I don't think it's my job to do the work of the poster. If he wants to make a point he should make it clear.

 

[WoodyinNYC]

Just a little footnote about the AAF Siemens order.

The cars AAF are getting aren't new untested designs. Siemens is doing the same thing they did for the ACS-64, Velaro (Acela IIs/Cali HSR), and Diesel locos (bets it'll be called the ACR-##?) which is to take their already successful european designs and convert them for use here in the states.

The AAF coaches will be a modified version of the Viaggio coaches ...

A few years back Amtrak (& some states) issued the specs for new bi-level cars. Soon after some were ordered from Nippon Sharyo. With the new specs for single-level cars, not so lucky. (Hey, back on topic AGAIN?)

iirc The All Aboard Florida RFP called for bidders to meet the new Amtrak specs, and why not. This could be the launch order for Amtrak's new single-level fleet. AAF and Siemens will be eating much of the set-up costs for the newly Americanized cars. AAF had no other easy way to get good equipment. And it puts Siemens in the best position to bid for any Amtrak order.

Now, if the specs were drawn to favor certain mass production rail car makers around the world, i don't know. But I'd expect that it should not be too hard for various bidders to meet the terms from AAF and Amtrak.

 

[neroden]

But now I'm confusing myself about Amtrak overhead allocated to the state-supported corridor trains. Did the post-PRIIA formulas limit all that, or what?

As I understand it, before PRIIA, Amtrak charged a negotiated price for each of these; this was at least the incremental cost (Amtrak made money on every deal), but less than the 'fully allocated' cost.

Now, each state has to pay, IIRC, 97% of the "fully allocated" cost? Or 95%? Something like that, anyway... they found a loophole where it didn't have to be 100%. But either way, most states are paying a lot more in overhead than they were before.

This is the reason every single state complained about the PRIIA cost-sharing rules. Although the worst impact was on states like New York and Michigan which had to pay for trains formerly covered entirely by federal funding, every single state was charged more than before. Because a bunch of overhead, formerly not allocated, was allocated to each and every one of them.

Expansion of the system will spread that overhead over more trains and reduce each state's cost per train.

I am not sure, however, whether more passengers -- or shorter trip times -- on an individual corridor will pick up any additional overhead. If it's really being allocated by train miles, they won't change the overhead allocation.

Additional frequencies should, though: more train miles, so more overhead allocated to those trains. So when the Cascades go from 4 frequencies per day to 7, this takes some overhead weight off of every other train. Same thing when Illinois adds Moline service.