Could a version of the Auto Train sans the auto be successful?

[Thunder Road]

In other words, a non-stop Superliner consist but only for passengers. Is there enough demand for a third Florida train, or perhaps some hypothetical system where the Star and Meteor ran opposite days, but this train ran every day?

I know WAS can handle Superliner equipment from the CL, but could it also handle a southbound Superliner train? If so, then it would probably be better to have the Auto-less Auto Train run right out of WAS so it could connect with the NEC.

 

[gswager]

Probably not. What is the percentage of passengers use end to end trip on Star and Meteor trains?

 

[Ryan]

Yes, (as Alan reminded me this week) the Cardinal used to run with Superliner equipment.

Interesting idea - not sure if there is the demand - part of the good thing about train travel is the large number of city pairs covered by a single train.

 

[battalion51]

I think you need look no further than the PRIIA Documentation to answer that question. The average coach passenger travels 499 miles, while the average sleeper passenger travels 888 miles, but sleeper passengers only account for 6.83% of the trips taken (on the Star, roughly similar numbers on the Meteor).

All that being said, what makes long haul trains successful in general is passengers that travel to intermediate stations. Most of the time the Silver trains have 240 coach seats available, but most of those seats will be "turned" two or three times and be utilized by folks making shorter trips. For example, on the Star, 19% of the passengers that got on the train in total were traveling between Tampa and another city in Florida.

The better goal I think for Amtrak to focus on is to try and build more Regional feeder trains that supply the Long Distance network. Continued growth and movement towards breaking even will come from short haul services, not long haul in my opinion.

 

[D.P. Roberts]

I wonder if an Auto Train could work on other routes. In other words, build some car carriers & charge passengers extra for bringing their car along with them. I think the difficulty would be in getting the cars on & off at midpoint stops, so you'd probably have to limit the stations where passengers could do this.

For example, I wonder what percentage of Auto Train users come from the Virginia area. If they find out that all of those people come from, say, the Chicago area, they could perhaps add a car carrier to the CL, and allow people to put cars on in Chicago.

 

[xyzzy]

91/92 carry a lot of people between the NEC and points as far as south as Raleigh. As another post mentions, they also carry intra-Florida traffic. But through South Carolina, 91/92 are usually half empty. 97/98 tend to carry more passengers over a longer distance. Most people would agree there is more demand for both short-haul and long-haul than Amtrak can currently accommodate with the consists of 91/92 and 97/98. Whether the additional demand is better addressed by lengthening the existing consists or extending 89/90 into Florida (again)... that's a matter of opinion. But it's all academic until new cars are available.

 

[battalion51]

Don't forget, the second you tack an Auto Carrier on to the train you instantly go from a Maximum Speed Permitted of 79+ MPH down to 70 MPH. It'd better be a very large, very lucrative market to start sacrificing speed on other routes. Amtrak would likely be better served by hauling mail or items for UPS before adding carriers to trains. And I won't hold my breath for the mail contract to come back...

 

[Nathanael]

Interesting question. I'd say the end-to-end traffic on the Meteor and the Star isn't enough to justify it. Nonstop trains are rarely the best use of trains; an extra stop can cost maybe a couple of minutes on schedule (at most) while bringing in hundreds of passengers daily.

In the 'old days', nonstop trains generally only existed on routes which already had lots of local, limited, and express trains -- and which had a particular character with a lot of end-to-end traffic. For instance, the "Twentieth Century Limited" ran nonstop from the Chicago metro area (which had 1 downtown, 1 suburban stop) all the way to the NY metro area (which had the same), but it was overlaying a whole lot of trains which stopped more frequently.

What would be the equivalent today? In the US, probably the NEC. The Acela is already providing the express role. But there's not enough traffic from Boston to south-of-Philly to justify a non-stop starting north of New York. And New York to Philly is stronger than New York to Washington, making it hard to justify a non-stop from NY to Washington. While NY to Philly is too short to bother with an exclusive nonstop train...

I don't think a proper non-stop would be successful in any corridor at the moment. However, as train service extending south of the NEC increases, there's probably a market for some more "limiteds", making only major stops. I could imagine a Norfolk Limited, stopping only at Norfolk, Richmond, DC, Baltimore, Philadelphia, and NY.

The Auto Train is only non-stop due to how time-consuming and annoying it is to load/unload cars at multiple locations.

---

A more interesting question would be whether additional Auto Train services would be successful. They've tried Chicago to Florida, but it couldn't be made fast enough to be popular. Chicago to Arizona was considered in a recent PIP, but rejected as having worse cost recovery than the existing Auto Train and not worth it.

I think there is one really natural extension: get a northern terminal of the Auto Train further northeast. It is known that lots of the customers are coming from the NY area. You'd probably want to retain the Lorton terminal, but you really want a terminal up in New Jersey or even Connecticut. This has been prevented by a lack of rail routes with sufficient clearance. It should soon be possible to get autoracks through DC due to CSX's clearance work on the Virginia Avenue tunnel. I believe the next major obstacle is Baltimore. If the multibillion dollar new tunnels (tunnels for freight and tunnels for passengers) which are currently planned are built in Baltimore, it would likely become possible to get an Auto Train all the way to the Jersey suburbs of New York, and I'm sure it would be possible to build a terminal there. (New York is the next major obstacle.)

 

[afigg]

I think there is one really natural extension: get a northern terminal of the Auto Train further northeast. It is known that lots of the customers are coming from the NY area. You'd probably want to retain the Lorton terminal, but you really want a terminal up in New Jersey or even Connecticut. This has been prevented by a lack of rail routes with sufficient clearance. It should soon be possible to get autoracks through DC due to CSX's clearance work on the Virginia Avenue tunnel. I believe the next major obstacle is Baltimore. If the multibillion dollar new tunnels (tunnels for freight and tunnels for passengers) which are currently planned are built in Baltimore, it would likely become possible to get an Auto Train all the way to the Jersey suburbs of New York, and I'm sure it would be possible to build a terminal there. (New York is the next major obstacle.)

It will be 10 years at least before the B&P replacement tunnel would be completed. What is not clear is whether CSX will be interested in contributing a large piece of the capital needed to build a parallel double stack freight clearance tunnel. CSX may decide that they are satisfied with the Howard St tunnel and running single stack trains up to Philly over the large capital investment to have double stack clearance through Baltimore. OTOH, if MD is willing to pay much of the cost for a new freight tunnel in return for the Howard St tunnel for light rail, getting hazardous freight traffic away from downtown Baltimore, and boosting container traffic capacity from the Baltimore ports, CSX may be interested.
If there is someday a route with sufficient clearance for the AutoTrain via CSX through DC, Baltimore, and up to NJ, I don't think it would have to go north of northern NJ. A new AT that departed from a station in northern NJ would pull in the greater NYC region, southern New England, Boston markets. It could still reach Sanford overnight, although a longer trip. But this is all academic because the Virginia Ave tunnel rebuild in DC has not even started yet.

As for the original question, I agree that a super express LD train from WAS to FL would be a major money loser. Now extending a Capitol Limited from WAS to Orlando, as discussed in a way in the CL PIP report, might be worth considering if Amtrak were to order enough Superliner I replacements. The PIP discussed combining the CL with Silver Star which adds all sorts of possible trip delays that could hurt both CL and SS ridership. Rather than combine the trains, leave the Star and Meteor as is and extend the CL to Orlando for a 3rd WAS/NEC to FL train. Terminate the new CL at Orlando to have the AT Sanford facility maintain the Superliners. Whether CSX would be willing to host an additional passenger train south of Petersburg is an open question. Adjust the schedule so the southbound CL gets to Orlando 8 AMish.

Drawback to this idea - besides the need for new Superliners - is that this would turn the CL into a 2 night trip and the 2 night CHI to west coast LD trains lose the most money.

 

[jphjaxfl]

The winter only Florida Special from New York to Miami which operated from mid December to mid April had limited stops even the first year of Amtrak operation and could fill up because the market was there. Auto Train grabbed some of the Florida Special business. Many of the people from Manhattan and the rest of the NYC don't own cars to take on Auto Train though. They usually rent when they get to Florida. If you could get back some of the traffic that Amtrak lost in the New York - Florida market, a winter only limited stop non Auto Train might work.

 

[dlagrua]

I do not believe that any LD point to point train to Florida would be successful. The Autotrain passengers want their car with them. Their market is mainly snowbirds and family vacationers. Passengers who insist on rail travel without their car would be on the Silvers but air travel still represents most of the passenger traffic to Florida.

The Silvers fill up (sometimes) by receiving passengers from locations all along the eastern seaboard to Florida. IMO adding another LD Florida service with only two boarding points would be an ill conceived idea.

 

[user 1215]

Purpose built end-to-end trains like the Florida Fun Train and ACES in New Jersey tend to be a bust. Even non-stop Acelas between WAS and NYP or WAS and BOS don't seem to be practical. The AutoTrain seems to be the only exception. Even then, the AutoTrain's business model was tried on other routes to abject failure.

Batallion51: Even though sleepers make up 6% of the riders, what percentage of the revenue do they make?

 

[Anderson]

I think a Florida Special style train that runs nonstop (or close to it) RVR-JAX might work. You'd need stops along the NEC (NYP, PHL, WAS) and in Florida (JAX, ORL, MIA), but I think this could work. A large portion of the Silvers' business is NEC-FL...it's just not aimed at a small number of city pairs.

What might actually have a market is a BOS-MIA train that's more-or-less R-only in the NEC (I'd allow a limited amount of tickets to be sold BOS-WAS and the like, partly to offset the buttload of passengers that those trains tend to pick up in DC...I don't mind some restrictions, but I have this aversion to seats going empty for 6-8 hours), runs R/D in a lot of Florida (likely restricted only down in Tri-Rail land), and either has no passenger stops in between, or only has one somewhere like Florence (where the train has to stop anyway).

Obviously, this would have to be a single-level set, but I do think there's a market for it, especially if by running non-stop they could whack 90 minutes-two hours off the schedule by removing both stop time and intermediate padding. The biggest downside is that they'd probably get stuck with a train parked in JAX (or wherever the first FL passenger-recieving stop was) for a while.

Hmm...thinking about this, I wonder if they couldn't limit themselves to 2-3 "turnover" stops:
(1) WAS

(2) ORL

(3) TPA (if applicable)

DC adds a lot of passengers, so allowing longer BOS/PVD-WAS traffic seems fine, and the engine change can simply be excluded in the departure time NB (so as to allow a quick departure there). Orlando is going to be one of the bigger destinations, and any intervening discharges that are done can simply be added in there. And Tampa is going to be a major discharge stop no matter what.

 

[battalion51]

Venture, the sleepers obviously have a larger financial contribution because you're getting rail fare plus accommodation charge, I don't dispute that the yield is different. However, the point is that isn't nearly as common for rooms to turn on a train, whereas coach seats tend to turn almost as soon as they're emptied. Part of the model here is that you could end up having a decent yield on that seat with the number of turns it makes on a single trip, whereas your yield is generally going to be on the single occupants of the room for sleepers since they don't turn. Revenue per sleeper mile and revenue per seat mile really are the most appropriate metrics.

 

[chakk]

Extend the existing Auto Train too far north (or too far south), and the equipment won't be able to make a round trip within 48 hours, necessitating at least one full set of additional equipment, with the resulting lower percentage hours of utilitization in service.

Also, if the southbound train departs its northern terminus too early in the day (to insure arrival at the southern terminus by mid morning), then it might be difficult for people to drive to it in time to get loaded without leaving another day earlier than they do now.

 

[battalion51]

Chakk makes a good point about the extension. Let's say it takes five hours to get from Jersey to Lorton (keep in mind maximum speed is 70 MPH, so even five hours may be lowballing it a little), that puts you at a 11 AM departure and 2:30 arrival. So the train would stop accepting oversize vehicles at 9 AM and stop accepting regular vehicles at 10 AM. You'd also likely be adding lunch to each segment arriving/departing from Jersey, so that increases your cost there. Now, as for the equipment sitting for so long the only way I could see this working out well is if you can figure out a way to do it with single level equipment you could run it as a late night/early morning train to Albany with the coaches, but that gets messy if you've got late trains. All in all, you're reaching the point of diminishing returns if you go to Jersey.

Now, if you were to go a bit further up to Springfield and try to capture the Massachusetts/Connecticut/Rhode Island/Upstate NY/Vermont/New Hampshire/Maine/Quebec market you might have something. The run is a bit longer, but that plays to your advantage for equipment utilization, but you'd have a very early departure from SPG, so that might alienate some of the appeal...

 

[Guest]

Would any extension, be it to NJ, CT, or even VT, still be able to support the original assertion of being a "non-stop Superliner consist" to FL ? I mean, to me, such an extension would require its own full set of consists, which would be in addition to today's Autotrain consists. Though, I guess, spares would be shared from the same pool.

Plus, would such a consist being of two-level equipment be enough of an added draw, over single level equipment (today's Silvers), to increase passenger volume?

 

[user 1215]

I think that the biggest problem with non-stop service is that it wouldn't necessarily be any faster. Maybe a few minutes saved from station stops, but not any more than that. It's hard enough watching CSX dispatch the Autotrain around 2-hour late Palmetto and Silver Stars. To add another "top priority" train would be devastating to the infrastructure.

Even the Acela isn't a HUGE time savings over the NERs. The only way to get TRUE high speed rail and TRUE time savings is by having a dedicated HSR ROW. That would be essentially saying, separate the Acela tracks completely from the NER tracks. If that's done, then it should be a bit easier to have more non-stop trains that can leapfrog others that are stopped in stations.

Outside the NEC, though, I think this is dern near impossible.

 

[Anderson]

Well, I think one thing we're missing is that the suggestion here would be able to operate over 70 MPH because it wouldn't have autoracks.

The other thing is that in a lot of longer-distance trains, you've got padding both at the ends and throughout the route. If you drop intervening stops, you can cut out some padding/recovery time in the middle.

 

[jis]

In India which has a huge LD train ridership market most end to end non stop trains (the Duronto Expresses) are not daily trains. They run anywhere betwwen twice to 5 times a week, even on routes where there are multiple stopping higher speed trains over and above the regular mail and express trains. This would suggest that end to end non stop LD service is not that easy to sustain, unless of course it is a full blown HSR with end to end travel times almost competitive with air, and of course cheap enough.

The usual problem also is that irrespective of how many stops a train makes it takes a slot, and you have to ask yourself what is the best use of a slot when those are either expensive or hard to come by.

 

[user 1215]

The Duronto Expresses tend to have a higher max average speed, but there are many Rajdhani Expresses that have better average speed than some of the durontos. And correct me if I'm wrong, but I'm under the impression that most of the passenger lines on IR are double tracked with passing sidings at stations, like in Japan.

To have true, useful end-to-end trains in the US, you have to be able to efficiently leapfrog over the slower trains.

The primary Shinkansen between Tokyo and Shin-Osaka (340 miles) has three levels of service - the Kodama (local, stops at all 16 stations), the Hikari (express, stops at 8 +/- stations), and the Nozomi (super express makes 5 stops). Amazingly, the difference in time between the three services are roughly 234, 173 and 153 minutes at 87, 118, and 133 MPH average speed, respectively. This goes to show that as mentioned before, there is a LOT of time lost in station stops - in both the acceleration/deceleration as well as the stop itself (though the Shinkansen prides itself on generally being able to disembark and board passengers in one minute - something Amtrak can't seem to do...) But all this is on a dedicated ROW. Three levels of service on the same track - not shared with commuters or regionals.

Compare that to Acela. On the 457 miles between Boston and DC, there are multiple transit trains, regionals and of course the Acela. The NER between BOS and WAS has about 24 stops and takes approximately 480 minutes (57 MPH). The Acela makes around 12 stops and takes approximately 390 minutes (70 MPH).

So, the best Acela is really only 17 MPH average slower than the slowest of the bullet trains. I believe this is due to the congestion and sharing of the ROW (not to mention the fact that the Acela can't reach it full speed very often).

BUT I digress as none of this can happen efficiently INSIDE the well-oiled NEC, much less in the Freight world outside of the NEC.

 

[chakk]

In other words, a non-stop Superliner consist but only for passengers. Is there enough demand for a third Florida train, or perhaps some hypothetical system where the Star and Meteor ran opposite days, but this train ran every day?
I know WAS can handle Superliner equipment from the CL, but could it also handle a southbound Superliner train? If so, then it would probably be better to have the Auto-less Auto Train run right out of WAS so it could connect with the NEC.

A non-stop Superliner between the Northeast and Florida that carries only passengers would not draw enough business to justify the cost of operation. The Auto-Train is a success precisely because it allows people to bring their own autos with them. If no autos are carried, then the train endpoint would have to be populated with several rental car agencies (at least at the Florida stop). And the cost of the rental car might still be more than the incremental cost of bringing along your own auto. Plus, the benefit of bringing that auto stuffed to the gills with any stuff that you might want to bring with you on your trip to Florida -- for the snowbirds, etc.

 

[Thunder Road]

Wow, I didn't think this would trigger such debate. So the answer is obviously no, but I learned a bunch about the AT just by reading the thread. Sometimes I'm convinced this place knows more about Amtrak than Amtrak knows about Amtrak.

 

[user 1215]

Sometimes I'm convinced this place knows more about Amtrak than Amtrak knows about Amtrak.

I think we as a group sometimes think we know more about Amtrak than Amtrak knows. In fact, I would venture to say that there is a strong possibility that Amtrak is inconsistent enough that lots of opposing views here are BOTH right!

 

[oldtimer]

As a retired Amtrak employee with 35+ years Amtrak time married to an Amtrak employee with 33+ years service we have learned the the only thing you can count on at Amtrak is the inconsistencies!

:wacko: :wacko: :help: :wacko: :wacko: