Amtrak Success Stories

Suburban Station wrote:

mkellerm wrote: It is more likely to go the other way; more short-distance trips would produce both higher ridership and lower load factors. The scenario above (fewer seats HBG-PHL, lower ridership PIT-PHL) would produce both lower ridership and (probably) lower load factor.

let's say a seat is full for 104 miles out of 353m, then it's full nearly 30% of the trip (more if it's resold but let's assume it's not). if that seat is then moved to the longer distance bucket but isn't filled, it drops to 0%. no?

Correct, but then ridership drops from 1 to 0. The issue as I understood it was that load factor went down (from 72% to 69%) but ridership went up 5%. That is consistent with either (a) additional capacity, as suggested by Mr. Stolberg, or (b) a change in the mix of trip lengths in favor of shorter trips holding capacity constant, which I suggested as an alternative.

jamesinclair wrote:

jstolberg wrote: The California trains have lower load factors for two reasons.
1. They have unreserved seating. More trains have to run on a given route to make sure that there are enough seats for everyone, since passengers are not turned away in advance.

The load factors for the California trains are:
Pacific Surfliner 30%
Capitols 27%
San Joaquins 38%

Hm, I hope your math is wrong because those numbers seem shockingly low, especially the pacific surfliner which is supposedly one of amtraks best routes.

Yes, there is no reserved seating and you can pay cash on board, but that means load factors could be higher than 100%, since nobody is turned away.

Also, I wonder how amtrak accounts for the thruway buses. The san joaquin line has an early departure and a late departure which are both simply buses stopping at the train stations (6:00am from fresno southbound and 8pm from san francisco southbound)

Those load factors are correct. The issue is not exactly that the trains are unreserved (the San Joaquins are reserved), but rather that the trainsets make multiple trips a day and tend to be sized for maximum loads. You may have, for example, a trainset on the Capitol route that averages 250 passengers on its peak segment going in one direction (producing an load factor of say 60% for that train, since most riders don't ride endpoint to endpoint) and averages 25 passengers on its peak segment going the other direction (for a load factor of 6% on that train). That trainset has an overall load factor of 33%. There were some great stringline graphics showing ridership by segment for each train in California that I posted last year:

http://www.railroad.net/forums/viewtopi ... 46&t=58312

Take a look at them, and its easy to see why load factors are so low in California.

Very interesting charts. If you find that 2010 charts are made available, please post them here as well.

I find it very interesting that the San Joaquin ridership apparently has a very large college student base. Huge spikes at the end of august (start of semester), Thanksgiving, end of march (spring break) and end of may (end semester).

This can be seen on page 7
http://www.dot.ca.gov/hq/transprog/ctcl ... l_Info.pdf

Vincent wrote:I did some quick calculations of the operating costs per train mile of various routes and got some interesting numbers. Using the Amtrak timetable I calculated the number of train miles operated per month for several routes and divided that number into the operating cost number from the MPR. It looks like most routes are costing about $50/train mile, but there are some significant deviations from that number. The Hiawathas are way above (about $80/train mile) while other Chicago based trains (Wolverines, Blue Water and CHI-STL) are very near the average. The Piedmont and Downeaster are way below the average (about $29/train mile). Any suggestions for why there is so much deviation?

I am pretty sure that this reflects the fact that that Hiawathas bear a disproportionately high share of the costs of the Chicago Terminal. The new reports using APT (Amtrak Performance Tracking) are not as informative as the earlier reports using RPS (Route Profitability System), which broke costs into "FRA direct" and "Remaining direct" costs. The second category was essentially an allocated share of common operational costs (shared stations, supervision and training, yard ops, etc.). Those shared costs are allocated according to some combination of ridership and train starts, and the Hiawathas make up a high percentage of both categories in Chicago. If you look at the June 09 MPR, you'll see that the Hiawathas had FRA direct costs of 7.1M and the Wolverines had FRA direct costs of 12.7M. This makes sense; the Wolverines require three trainsets, the Hiawathas two; the Wolverines have food service cars, the Hiawathas don't. In the Remaining direct costs, however, the Hiawathas are charged $7.6M and the Wolverines only $7M. These (relatively) higher shared costs are what bump up the per-mile cost of the Hiawathas.

On the other hand, the Piedmont benefits from the fact that many of its expenses are off-book from the perspective of Amtrak, since NC owns and maintains the equipment and may provide some other services.

All of this shows how difficult it is going to be to implement the PRIIA requirement that states fully fund losses from non-NEC corridors under 750 miles. You are looking at states having to cough up $200M or more a year just to maintain current service levels under the current accounting scheme. That's clearly not going to happen, and it will be interesting to see how that is resolved.

jstolberg on June 5th wrote:The state of Pennsylvania applied for a $1.6 million planning grant to study the possibility of increasing service between Cleveland and Buffalo. http://www.prnewswire.com/news-releases ... 83244.html

To finally answer Mr. Benton's question as to why Pennsylvania was applying for a grant to study the lake shore line through Erie instead of Pittsburgh to Cleveland, perhaps it's because the State of Ohio is paying for the Pittsburgh to Cleveland study.

jstolberg wrote:To finally answer Mr. Benton's question as to why Pennsylvania was applying for a grant to study the lake shore line through Erie instead of Pittsburgh to Cleveland, perhaps it's because the State of Ohio is paying for the Pittsburgh to Cleveland study.

I believe PA came out in support of the Pitt-Cleveland (if not actual money). It's a study, and since money was available for it, why not throw a bone to the state's third largest city? PA's priority still has to be East-west for the time being. so long as it takes 7 and a half hours to travel between the state's two largest cities, there's a problem. does anyone know the status of the presumably underway Harrisburg-Pittsburgh study?

jstolberg wrote:North Carolina started a second Piedmont train running between Raleigh and Charlotte on June 5th, and the number of passengers didn’t double, it tripled! The new mid-day train didn’t steal passengers from the longer New York to Charlotte Carolinian either. The number of passengers on the already crowded Carolinian rose by 8% from June of last year to an average of 460 people per train. Having three trains per day makes the service much more convenient.

From your lips to Georgia and Florida's ears.

I've had to drive a couple of times recently from Fla to SC and back when I would have gladly paid Amtrak 2x or 4x my gas outlay to ride but Amtrak couldn't do the round trip without 2am layovers and 5am depart times. A usable scheduling is more valuable than all the slick pretty marketing in the world.

Matt Johnson wrote:Now, if they could just extend that Lynchburg train to Roanoke, I could actually use it to visit Virginia Tech! (Much easier to get to Blacksburg from Roanoke than from Lynchburg...)

Here is the latest on extending the Lynchburg service further west.

http://www2.wsls.com/sls/news/local/lyn ... ns/117026/


I live in Richmond and my son goes to Virginia Tech so I'd love to see this happen......and soon

electricron wrote:

Vincent wrote: It looks like most routes are costing about $50/train mile, but there are some significant deviations from that number. The Hiawathas are way above (about $80/train mile) while other Chicago based trains (Wolverines, Blue Water and CHI-STL) are very near the average. The Piedmont and Downeaster are way below the average (about $29/train mile). Any suggestions for why there is so much deviation?

There's far more train frequency for the Hiawatha trains sharing tracks part of the way with Metra trains, while the Piedmont runs mostly on NCRR owned tracks paying no track access fees. I can't fathom a reason off-hand why the Downeaster costs are so low, because they have high train frequencies and must pay trackage fees too? Maybe Amtrak's accounting tricks are at work again?

One reason the Downeaster's costs are lower may be that the route is divided between two owners. Of the 116 route-miles, about two thirds, Portland to Plaistow, NH, is on Pan Am Railways. The Downeaster does pay trackage fees for this part, but they may have been negotiated lower (don't quote me on this) since the Downeaster paid for all the track upgrades that Pan Am now enjoys, too. The remaining third, Plaistow to Boston North Station, is on MBTA-owned tracks. My understanding is that MBTA does not charge the DE for trackage or terminal costs, in part because the DE service complements what MBTA provides between Boston and Haverhill.

July is usually Amtrak’s best month. This July was spectacular.

The number of passengers riding Amtrak rose by 10% nationwide over July of last year. On long distance trains and state-sponsored corridor trains, the number of passengers rose by 12%. Revenues were up by 18%. The data is available at http://www.amtrak.com/servlet/BlobServe ... onthly.pdf Meanwhile, the price of gasoline was only 11 cents per gallon higher for the 4th of July weekend than last year. And while airline revenues for domestic travel rose 13% relative to last year, there were actually 1% fewer passengers in airline seats according to the Air Transport Association. Clearly, the American people are changing the way they travel.

Northern destinations proved popular during the hot month of July. Travel on the Vermonter was up by 42%. The Ethan Allen carried 17% more riders and the Adirondack to Montreal carried 25% more. The Adirondack carried an average of 247 passengers per train. Travel through upstate New York was up 37% on the Empire Service and Maple Leaf between Albany, Niagara Falls and Toronto, and the Lake Shore Limited carried 15% more riders. Travel northward through central Connecticut was up by 16%, and after the fireworks were over in Boston, all trains back to New York were sold out until July 6th.

In the southeast, the new Piedmont service continued to do well. With two Piedmont trains, ridership was up by 130% over last year, while the Carolinian carried 11% more passengers.

Travel on the new train to Lynchburg continued to be strong, at 188% of the pre-launch estimates. Travel on the Crescent rose too, with 14% more riders. The State of Virginia and the Norfolk Southern Railroad have just completed six projects to increase capacity and reduce travel time on the Crescent corridor. “Funded by $43 million from the Virginia Department of Rail and Public Transportation (DRPT) and nearly $20 million from Norfolk Southern, the projects lengthened or built new passing tracks between Manassas, Va., and Front Royal, installed five miles of double track near the Virginia Inland Port, improved signal and traffic control systems, and increased train speeds through Riverton Junction near Front Royal.” http://www.drpt.virginia.gov/news/details.aspx?id=512

In Michigan, travel on the Wolverine was up by 23% and the number of passengers on the Blue Water was up by 50%. They were busy trains with the Wolverine carrying an average of 291 passengers per trip and the Blue Water carrying 314. Amtrak worked to ease the crowding by adding an extra daily train between Chicago and Kalamazoo during the Labor Day weekend. http://www.nilesstar.com/2010/08/31/amt ... s-weekend/

In Illinois, travel on state-sponsored trains was at record levels. The Chicago-Quincy corridor was up by 18%, the Chicago-Carbondale corridor had 17% more passengers and the Chicago-St. Louis corridor had 16% more passengers. Construction has now begun between Springfield and Alton to replace the existing Union Pacific mainline.

Travel on the Missouri River Runner to Kansas City hit a new record, up by 26%, benefiting from capacity improvements and improved on-time performance that has increased the average effective speed on this line by 8 miles per hour since 2008.
The average speed on the Heartland Flyer from Oklahoma City to Ft. Worth is up by 4 miles per hour and ridership there is up by 43%. The state of Kansas would like to have some service in between, considering either extending the Heartland Flyer through the late evening and early morning hours arriving in Newton at 1:46 am northbound and leaving Newton at 4:20 am southbound, or running daytime trains leaving Kansas City at about 7 am and arriving in Ft. Worth at 7:30 pm with a return trip leaving Ft. Worth at 7 am and arriving at 7:30 pm. The first option would cost less and would connect the Southwest Chief (which carried 29% more passengers this July) with the Texas Eagle (which carried 14% more passengers http://www.ksn.com/news/local/story/Kan ... cRi3Q.cspx

Out on the west coast, ridership on the San Joaquin was up by 11% aided by the addition of a fifth car to some trains. The design of the next generation of California cars was approved by Amtrak and the Federal Railroad Administration on August 31st and will be able to run at speeds up to 125 mph. http://www.highspeed-rail.org/Documents ... elease.pdfRidership on the California Zephyr was up by 36% and first class tickets increased by 27%.

In the Northwest, the Cascades service carried 18% more passengers than last year. A Memorandum of Agreement signed September 3rd between Washington State DOT and the BNSF will enable some of the $590 million in stimulus money to be spent upgrading the BNSF track and increasing capacity so that the number of Cascades trains between Portland and Seattle can be increased from 4 to 6 trains per day. http://www.wsdot.wa.gov/News/2010/09/3_ ... eement.htm About a year ago, service between Seattle, Washington and Vancouver, British Columbia was increased from one train per day to two. That second train has exceeded expectations and the provincial government in Vancouver wants to double the service again to 4 trains per day. http://www.theprovince.com/sports/Daily ... z0xCqgWiPC However, the national government in Ottawa has not determined yet whether they will continue to waive the Canadian border crossing fees after September 30th and the schedule for October currently only shows one daily train. In the meantime, train tickets to Vancouver during the month of September are now selling at a 25% discount.

The San Joaquin just missed 100k, with 98,377. It might hit it next month, as the line appears to have a back to school bump (valley kids going to college in the cities?).

Still, 10,000 better than last year, shows that the central valley wants more rail. Hurry up with HSR already!

jstolberg wrote: Northern destinations proved popular during the hot month of July. Travel on the Vermonter was up by 42%. The Ethan Allen carried 17% more riders and the Adirondack to Montreal carried 25% more. The Adirondack carried an average of 247 passengers per train. Travel through upstate New York was up 37% on the Empire Service and Maple Leaf between Albany, Niagara Falls and Toronto, and the Lake Shore Limited carried 15% more riders. Travel northward through central Connecticut was up by 16%, and after the fireworks were over in Boston, all trains back to New York were sold out until July 6th.

There was no mention of the Downeaster. What would be the reasoning for that?

Adams_Umass_Boston wrote:There was no mention of the Downeaster. What would be the reasoning for that?

The Downeaster was the only regular Amtrak route — short- or long-distance — to see a decline this July versus July 2009. Ridership was down .6 percent; revenue was down .2 percent. The route even underperformed the anticipated results for revenue; "budget" for the Downeaster was a very small increase in revenue. However, the route still outperformed the ridership "budget," which was a decline of 5.6 percent.

Anybody want to hazard a guess why?

One possible explanation is poor OTP by the Downeaster which as memory serves was about 62% in July(?). The second possible explanation is a little more complex but also plausible.

The Downeaster's success in encouraging a modal shift by daily commuters and the occasional leisure traveler has already borne substantial fruit which was evidenced in ridership gains last year, that are now happenning elsewhere. In some senses the Downeaster in its present, 'five-a-day each way', format is a mature service that has not seen major changes for about two years. Some of the other services that are seeing performance increases were underperforming (LD routes), are new service (Lynchburg), are benefiting from route improvements (Midwest in general), or have added another frequency (Piedmont).

The fact that the Downeaster continues to hang on to ridership gains made last year in the face of substantially lower gas prices is something that NNEPRA is rightfully proud of. It is however still disturbing that other services at this time appear to be making gains that seem to elude the Downeaster. There is still plenty of room for debate as to whether or not the Downeaster simply made these gains earlier than its other Amtrak cousins. The LD routes in particular seem to be benefitting from Amtrak's corporate focus on product improvement which coincides with the most favorable conditions for rail travel in the United States in the past quarter century, or frankly, since A-Day. So in one sense perhaps the Downeaster was simply the leader and it is experiencing trends prior to other comparable Amtrak services. This theory still does not explain issues with prevailing economic conditions which one way or another should have a uniform impact. Conversely, this may also be an indication of regional economic differences.

There is one third possibility which would especially help explain ridership gains on LD routes. The U.S. traveling public may be in the process of undergoing a modal shift, where air travel once held an advantage for regional travel, now because of costs for fares, parking, additional fees, the time consuming nature of air travel both in terms of accessibility and security there are more and more passengers who are taking rail instead. The aging population may also help explain this change as these individuals are less capable of tolerating the strains of air and car travel. Although the Downeaster would certainly benefit in the long run from demographic changes the modal shift away from air travel is not as likely to benefit the service because they don't compete with air transportation.

The most likely explanation seems to be related to the inflated growth in ridership from last year. The year over year comparisons next year may be more telling as to the actual direction the service is taking. The OTP issues are still very troubling but anecdotal evidence still seems to point towards MBTA dispatching and schedule conflicts as being the primary issue. There is some track work being done that may help mitigate these issues in the near future.

I don't recall how many 90+ degree days there were in the NH/ME
area of the Downeaster route in July. PanAm puts a 40MPH
restriction in force every day that the temperature reaches 90.
This results in most of the Downeasters being very late for the
rest of the day, often an hour or more. This could be discouraging
passengers and cutting revenue.

There were quite a few 90+ days in August, so I would not be
surprised if the Downeasters took another hit in August.

One other issue of concern is the change in weekend consist
assignments. One trainset has five cars and the other has four.
During the week, the five car set handled #680 and #685, but
on the weekend, they put the five car set on #692 and #697, and
even with the five cars #692 quite often sold out. However, since
at least mid-August, they run the four car set on #692 and it sells
out the day before and passengers are turned away. There are
probably some logistics issues at Portland to swapping out the
train sets, but they used to do it. Some bean counter probably
decided they could save an hour or so crew time.