August 2013 Monthly Performance Report

[afigg]

The August 2013 Monthly Performance Report has been posted for those interested in such things. The ridership and revenue numbers were pretty good for the month with 2,851,979 passengers, a +3.0% ridership and +7.5% revenue increase over August 2012. There were not any notable changes in ridership trends from the previous months for the specific corridor or LD train services. It was a good month in that adjusted operating loss did not increase much in August, The projection for the end of the fiscal year is an adjusted loss of $365 million, $49 million less than the budget.

The LD trains had a good month with a +4.7% ridership increase which put them into a +0.4% increase for the year to date.

Amtrak is spending less on capital projects which they are attributed to delays in ramping up the projects. But there is this comment in the summary: “[SIZE=11pt]Cash balances are above budget as a result of the strong collection of receivables.” [/SIZE]Amtrak should start the next fiscal year with a good cash balance which will help Amtrak ride out a government shutdown.

I am going to just summarize the ridership growth (or decrease) breakdown by category for now. Others are welcome to post their own analysis.

August 2013 and Year To Date Ridership % and YTD Revenue % change

August YTD YTD Revenue
Acela: +1.0% -1.7% +4.5%
Regionals: -1.8% +0.9% +6.6%
Corridors: +5.1% +2.5% +4.7%
LD: +4.7% +0.4% +1.8%
System: +3.0% +1.3% +4.4%

 

[henryj]

Is August here already. lol.

 

[Train2104]

Things I see as compared to last August...

-Huge gains on the Lincoln Service corridor. Trackwork last year?

-10.8% gain in Keystone revenue even though ridership dropped 0.1%. How big of a fare increase did they implement?!

-Other large gainers: Michigan services (trackwork?), Pacific Surfliner (service improvements?), Pennsylvanian (?)

-Norfolk is eating into Newport News

-Ethan Allen ridership fell 5%. Not much, but it's the worst of the short distance routes.

-Large increases for Sunset Limited and Crescent.

-Cardinal ridership drop by 7%, while sleeper ridership increased 15%. Strange.

-In terms of lowest loss per passenger mile for LD's, Auto Train and EB are the best, Sunset and Cardinal are the worst. Those triweeklies really need major changes. I didn't expect the EB up there, I thought a Florida train would be there.

-Adirondack made a profit, though small.

-Hoosier State's loss per passenger mile dwarfs all the other routes, even the LD's. The state subsidies should help a lot.

-The Springfield Shuttles do worse financially than I thought

-Most minutes of delay for LD's: EB, CZ, TE, CS, SWC in that order. Somewhat expected given their length

-Most minutes of delay for corridors: Empire Service, Surfliners. Not too accurate because more trains naturally means more delay minutes.

Man, Amtrak implemented a lot of fare increases last year!

 

[afigg]

-Huge gains on the Lincoln Service corridor. Trackwork last year?

...

-Norfolk is eating into Newport News

...

-Cardinal ridership drop by 7%, while sleeper ridership increased 15%. Strange.

...

-Adirondack made a profit, though small.

...

Man, Amtrak implemented a lot of fare increases last year!

Yes, there was track work on the Lincoln service corridor last year as there is again this fall. There will be large swings in month to month ridership numbers for the Lincoln service and Wolverine corridors until the major disruptions for track upgrade work are wrapped up.
Norfolk and Newport News: yes, there has been an erosion in Washington-Newport ridership, but the FY12 number covered all 5 Virginia Regionals, 3 that terminated at RVR and the 2 that go to NPN. It appears that they assigned all the passengers from the RVR train(s) that was extended to NFK to the new Washington-Norfolk category, so there may not be that much reduction in passengers at NPN. There has been respectable overall growth however. In FY12 YTD, WAS-NPN was 575,532 passengers. In FY13 YTD, adding WAS-NPN 538,338 and WAS-NFK 117,127 equals 656,515 for a 14% increase.

Cardinal: In the ridership discussion section, there is this comment: "The Cardinals ridership was down 8% due in part to less coach capacity this year." Is the Cardinal running with one less coach car this year?

Adirondack "profit": have to be careful interpreting the Route Performance Report. The total revenue column for the state supported corridors includes the state subsidy payments since Amtrak treats those as revenue. Which will cause swings in route performance reports in FY14 for contribution/loss columns, depending on when the state payments were made. For FY13 YTD, the Adirondack had $6.5 million in ticket revenue while the total revenue in the route performance report is $13.8 million agsinst a total cost of $12.7 million.

Finally, oh, yea, Amtrak implemented a lot of fare increases this past year. I figure that many of the increases for the state corridor trains were to reduce losses so the states would not have to make as large a subsidy payment in FY14.

 

[Nathanael]

Others are welcome to post their own analysis.

UP is currently providing very good on-time performance, with the only noticeable problems being on the Sunset Limited / Texas Eagle.

None of the other hosts are providing good on-time performance.

The commuter & state-owned railroads:

- NM RailRunner is suffering from slow orders.

- So does the MBTA on the Downeaster route.

- Metrolink / Coaster seem to be dispatching their own trains ahead of Amtrak (well, they are on tighter schedules).

- So is Metra, and it has problems with slow orders.

- Same two problems for TriRail.

- Same two problems for Metro-North. Ouch. With a four-track railroad, they should be able to handle Amtrak better than that.

- Sunrail is still under construction, causing massive delays.

- The Michigan DOT railroad is under construction (as of August, still suffering the slow orders from before the purchase)

The Class Is:

- As noted, UP is dispatching Amtrak very well.

- BNSF has a lot of slow orders. Probably mostly due to current construction work, but they clearly don't want to maintain the Heartland Flyer line.

- CP track is suffering from severe slow orders. I thought ARRA money was supposed to fix that in NY...

- NS seems to be having problems almost entirely at the Chicago terminal end, which is getting hammered, as usual, with slow orders and freight train interference. Perhaps Englewood Flyover and the other projects in the area will help. Trains not heading to Chicago are doing well.

- CN seems to be dispatching freight trains ahead of passenger trains. There are also signal problems and slow orders.

- CSX is laden with slow orders everywhere, *and* is dispatching freight trains ahead of passenger trains. And there are signal problems. And this is systemwide. Yuck.

The shortlines:

- Buckingham Branch has severe "freight train interference". No surprise.

- NECR has slow orders (didn't we just have that line rebuilt with ARRA money? Maybe this is on the Massachusetts segment)

- Pan Am and Vermont Railway are both doing excellent dispatching of Amtrak.

Amtrak:

- It's worth noting that delays due to "passenger holds" are disproportionately high along the Empire Corridor, including the LSL. Some effort should be taken to reduce that. Hopefully the high-level platforms at the new Rochester station will help some. Dwell times might need to be increased.

Finally, the Empire Corridor is really being treated poorly. In addition to the previously-mentioned problems on both CSX and Metro North and the delays for "passenger holds", there are 11,000 minutes of host-responsible delays classfied as "other" (not freight train interference, not passenger train interference, not slow orders) on the Empire Corridor. Huh? This is way high. What is going *on* on the Empire Corridor?

 

[jis]

- Same two problems for Metro-North. Ouch. With a four-track railroad, they should be able to handle Amtrak better than that.

MNRR at present is effectively a two track railroad for about 5 miles (west of Bridgeport) and a three track railroad at its east end, slowly trying to become a four track railroad again.

Amtrak:

- It's worth noting that delays due to "passenger holds" are disproportionately high along the Empire Corridor, including the LSL. Some effort should be taken to reduce that. Hopefully the high-level platforms at the new Rochester station will help some. Dwell times might need to be increased.

High platform won't fix that particular problem since it is a pure operations problem involving conductors insisting on checking tickets before letting people board and also using only one door or on good days two doors to do so. What they need is a swift kick in their you know what to get them to start behaving like they are operating a corridor and not something in the middle of the sticks of Nebraska. Empire Corridor trains should be operating like NEC at least at HL platform stations, but even there they don't. Even at low level platforms they operate like they have the entire day available to them to board trains.
Amtrak needs to have a category calls "pointless Amtrak staff hold". I wonder if the "host responsible" delays is an euphemism for this odd behavior of Amtrak staff on that corridor, the host being Amtrak in the Hudson valley.

What happens is that they go through these inordinate delays - upto 10 minute stops to board 40 passengers, and then miss the slot and then try to blame the whole thing on CSX, or even themselves now that they don't have CSX to kick around between Poughkeepsie and Albany.

 

[henryj]

LD performance is looking the same. Total revenues 517.8 million, Total Operating costs 621.1 million for an estimated operating loss of 106.3 million or an 83% recovery which isn't too bad actually. Biggest losers are the CZ 27.6 mil, SWC 17.2 mil, Capitol 11.3 mil, Sunset 11.2 mil, Crescent 10.4 mil, CS 9.9 mil, LSL 9.3 mil, Star 8.1 mil, EB 7.1 mil, then you have the Meteor, Cardinal and Eagle at the 4 mil mark and the Palmetto and CONO break even and of course the Auto Train makes buckets of money at 19.4 mil. Of course Amtrak shows fully allocate costs at 1,058.5 billion and fully allocated loses at 543.7 million if you want to believe any of that. One more month to go in this fiscal year.

 

[WoodyinNYC]

The August 2013 Monthly Performance Report ... The projection for the end of the fiscal year

is an adjusted loss of $365 million, $49 million less than the budget.

Amtrak is spending less on capital projects ... But there is this comment in the summary:

Cash balances are above budget as a result of the strong collection of [/size]

receivables. Amtrak should start the next fiscal year with a good cash [/size]

balance which will help Amtrak ride out a government shutdown. [/size]

===========================================[/size]

Once it gets past the craziness of a government shutdown, then

it will have to get past the craziness of a government refusing

to pay its bills, and THEN Amtrak will have to face the craziness

of a Congress that can't pass a Transportation bill.

So to pay CAF for the desperately needed new passenger cars,

Amtrak will at least have some of this cash on hand. If that's not

enuff, well, we and CAF and Amtrak will probably just have to

wait until more price increases can squeeze out a few more millions

before the single-car order can be fully paid and the cars released

to service.

Oh, brother.

 

[Anderson]

I could repeat some of what afigg has said, but in general:

(1) I don't have a good read on WAS-NPN ridership changing. As it stands, the 14% increase in ridership on the combined ridership of the two lines allows enough for ridership to/from NPN to be close to flat while still adding Norfolk riders.
-There was a news report claiming a major drop in ridership on the Peninsula, but it also included this gem:

Amtrak estimates more than 320,000 passengers have taken trips out of the Peninsula station since last December 2012. That's down from the estimated 467,000 passengers during the same period the previous year.

467,000 passengers on the Peninsula? A drop of 147,000? Those numbers don't show up anywhere that I can find. I think someone just made up numbers...NPN didn't even have 147,000 riders to lose, and WAS-NPN hasn't been off by that amount at any point...and on the quote from Kevin Page, it's pretty hard to respond to off-the-wall numbers like that.
-As I think I've mentioned before, the last annual figures I have for 125/174 (and their weekend friends) was 64,973 for FY11. It was running 39.9% above that Oct-Jan in FY12, which would translate into about 90,869 for the full FY12 (+25,895 over FY11). For reference, WAS-NPN added 66,336 FY12 over FY11. If we assume that the trend on 174/125 held, you'd have 40,441 riders to allocate over the other four trains. That number feels reasonable to me. Also, if we look at FY13 thus far, per the DRPT numbers, through May there were 22,382 O/D out of 73,126 total on that train. That puts December-May RVR-north somewhere in the 50,000 range over about six months, or on course for about 100-110k for all of FY13 (mind you, this won't quite show in the final reports since Oct-Dec weren't listed as part of the NFK-WAS data). This would not be an unreasonable number to expect/hope for on that segment IMHO.
--I've got no way to determine what the situation "actually" is on my side of the James River until the station data spits out in a few weeks.

(2) The Springfield Shuttle situation is very complicated, and I believe due to some messy cost/revenue allocation (i.e. the Shuttle doesn't get any credit for passengers passing through beyond New Haven like the VA Regionals get) and due to the fact that the line is basically a commuter line (fares run $16-41 for 63 miles; the $16 special/$21 low-bucket fare is on par per-mile with the roughly 70 mile NYG-NHV one-way $20.50 peak fare). It also runs what? 2-3 cars long but still has a full operating crew? That inhibits making a profit.

(3) A number of trains have had substantial increases in fares over the last few years as part of the PRIIA 209 transition. While there's been hand-wringing over the required state support, my understanding is that Amtrak and several states had a sit-down-and-talk and decided to phase in some substantial fare hikes. This leads me to...

(4) In the case of the Surfliners, ridership took a massive hit after the 20-25% fare hike starting during FY11/12 (IIRC it was implemented over the summer of FY11). We're getting a bit of a bounce-back from that hit as folks get used to the higher fares as well as a resumption of normal-ish growth amid less draconian fare hikes.

(5) The loss per passenger mile situation is a bit wonky, but from what I can tell, with the Builder that lower number is due to the batches of through traffic spreading the losses out over lots of passenger-miles. Also not helping the Meteor's case, at least, is relatively empty space south of ORL.

(6) Finally, ridership through August FY13 was +366,675 over FY12. Ridership in September FY12 was 2,487,448. Unless we see a drop of at least 14.7% for September, we're going to have another annual record on our hands.
-Routes likely to be up for the year are a lot of the state corridors. On the LD side, the Cap, Sunset, Starlight, Sunset, and Palmetto are probably safe for a rise in ridership. The Chief, Zephyr, Meteor, CONO, Crescent, Eagle, and Auto Train are "on the bubble" but are close enough to break-even that I can't say for sure (a deficit of 2-3k can be made up with a good month); the Zephyr would be safe but for the disruption (which may scuttle a ridership increase). The Builder, Star, and Cardinal are likely off for the year.
--Takeaway: Unless we have a lousy September, we've got about 17,000 riders of "buffer" on the LD trains. We should be BARELY up for the year.

 

[afigg]

None of the other hosts are providing good on-time performance.

...

- So does the MBTA on the Downeaster route.

...

- Same two problems for Metro-North. Ouch. With a four-track railroad, they should be able to handle Amtrak better than that.

- Sunrail is still under construction, causing massive delays.

...

- NS seems to be having problems almost entirely at the Chicago terminal end, which is getting hammered, as usual, with slow orders and freight train interference. Perhaps Englewood Flyover and the other projects in the area will help. Trains not heading to Chicago are doing well.

...

- CSX is laden with slow orders everywhere, *and* is dispatching freight trains ahead of passenger trains. And there are signal problems. And this is systemwide. Yuck.

Good in-depth look at the extensive data in the OTP section! Many of the "slow order" delays in the August report are likely due to summer trackwork. Should plot the All Stations OTP by month over the course of several years. I expect the OTP will show a visible drop in the spring-fall track work season. As for some of the host railroads:
- There is a $20 million HSIPR project for bridge replacement and double tracking on the MBTA Haverhill line which is affecting the Downeaster. After years of reports of little visible progress, the pace of work picked up this summer with the bulk of the project supposed to be completed by this fall. That project may be the major reason for the MBTA delays for the Downeaster.

- As jis noted, the New Haven Line is currently a 2 track railroad over a 6-7 mile segment. And got hit with slow orders due to concerns about track maintenance in the aftermath of the derailment accident. The catenary and bridge replacement project that is causing the 2 track segment is supposed to be completed by early 2014. Then the catenary replacement project will move to shutting down 1 track at a time over the remaining old catenary segments. So the NHV Line is going to have changing 3 track segments through circa 2017. The MNRR delays should eventually improve, but the derailments, power failures have shown MNRR has a lot of catch up state of good repair work to do.

- Sunrail: Phase 1 service supposed to start next May so the track work for that segment should be completed before then. But there is a Phase 2 Sunrail extension to open by 2016, so there may be more Silver service delays in central FL in the years ahead.

-NS delays: Work was supposed to start this summer in Indiana on the $71 million IN Gateway track improvement project, but I have seen no reports on whether any track work has started or not.

There are a number of HSIPR and state funded track improvements projects that are to be completed over the next 3-4 years. Which should help reduce some of the delays. I can see where NS, BNSF, CP, MBTA, MNRR, BBRR get better on delays by then, but CSX only marginally so.

 

[Nathanael]

High platform won't fix that particular problem since it is a pure operations problem involving conductors insisting on checking tickets before letting people board and also using only one door or on good days two doors to do so.

At Syracuse they do board every door every car, pretty reliably.
I've seen the bizarre behavior you describe at Utica.

What they need is a swift kick in their you know what to get them to start behaving like they are operating a corridor and not something in the middle of the sticks of Nebraska. Empire Corridor trains should be operating like NEC at least at HL platform stations, but even there they don't. Even at low level platforms they operate like they have the entire day available to them to board trains.

Amtrak needs to have a category calls "pointless Amtrak staff hold". I wonder if the "host responsible" delays is an euphemism for this odd behavior of Amtrak staff on that corridor, the host being Amtrak in the Hudson valley.

That's a possibility.
It is notable that these inordinate "passenger hold" delays are only noticeable on the Empire Corridor. So it's a problem with a specific operating division -- you don't see the same delays on Chicago corridors or California corridors or even the Vermonter.

 

[afigg]

(5) The loss per passenger mile situation is a bit wonky, but from what I can tell, with the Builder that lower number is due to the batches of through traffic spreading the losses out over lots of passenger-miles. Also not helping the Meteor's case, at least, is relatively empty space south of ORL.

(6) Finally, ridership through August FY13 was +366,675 over FY12. Ridership in September FY12 was 2,487,448. Unless we see a drop of at least 14.7% for September, we're going to have another annual record on our hands.

-Routes likely to be up for the year are a lot of the state corridors. On the LD side, the Cap, Sunset, Starlight, Sunset, and Palmetto are probably safe for a rise in ridership. The Chief, Zephyr, Meteor, CONO, Crescent, Eagle, and Auto Train are "on the bubble" but are close enough to break-even that I can't say for sure (a deficit of 2-3k can be made up with a good month); the Zephyr would be safe but for the disruption (which may scuttle a ridership increase). The Builder, Star, and Cardinal are likely off for the year.

--Takeaway: Unless we have a lousy September, we've got about 17,000 riders of "buffer" on the LD trains. We should be BARELY up for the year.

Despite the power failure and NYP-BOS service disruption on the New Haven line, no reason to think that September won't be a growth month for ridership overall. I expect Amtrak will issue a press release by mid-October announcing another record setting year for ridership. With the loss in business from Sandy (~313K passengers according to the Oct & Nov reports) and the multiple disruptions on the NYP-BOS corridor by the Boston Marathon bombing manhunt, MNRR derailment & power failure, and the various routine weather related disruptions, a system increase of ~400K passengers is solid but steady growth.
Looking ahead to FY2014, growth on the NEC is likely to remain sluggish due to Amtrak's aggressive maximize revenue strategy, although it should get an artificial bounce if there are no Sandy level weather events and MNRR can avoid more shutdowns. The state corridor trains will continue to show the largest overall growth. Once they get pass the 10 day track work disruptions, the Lincoln service could show the largest steady growth over the next several years.

With regards to the Meteor, once they open the new Miami Central Station and the Silvers shift to the new station, the higher profile location in combination with the publicity about AAF, could boost both in-state and long range ridership from Miami. With Sunrail making more people aware of train service, the in-state traffic could get a bump for a couple of years until AAF starts operation. But the net benefits of the extra capacity from the Viewliners, new station in Miami, and SunRail likely won't take effect until FY15.

 

[Nathanael]

(2) The Springfield Shuttle situation is very complicated, and I believe due to some messy cost/revenue allocation (i.e. the Shuttle doesn't get any credit for passengers passing through beyond New Haven like the VA Regionals get) and due to the fact that the line is basically a commuter line (fares run $16-41 for 63 miles; the $16 special/$21 low-bucket fare is on par per-mile with the roughly 70 mile NYG-NHV one-way $20.50 peak fare). It also runs what? 2-3 cars long but still has a full operating crew? That inhibits making a profit.

Worth remembering: when "New Haven Hartford Springfield Rail Rail Project" starts operating (several years from now) I would expect the Springfield Shuttles to be replaced by the new commuter trains. It won't make much sense to run shuttles on the same route as the commuter train, with fewer stops, more crew, fewer seats, and different ticketing. Amtrak will probably retain the one "Shuttle" which runs through south of New Haven, and the Vermonter. The cities in Massachusetts along the "Knowledge Corridor" are agitating to get a second frequency in addition to the Vermonter, so I wouldn't be surprised if we end up with Amtrak extending the one remaining "Shuttle" north, and leaving the Springfield-New Haven local traffic entirely to the state of Connecticut.

 

[jis]

Actually it would be a choice for Connecticut (and not Amtrak's) to make as to whether they want to retain Amtrak as an operator of Amtrak branded trains on that route. If the operating agreements with SLE are the model for what will happen on NHV - SPG it may very well be that either way Amtrak remains the operator east of NHV on both SLE and SPG, irrespective of whether the service or part thereof is branded Amtrak or not.

 

[Anderson]

afigg,
I agree that full-year growth of around 400,000 would be good. Considering the awful luck the NEC has had, I think it's fair to estimate somewhere around 200-250k in lost ridership from the various incidents as a lowball in the extended NEC.

On the Empire corridor delays:
One possibility is that some delays from further up are "spitting out" at/beyond ALB and/or that at least some of the delays are a result of trains being out-of-slot. Remember, about half of the trains on that route are coming from beyond ALB, and two of those have a habit of getting stuck at the border. From what I can tell, there's not a dedicated code for "we were delayed by dispatch because we were already out of slot on a busy line".

Another possibility is that the traffic at a few stations along that line tends to be VERY directional (IIRC, 95%+ heading to/from NYP). That would certainly lead to issues with conductors checking tickets. For comparison, on Regionals down in VA, the procedure tends to be two doors: One for BC/quiet car (and one other coach) and one for other coaches, with coach traffic split between in-state traffic and those traveling past WAS. It may be hard to manage such a split on the Empire trains for some reason.

 

[Anderson]

I did have one other thought: The Star, Palmetto, and Crescent all spiked pretty hard in August (+7.8%, +7.8%, and +16.3%). All three had NEC sales opened up to them. It looks like somewhere between 5-8k trips shifted from the NE Regionals to the LD trains.

 

[jis]

Was there a corresponding reduction in ridership on Regionals? If not then the experiment apparently addressed a latent unsatisfied demand increasing overall ridership on the route. If so then it just off loaded some of the overcrowding on Regionals.

 

[Anderson]

There was a drop on the Regionals of 12k, but a lot of that was attributable to the PHL-BOS situation. Apparently, NYP-PHL was up while NYP-WAS was flat. I do find Amtrak mentioning the spike on the Star in the notes without mentioning this to be a bit sketchy.

 

[Nathanael]

Actually it would be a choice for Connecticut (and not Amtrak's) to make as to whether they want to retain Amtrak as an operator of Amtrak branded trains on that route. If the operating agreements with SLE are the model for what will happen on NHV - SPG it may very well be that either way Amtrak remains the operator east of NHV on both SLE and SPG, irrespective of whether the service or part thereof is branded Amtrak or not.

Quite likely.... but in that case it becomes a "contracted service". It goes off of Amtrak's books and disappears from the Monthly Performance Reports; Amtrak isn't bothered by the ridership or revenue numbers, and quite likely doesn't worry about finding rolling stock.

 

[Nathanael]

On the Empire corridor delays:

Another possibility is that the traffic at a few stations along that line

Most stations

tends to be VERY directional (IIRC, 95%+ heading to/from NYP). That would certainly lead to issues with conductors checking tickets. For comparison, on Regionals down in VA, the procedure tends to be two doors: One for BC/quiet car (and one other coach) and one for other coaches, with coach traffic split between in-state traffic and those traveling past WAS. It may be hard to manage such a split on the Empire trains for some reason.

Yeah; trying to do that would be really inefficient on the Empire Service. You have to use a more efficient "board all doors, check tickets in the train" procedure. Luckily the stations are spaced quite widely and there's time to do so. The crowds do mean that if you aren't boarding all doors at a high platform, it is going to take unduly long to board....

 

[Anderson]

Well, and the stats at those stations Hudson and Rhinecliff indicate you're probably looking at 100+ passengers boarding one or more of the morning trains (and disembarking on the evening trains). You've got just under 500 pax/day at each station on average, but you're probably looking at a strong bias towards weekdays that would put you at something more like 600/day instead. Eastbound traffic is going to concentrate towards 230/232/234 (arr. NYP 0730, 0826, and 0920 respectively); westbound traffic will likely concentrate towards 237, 239/293, and to a lesser extent 241 and 255/291 (dep. NYP at 1640, 1747, 1915, and 1515). Looking at the timing distribution, you've probably got 75-100 on each of the morning trains (and on the two main evening trains) for each of those stations.