Wednesday, April 25, 2012

Amtrak's Vermonter had 100% on-time performance in March

And it's not a lie!

Amtrak's Vermonter was 100% on time in March of 2012!!  Given murphy's law, that's pretty amazing!  The Ethan Allen did well too, at 85.5% on-time.  Averaged together, Amtrak trains in Vermont were 92.7% on-time.  The 12 month average for the Vermonter is 95.6%, which is pretty darn good.
This is big.  Worth a celebration!  (Worth a ride!)

On-time performance tops everything.  We'd like a faster train.  We'd like more frequent trains.  We'd like trains to Montreal and up the Western Corridor to Burlington.  But The only thing worse than a slow train is a late train!  We'd rather get there on the advertised.
This is not just me talking.  When the New England Central Railroad improved dispatching and track and the Vermonter started running on-time ridership started climbing and climbing and was up 30% over three years.  Now that the Vermont Rail System has improved the track of the Ethan Allen, ridership on that train has started climbing as well (up 9.9% in March).  (In Vermont as in most of the country, Amtrak operates as a tennent over railroads that are maintained and dispatched by freight railroads).
This achievement did not happen by accident.  It reflects the recent investment in track on both routes -- and the commitment of both of Vermont's freight railroads, the Agency of Transportation and Amtrak.  And thank you to the conductors and engineers who keep everything moving on-time every day!
The Vermonter is benefiting from the track work funded by stimulus funds.  Seventy five percent of the project in Vermont was done last summer; work on finishing the rest is now underway.  When it's done, and after a visit from the track geometry car (measuring to make sure everything is aligned), then the speeds in Vermont will be increased in time for the fall timetable change which will see 27 minutes removed from the schedule within Vermont.

Looking through the Amtrak Status Maps archive, it's actually reasonably accurate even with the potential for misleading; at 611 route-miles it's only considered late if it is 30 minutes or more past the scheduled arrival time rather than 10 minutes for most corridor trains, such as the Surfliner. 55/57 were always on time and averaged 2 minutes early, often eight minutes early (due to padding at the end). Latest arrival in Washington, D.C. was on March 29, 7 minutes late. 56/54 had three days exceeding 10 minutes, March 10th, 14th, and 24th for 17, 20, and 12 minutes respectively. However, thanks to the 12 minute padding it was, on average, 2.5 minutes early. All-point OTP was reasonably good as well, though I'm certainly not going to calculate that.

Amtrak's actually seen a substantial improvement in on-time performance lately, though this may be due to the rather mild winter not posing much of a problem for routes vulnerable to weather disruptions.  This could, however, pose issues for Amtrak's retention of contract in the future on routes where it has not been improving and has, in fact, been declining.

Tuesday, April 24, 2012

MBTA rolling out smartphone ticketing

Rolling out this fall:

Boston rail commuters will soon have a mobile alternative to traditional paper tickets, allowing them to use their smartphones to buy and display their train tickets. The Massachusetts Bay Transportation Authority (MBTA), which serves 1.3 million people a day, will launch the U.S.’s first smartphone rail ticketing system this fall through a partnership with Masabi, a London company which has been rolling out mobile ticketing services in the UK.
Users will be able to buy their tickets and passes via their smartphone instead of lining up at vending machines. When it comes time to show their proof of purchase, the smartphone application will display an animated watermark with a background color that changes according to the day. The app, which will work on iPhone, Android and BlackBerry devices, also produces a bar code that can be scanned for closer inspection. Smartphone-equipped train conductors will be able to do quick visual inspections of the tickets or scan the tickets to be sure.
The system will be tested this summer with a pilot group before a full launch this fall. It can also work in conjunction with the MBTA’s contactless CharlieCard, allowing monthly pass holders to link their cards to their account and charge them up through their smartphone. The move to mobile ticketing should speed up the ticketing process commuters, many of whom are forced to buy tickets on the trains because there are no vending machines at their stations. It also allows the MBTA to not have to buy or maintain additional vending machines and lowers the cost of handling cash.

Of course, as those who have been following this would know, the article is wrong when it refers to MBTA being the first to have a smartphone rail ticketing system this fall; Amtrak has been rolling it out this year with the Capitol Corridor and San Joaquins recently switching over to e-ticketing and use of smartphones. They would, however, be the first commuter rail system to do so. I am astounded, however, that the initial press release mentioned that less than half of MBTA's stations possess ticket vending machines. Sure, putting a pair in any individual station would cost about a hundred thousand dollars, but that's still capping out at about ten million dollars systemwide to emplace them. By removing the need for conductors to ever sell tickets, MBTA could reduce to a single conductor per train, with the reduced operating costs easily paying for the capital expense (this may require some platform reconstruction at certain stations as well).

Monday, April 23, 2012

Super-commuters and rail

Courtesy of LA Metro comes word of an interesting study on an interesting emerging phenomena: "Super-commuters
The twenty-first century is emerging as the century of the “super-commuter,” a person who works in the central county of a given metropolitan area, but lives beyond the boundaries of that metropolitan area, commuting long distance by air, rail, car, bus, or a combination of modes. The super-commuter typically travels once or twice weekly for work, and is a rapidly growing part of our workforce. The changing structure of the workplace, advances in telecommunications, and the global pattern of economic life have made the super-commuter a new force in transportation.
These "super-commuters" represent an interesting market which should be highly amenable to incorporating rail into their infrequent commutes and a shortcut method of identifying rail corridor which should be prioritized for increased rail service.

Somewhat surprisingly, Houston has both the greatest number of super-commuters and the greatest percentage growth in 2002-2009 with a total of 251,200 super-commuters in Harris County, almost doubling its initial amount. Coming for the most part from Dallas-Fort Worth, Austin, and San Antonio (collectively, 47% of the super-commuters), and with Dallas itself holding a substantial number of commuters from these cities, it seems obvious that a Texas Triangle high speed rail system should attract a very high level of ridership. Certainly it would seem to justify the approximately 1.25 billion dollar expense of an intercity passenger rail line between Austin and Houston.

Los Angeles comes second for super-commuters with 233,000 and a 76.7% increase and, rather interestingly, all but the very last location, El Centro, already possess a developed intercity rail service to Los Angeles or will possess it with the introduction of the California high speed rail system. Unsurprisingly, San Diego leads with 78,300 such super-commuters but Santa Barbara and San Luis Obispo contribute another 10,500 and 5,800 respectively. These numbers indicate that, although the high speed rail project will of necessity consume the large majority of rail funds in and for California, substantial investment should still be made with the Pacific Surfliner to reduce travel times and increase reliability of the service, which would, in turn increase the patronage of the high speed rail system when it arrives.

Somewhat surprisingly with Los Angeles, in my opinion, is the sheer number of commuters from Northern California, over 60,000 of them. Predominately they originate in San Francisco, 35,700, but San Jose and Sacramento contribute another 12,500 and 10,400 respectively. Given that the study indicates these super-commuters are more likely to be middle-class, with earnings under $40,000 annually, it's highly probable that a cheaper, yet still time and productivity competitive, rail option will succeed in becoming the preferential mode of travel.

Friday, April 20, 2012

CAHSR's claimed emissions reductions have been lowered

Brad Plumer recently posted on how high speed rail is not the most efficient way to cut carbon emissions, something I'd previously noted. Unfortunately, it actually overestimates the cost-effectiveness of CAHSR as a means of reducing CO2 because the Authority has not updated their website to reflect revisions to the plan. The new business plan estimates only a three million short ton annual reduction of CO2 emissions, doubling the cost per pound of CO2 in his estimates to $500 per ton.

Furthermore, I do not trust that number. Applied per rider, at the high end estimate of 24.2 million passengers in 2029, it's about 250 pounds of CO2 reduced per passenger which would equate to an avoidance of 272 miles driven per passenger using EPA estimates if every single rider were to have been diverted from driving. Unfortunately for such estimates, a substantial number would be diverted from air travel and according to the International Civil Aviation Organization only 160 pounds of CO2 would be avoided per passenger in such a case (LAX-SFO, economy class). Induced travel, that is, passengers who would not otherwise be making the journey, might actually be considered carbon-negative thanks to emissions caused by their travel to and from the rail system; certainly they cannot be considered to be the source of an emissions reduction, making the estimates further outlandish.

Wednesday, April 18, 2012

Los Angeles Metro: Steampunk before there was punk

While looking for something else, I came across this rather fascinating article, with film, about Southern California Rapid Transit District's (the predecessor to LA Metro) experiment with steam-powered buses in an attempt to cut down on air pollution.  Obviously they didn't go into fleet use, but it's still a rather interesting bit of history I'd never heard of before.

Tuesday, April 17, 2012

All Aboard Florida is predicting ridership in excess of 3 million

An interesting tidbit from a recent PR piece they've released:

Beyond the financial benefits, All Aboard Florida passenger rail also promises long-term environmental benefits. Giving frequent business travelers and families the option to choose an energy-efficient rail service over often-congested roadways and airports lowers the state’s overall carbon footprint.  Initial studies predict this new rail service will take more than three million cars off Florida’s roadways annually.

Three million cars off the road would mean just north of 3.5 million or so passengers assuming an average occupancy of 1.2 per vehicle, which is rather interesting in light of my previous attempt to estimate the ridership they'd need to break even.

I do have to say that the paragraph which immediately followed the above does not fill me with confidence:

All Aboard Florida will select equipment utilizing clean diesel engines that meet stringent Tier 3 federal emission standards.  Clean diesel systems are both fuel efficient and environmentally friendly.  Clean diesel-powered vehicles produce up to 25 percent less CO2 emissions than that of comparable gasoline engines, resulting in better air quality, and 30 percent better fuel economy, resulting in a smaller environmental footprint.
Look like the PR person did a bit of context-free copypasta with that diesel vs gasoline comparison.

Saturday, April 14, 2012

More HSR operating surpluses

The collection of high speed rail operating surpluses has been updated with Spanish figures for 2009-2010 (with the caveat that they are for the entire Renfe passenger system due to a merging of departments) as well as SNCF and DB Bahn for 2011.

Reworking the Starlight for the Daylight

For better than a decade now, California has been trying to bring back the old Coast Daylight as an extension of the current Surfliner from San Luis Obispo to San Francisco (via Oakland or Emeryville). This would provide convenient one-seat day travel between Los Angeles and the Bay Area, which is currently not satisfied by either the Coast Starlight (arriving as it does, at 10pm or later if behind schedule) or the bus connection between Los Angeles and the San Joaquins at Bakersfield.

The current schedule anticipation is for departing/arriving at ~7:30am/pm. With only a three hour gap in departure times from Los Angeles, the number of coach passengers will be fairly minimal, resulting in higher losses on the route, as the vast majority of current California intrastate defects to the cheaper and more convenient option. With a cost recovery of only 49% (page 14), any revenue loss puts it in rather poor territory and more at risk for cutting of the route. This raises the question of how to prevent a reduction of cost recovery, and ideally improve it, for which I see three possible solutions.

The first, and worst, is to simply truncate the route at Emeryville using the Daylight as a connection to Los Angeles. This avoids the expenses inherent to a poorly patronized train service, but also cuts most current connections (retaining only the connection with the thrice weekly Sunset Limited) and loses the added revenue of sleeper car patrons and and likely the loss of some ridership that extends beyond Emeryville and does not wish to make a transfer.

The second solution would be to attach some sleepers to the Coast Daylight and switch the cars in Emeryville onto a truncated Starlight, as is done currently in San Antonio between the Sunset Limited and Texas Eagle. This does retain most of the business and ridership, but would require a degree of schedule realignment on the southbound leg in order to connect with a departing Daylight.

The third solution, and in my opinion the best, is to reschedule the entire Starlight so that it runs overnight between Los Angeles and San Francisco as a true double frequency of the Daylight. Not only does this reduce "passenger poaching"but it also opens the train to a new market of those who are open to traveling to San Francisco or Los Angeles by rail, but do not wish to spend the entire day doing so. Importantly from a cost recovery perspective, members of this market may be more willing to pay extra for base fares and more likely to upgrade to sleepers which can greatly increase revenue (even a roomette will increase revenue by 50% if shared and double or treble it if occupied by only a single person).

If the latter is true, and Amtrak is able to capitalize with additional sleepers based on demand (certainly a possibility if they join the mass railcar order California and the Midwest states are putting together), the results could very well be greatly beneficial for overall route performance and Amtrak's bottom line. Positing a northbound 8.5-10.5 hour schedule pushback, for a Los Angeles departure between 7-9pm results in the following arrival times: San Jose 5-7am, Emeryville 6-8am, Sacramento 8:30-10:30am,  Klamath Falls 5-7pm, Eugene (terminus of Cascades) 9:15-11:15pm, Portland 11:20pm-1:20am, Seattle 6-8am.

Heading south, for an arrival of 7-9am arrival time in Los Angeles, the schedule would need to be adjusted by 10-12 hours, leading to the following departures: Seattle 7:45-9:45pm, Portland 12:30am-2:30am, Eugene 3:10-5:10am, Klamath Falls 8-10am, Sacramento 4:30-6:30pm, Emeryville 6:30-8:30pm, San Jose 8-10pm.

Under such a schedule, the current connections to the Empire Builder at Seattle and Portland are broken, but a connection is made with both the arriving and departing California Zephyr adding Denver and Salt Lake City to the list of destinations easily accessible from Southern California solely by rail without significant loss. The connection with the Sunset Limited is also broken with the Starlight, but is replaced by the Daylight, which should result in no significant change.

Portland does not fair well with the midnight trains compared to its current scheduling, however increased Cascade service should fill in the gap for what is, in any event, an off-peak travel time train. Any losses with Portland, however, should be more than made up for by the fact that the entirety of California, from Dunsmuir to San Diego, now lies within reasonable travel hours, with hopefully an increase in travel numbers.

Since this would be at the behest of the state of California, if Amtrak should prove reluctant to change the schedule for fear of greater drops in revenue than remaining on the current schedule with the Daylight added, it would be appropriate for California to pick up the tab, since the schedule change would mostly be to California's benefit and 55% of the route lies within California. This would most likely take the form of crediting increased revenue originating or ending in California against revenue losses, with the state making up any remainder.

Friday, April 13, 2012

Tea Party endorses All Aboard Florida

All Aboard Florida updated with a small list of endorsements today but only one really caught my eye: TEA Party Miami has endorsed their endeavor.

CAHSR turns into a cat about Anaheim

It's in the plan, then it's out of the plan, then it's in the plan...will they just make up their minds?

The state's High Speed Rail Authority board on Thursday approved a new business plan for the bullet train route with one notable amendment – adding Anaheim back on as the route's southern terminus.
Anaheim Councilwoman Kris Murray testified before the rail authority in Sacramento advocating for the trains to travel all the way to Anaheim. After the meeting, she said she's thrilled to see Anaheim back in the project's initial phase, although the Los Angeles-to-Anaheim leg of the journey might not be at top speed.
"It's great news for us that Anaheim is back in the plan as a one-ticket ride all the way from San Francisco," Murray said. "It's great news that the rail authority recognized the importance not only Anaheim but the Orange County region.
Since it was publicized last week that the initial phase of high-speed rail would run from San Francisco to Los Angeles, stopping short of Anaheim to save on costs, the rail authority was flooded with calls and letters asking that Anaheim be put back in the first phase.
A letter from Lucy Dunn at the Orange County Business Council seemed to make a strong impact with the Rail Authority board, which voted unanimously to put Anaheim back in the plan, Murray said.
The rail agency had been looking at ways to cut the cost of the project, which had soared to $100 billion at one point. The current plan, detailed in a 212-page document, is estimated to cost about $68 billion, and could whisk passengers on the 520-mile route at up to 220 mph.
Murray said decisions about "electrifying" the L.A. to Anaheim segment of the line – needed for the train to run at full speed -- will be made at a later date. But the plan, as drafted, would allow passengers to travel from San Francisco all the way to Anaheim with one ticket – and without having to change trains in Los Angeles' Union Station, she said.

I'm honestly curious as to how they planned on actually deleting Anaheim from the system given that it is mandated as part of Phase I by law.

Now, presuming that they can get the appropriate waivers to share track with active freight operations, extension to Anaheim isn't very hard at all. It would be about 270-300 million dollars to string overhead catenary from Los Angeles to Anaheim, or another 640 million if you want to charge the Los Angeles Union Station through tracks to Anaheim as well. 

This does impose limitations on the number of trains that they can send through to Anaheim, of course, but unless they were going to have all trains originate/terminate in Anaheim, three trains per hour per direction, the HSR capacity under the shared track plan, was excessive and the dedicated plan's five was simply ludicrous. Browsing online, even Paris-Lyons tops out at 2 trains per hour, though this may be artificially low due to extremely high track fees per train on that line. Even without other improvements, Los Angeles-Anaheim should be able to handle two HSR trains per hour peak with the Los Angeles-Fullerton corridor finishing its triple track project this year (currently, including the 91 line, the peak use is 5tph, from 4:10-5:10pm, 3 Metrolink and 2 Amtrak) as well as the LAUS through tracks.

Until such time as Phase II connects to Sacramento, two trains per hour should be adequate capacity for Los Angeles to Anaheim. I do not consider Phase II to San Diego to require additional capacity for the simple reason that I do not expect any trains to actually make the journey from Anaheim to San Diego; it would be faster for passengers to take the Surfliner south along the LOSSAN route (admittedly, I am assuming that in the decades until such a Phase II connection exists, there would be 20 minutes worth of improvements made, but I feel that is a safe assumption).

Thursday, April 12, 2012

A-Day vs Today

BBnet3000 asked about seeing "Then and Now" figures for the rest of Amtrak and since I had planned on going through and doing so anyhow, here they are. As before, these are taken from Amtrak's original timetable and There are no adjustments or allowances made for additional stops, different station locations (though I've tried to make note of that where it occurs), or route changes. Lastly, this does not reflect additional or lesser frequencies or improvements to slower services on the same route (such as the Northeast Regionals).

Boston-New York:
1971: 3 hours, 48 minutes #3001 Turboservice
2012: 3 hours, 25 minutes #2163 Acela Express
Variance: -23 minutes/10.1%

New York-Springfield:
1971: 3 hours, 5 minutes #184-078
2012: 3 hours, 10 minutes #495 Northeast Regional/Shuttle
Variance: +5 minutes/2.7%

New York-Washington, D.C.:
1971: 2 hours, 59 minutes Metroliners (most scheduled)
2012: 2 hours, 42 minutes #2150 Acela Express
Variance: -17 minutes/9.5%

1971: 1 hour, 39 minutes #605, #614
2012: 1 hour, 37 minutes #607 Keystone
Variance: -2 minutes/2%

New York-Albany:
1971: 2 hours, 40 minutes all trains (Grand Central Station)
2012: 2 hours, 20 minutes #234, #237 Empire Service
Variance: -20 minutes/12.5%

New York-Buffalo:
1971: 7 hours, 30 minutes all trains (Buffalo Central)
2012: 7 hours, 55 minutes #63 Maple Leaf (Buffalo-Depew)
Variance: +25 minutes/5.6%

New York-Pittsburgh:
1971: 8 hours, 34 minutes #48
2012: 9 hours, 13 minutes #43 Pennsylvanian
Variance: +39 minutes/7.6%

New York-Chicago:
1971: 17 hours, 5 minutes #49
2012: 19 hours #49 Lake Shore Limited
Variance: +115 minutes/11.2%

1971: 8 hours, 18 minutes #49
2012: 9 hours, 25 minutes #30 Capitol Limited
Variance: +67 minutes/13.5%

New York-New Orleans:
1971: 28 hours, 8 minutes #2-126
2012: 29 hours, 48 minutes #20 Crescent
Variance: +100 minutes/5.9%

New York-Orlando
1971: 21 hours, 10 minutes #145-91
2012: 21 hours, 31 minutes #98 Silver Meteor
Variance: +21 minutes/1.7%

New York-Tampa
1971: 24 hours, 20 minutes #135-21
2012: 25 hours, 32 minutes #91 Silver Star
Variance: +72 minutes/4.9%

New York-Miami
1971: 25 hours, 15 minutes #58-142
2012: 26 hours, 46 minutes #98 Silver Meteor
Variance: +91 minutes/6%

Boston-Washington, D.C.:
1971: 7 hours, 4 minutes #108-3002, 3001-107 Turboservice/Metroliner
2012: 6 hours, 30 minutes Acela Express (several)
Variance: -34 minutes/8%

1971: 17 hours #549-49
2012: 17 hours, 30 minutes #30 Capitol Limited
Variance: +30 minutes/2.9%

1971: 6 hours, 10 minutes #304
2012: 8 hours, 32 minutes #50 Cardinal
Variance: +142 minutes/38.4%

1971: 5 hours, 15 minutes #9, #10
2012: 5 hours, 21 minutes #59 City of New Orleans
Variance: +6 minutes/1.9%

1971: 5 hours, 50 minutes #356
2012: 5 hours, 30 minutes #352, #354 Wolverine
Variance: -20 minutes/5.7%

Chicago-St. Louis:
1971: 5 hours, 15 minutes #2
2012: 5 hours, 20 minutes #301 Lincoln Service
Variance: +5 minutes/1.6%

1971: 1 hour, 30 minutes all trains except #31/32 with baggage (10 extra minutes)
2012: 1 hour, 29 minutes all trains
Variance: -1 minute/1.1%

Chicago-New Orleans:
1971: 17 hours, 30 minutes #1
2012: 19 hours, 15 minutes #58 City of New Orleans
Variance: +105 minutes/10%

Chicago-Los Angeles:
1971: 40 hours #18
2012: 43 hours #4 Southwest Chief
Variance: +180 minutes/7.5%

1971: 49 hours, 35 minutes #18-2
2012: 51 hours, 40 minutes #6 California Zephyr (Emeryville)
Variance: +125 minutes/4.2%

1971: 46 hours, 50 minutes #32
2012: 45 hours, 15 minutes #8 Empire Builder
Variance: -95 minutes/3.4%

1971: 7 hours, 55 minutes #31/32
2012: 8 hours, 5 minutes #8 Empire Builder
Variance: +10 minutes/2.1%

Los Angeles-New Orleans
1971: 44 hours #2
2012: 45 hours, 40 minutes #2 Sunset Limited
Variance: +100 minutes/3.8%

Los Angeles-Seattle
1971: 31 hours #11/12
2012: 29 hours, 25 minutes #14 Coast Starlight
Variance: -35 minutes/1.9%

1971: 3 hours, 30 minutes #195/196/198/199 (Tacoma and Vancouver only intermediate stops)
2012: 3 hours, 30 minutes all Cascades
Variance: 0 minutes/0%

Tuesday, April 10, 2012

Amtrak Cascades to receive unified management

Passenger rail service on ideal track:

Faster, more reliable passenger rail service between Vancouver, B.C., and Eugene, Ore., is one of the likely outcomes of a new partnership between the Washington and Oregon transportation departments.
The two state agencies have agreed to manage the route for Amtrak Cascades service as one continuous rail corridor, instead of two separate state operations.
This is a welcome move that should eventually lead to improved service for a growing customer base of rail passengers.
The goal is to have a unified plan for governance, scheduling, budgeting, capital planning and fleet management. This should lead to increased efficiencies and cost savings.
As gasoline prices continue to soar, more people are turning to passenger rail service as a viable transportation alternative. Amtrak Cascades ridership in 2011 totalled 850,000 passengers, which is a 51 percent increase in the past 10 years.

It would be interesting to see if they expand this further into a unified operation with British Columbia as well, though that would come with certain issues. I'm not personally aware of how this would immediately affect any plans or other issues with the rail service, other than reduction of redundancies. One potential, however, is a joint selection for future operators in accordance with the Passenger Rail Improvement and Investment Act.

It does point to an interesting trend in centralization amongst various regional groups however, with LOSSAN, the San Joaquins, the Northern California Unified Service, and now this being bandied about within the past year.

GAO: NJ governor was less than truthful about ARC cancellation

From the New York Times:

The report by the Government Accountability Office, to be released this week, found that while Mr. Christie said that state transportation officials had revised cost estimates for the tunnel to at least $11 billion and potentially more than $14 billion, the range of estimates had in fact remained unchanged in the two years before he announced in 2010 that he was shutting down the project. And state transportation officials, the report says, had said the cost would be no more than $10 billion.
Mr. Christie also misstated New Jersey’s share of the costs: he said the state would pay 70 percent of the project; the report found that New Jersey was paying 14.4 percent. And while the governor said that an agreement with the federal government would require the state to pay all cost overruns, the report found that there was no final agreement, and that the federal government had made several offers to share those costs.
A spokesman for the governor, Michael Drewniak, said Mr. Christie’s statement of costs had included $775 million to build a new portal bridge, which was required as part of the project. The 70 percent state share, he said, included the costs that would have been paid for by the Port Authority of New York and New Jersey, which is run by both states, as well as federal highway and stimulus funds earmarked for New Jersey. Counting those costs, which the report does not do, would put the state’s share at 65.5 percent.

The GAO report itself.

Monday, April 9, 2012

The problem with grade separations

Rail crossings are a real annoyance for those sitting at them, especially when one is unfortunate to get stuck behind a mile long train going only a few tens of miles per hour. Such prolonged stops can easily introduce produce major traffic snarls, extending the misery, although a passenger train will not produce a delay longer than a typical 90 second red light. Additionally, there safety issues whenever a crossing between two directions of traffic occurs, resulting in the occasional strikes of vehicles or pedestrians by trains, leading not only to deaths of the trespassers, but also delays to passengers aboard the train as well as passenger fatalities in extreme situations. Conceptually, at least, separating rail lines from roads where they currently cross is a good idea.

However, the problem is that these are extremely expensive and generally paid for by rail funds.

Consider, for instance, OCTA's current grade separation projects. Nearly 600 million dollars buys only seven grade crossings with minimal benefit to rail operations. While it does prevent the occasional delay from vehicular and pedestrian strikes, the benefit as seen by rail travelers is minimal to nonexistent. To make matters worse, that sum could easily pay for the electrification of the rail line from Los Angeles Union Station to Laguna Niguel/Mission Viejo Metrolink station with all the accompanying benefits of electrification, including significantly faster acceleration and average travel times, as well as agency benefits from the resulting higher ridership and fare figures and lower operating costs.

Thanks to the positive train control mandate, however, there is the potential to eliminate safety issues arising from most vehicular strikes. With readily available technology, one can use a variety of means, from radar to inductive loops, to determine whether a particular crossing is blocked or free of vehicular traffic. In addition to raising gates to permit the egress of offending vehicles, as Amtrak's four-quadrant gates in Connecticut do, a temporary "mini-block" may be overlaid upon the normal block with the gate crossing transmitting its normal information to the system. With this mini-block system, the engineer is easily notified as to whether any problems such as an obstructing vehicle await at the crossing and may take appropriate action, or even have the train automatically brake in response to a sudden change in situation. With the use of additional safety measures such as medians and quad-gates, this reduces the risk of vehicular strikes to last-second ramming through the gates, for which additional safety measures are available. This additionally prevents accidents arising from human error in crossing repair. To my knowledge, the positive train control system used in Michigan, ITCS, operates in this fashion.

Delays from suicidal or terminally oblivious pedestrians will not, of course, be affected by such a system.

Doing this does require that a train be able to stop in time, of course. FRA regulations require that lights are flashing for a minimum of 20 seconds prior to the arrival of a train, which comes out to a distance of 2,317 feet for a train traveling at 79 miles per hour. Courtesy of this document*, page 21, we have an equation for determining the stopping distance in Imperial units, provided we know the braking rate, which can be estimated for the FLIRT at approximately 2.8 mphps. For our 79 mile per hour train, that comes out to 1,634.5 feet, allowing for a a shade under 6 seconds for complete clearance of the tracks prior the train necessitating a halt. Of course, stopping distance may take longer depending upon equipment and this is only the minimum warning time, with many signals providing a much greater warning period. The point is not to establish a hard and fast rule about how it ought to be done, but merely to demonstrate that it can be done and provide for rail nearly all the safety benefits of grade separation at a fraction of the cost.

*Worth reading if purely for the amusement factor of the Department of Defense spending money to definitively say "No, the TGV is not suitable for hauling M1 tanks."

Saturday, April 7, 2012

Then and now

Los Angeles-Santa Barbara: 2 hours, 14 minutes Southern Pacific fastest scheduled time (#98)
Los Angeles San Luis Obispo: 4 hours, 32 minutes Southern Pacific fastest scheduled time (#98)

1964 (via)
Los Angeles-Santa Barbara: 2 hours, 15 minutes Southern Pacific fastest scheduled time (#99)
Los Angeles-San Luis Obispo: 4 hours, 33 minutes Southern Pacific fastest scheduled time  (#99)
Los Angeles-San Diego: 2 hours, 40 minutes ATSF fastest scheduled time (#79)

1971 A-Day

Los Angeles-Santa Barbara: 2 hours, 13 minutes Amtrak fastest scheduled time (#98, #12)
Los Angeles-San Luis Obispo: 4 hours, 30 minutes Amtrak fastest scheduled time (#98, #12)
Los Angeles-San Diego: 2 hours 45 minutes Amtrak fastest scheduled time (#11, #12)

Los Angeles-Santa Barbara: 2 hours, 23 minutes Amtrak fastest scheduled time (#14)
Los Angeles-San Luis Obispo: 5 hours, 5 minutes Amtrak fastest scheduled time (#14)
Los Angeles-San Diego: 2 hours 28 minutes Amtrak fastest scheduled time

Now, to give Amtrak some credit, some of the delay is due to extra stops and there has been a major increase in frequencies as well. Additionally, this may ignore periods of faster running. For instance, while today, only the oft delayed Pacific Surfliner Express ran San Diego-Los Angeles in 2 hours 28 minutes, with the remaining trains running 2:40 or longer, from 1975 until 1980 the San Diegans 12 daily trains ran at 2 hours, 35 minutes.

With that said, after billions of dollars in investment including the deployment of coach fleets designed to minimize dwell times (California and Surfliner cars), the fact that it takes just as long, or longer, now as it did then is absolutely atrocious. 

Thursday, April 5, 2012

Oil boom leads to Amtrak passenger boom in North Dakota

ND train depot booms with oil patch passengers

The number of people getting on and off Amtrak's Empire Builder trains at the seven stations in North Dakota dropped 10.6 percent in fiscal 2011, to about 111,000 passengers, compared with the previous fiscal year. However, at the stations in the heart of the oil patch, Williston and nearby Stanley, ridership increased despite several months of disruptions due to flooding and track damage.
Williston's ridership grew by more than 5,330 people in the last fiscal year, and it served nearly 30,000 passengers. Stanley depot served 6,146 passengers, or nearly 1,600 more than it did during the previous 12 months, railroad records show.
In Minot, which is on the fringe of the oil patch, passenger service was interrupted for five months last year due to flooding. The number of riders who passed through the station dropped from 40,360 in fiscal 2010 to 29,179 in fiscal 2011.
Amtrak spokesman Marc Magliari said the Williston stop on Empire Builder is on track to be among the busiest along the Chicago-to-Pacific Northwest Empire Builer route, surpassing ridership in passing through stations in some bigger cities, such as St. Paul, Minn.
Some 4,500 people got on or off the train in Williston in January, which were the latest figures available. Magliari said the railroad expects the numbers to climb as oil production increases.
Ridership on the Empire Builder from October to February was jumped more than 10,000 passengers to 207,417 compared to the same period a year earlier, Magliari said.
The 5.1 percent ridership increase on the line -- much of it attributed to oil patch passengers in western North Dakota -- compares to a 3.2 percent rise in ridership nationally, Magliari said.
North Dakota's oil rush is boosting ridership all along the railroad's network as itinerant oil workers travel by train to the state and head back home during time off, Valley said.

Monday, April 2, 2012

Amtrak costs per train-mile per route

After doing a bit of research with my last post and being both intrigued and perplexed by some of the disparities, I thought I'd go through and expand this to the entire Amtrak network and see how it broke down per individual route. The methodology involved is using the October 2008 schedule and assuming that it holds true for the entire year as a simplification while using Amtrak's FY 2009 annual report for cost information. Weekly one directional runs were calculated and it was assumed that the other direction would mirror image it. Total train miles were calculated by multiplying the weekly run by 52. There was no attempt to take note of special days of extra running or not running and it was assumed every single train ran and was not annulled. This was actually a fairly accurate simplification as it lead to a total of 37,943,672 calculated train miles against a reported 37,439,000 train miles. Additionally, the train miles and costs of the Northeast Regional, Washington-Newport News, and New Haven-Springfield were combined together into one accounting.

Draft Revised Business Plan is out: The Sanity Strikes Back

The Draft Revised Business Plan for CAHSR is now available at their website. There is also a summary of changes. Some stream of consciousness commentary as I read through it:

They have indeed dropped the Initial Construction System as an entity and moved straight to building an Initial Operating System, which is IOS-South down to the San Fernando Valley. Interestingly, they plan on moving the San Joaquins over as soon as what used to be the ICS is finished and tie it in with Altamont Commuter Express.

The new price tag is $68.4 billion and is "for delivering the San Francisco-to-Los Angeles/Anaheim system, in accordance with Proposition 1A performance standards."

Lot of talk of building up ridership base with the blended plan and bookend improvements. Referring to it as a blended system which is far better, in my opinion, than their previous "We're separate from everyone and everything" approach.

The ICS is now "IOS-First Construction" thanks to switching the San Joaquins, although this apparently isn't distinct from the actual IOS?

San Joaquins, Capitol Corridor, and ACE are referred to as "Northern California Unified Service." I'm not sure if this is supposed to be just an ease of use term or whether they plan on actually unifying the responsible agencies.

New completion date: 2028, five years earlier. Up to 9 trains per hour, average ticket price of $81 between LA and SF in 2010 dollars. Cost in 2011 dollars is 53.4 billion.

San Joaquin Regional Rail Commission, Caltrans Division of Rail, Capitol Corridor Joint Powers Authority, and Sacramento Regional Transit have developed the NorCal Unified Service, working on a MOU for improvements to non-HSR system for increased speeds and frequencies to San Jose, Oakland, and Sacramento by the 2018 opening when they'll use the former ICS until the IOS begins.

Currently, the IOS is defined as extending from Merced to the San Fernando Valley, and high-speed revenue service would only start once the full IOS is built and operable. Should ridership and revenue forecasts and financial projections demonstrate that revenue service compliant with Proposition 1A could begin earlier, with a shorter IOS, appropriate reviews would occur to consider and implement earlier service, if appropriate.
There's actual discussion of Phase 2 in this business plan. No costs or anything however.

 Most of the cost reduction seems to be by getting rid of the full build out and being satisfied with a blended system. A slight reduction in inflation estimates, not enough to vastly change anything.

Constant dollar cost of IOS is $26.865-31.339 billion in constant 2011 dollars. Breakdown of costs on page 3-8.

A mildly odd tidbit
An important step forward in demonstrating the viability of the model and the reliability of its outputs was to use it to test actual circumstances in the Northeast Corridor. To do that, the Authority developed a California HSR scenario that has service levels comparable to those offered by Acela service between Washington D.C. and Boston. The model forecasts 2.7 million annual interregional riders on California HSR with Acela-like service in 2008, which is 79 percent of the ridership on the Acela in 2008. A comparison of mega-region population shows that the California HSR corridor had 76 percent Northeast Corridor population in 2000. The outcome therefore could be explained by the difference in population between the corridors.

They need to hype the predicted advanced fare prices like crazy instead of letting the media focus on the average fares ($52 vs $81 LA-SF as example).

So the new look of CAHSR: Dedicated HSR from Los Angeles to San Jose, blended between Anaheim and LA and San Jose to San Francisco.