Fred Frailey Blog

Amtrak and CHSRA team up for new high speed trains

  • Comments 25

Amtrak and the California High Speed Rail Authority are joining forces this coming week to begin the process of ordering up to 62 sets of high-speed trains for use in both the Northeast Corridor and in the Golden State. Thirty-two of the train sets would be earmarked for the NEC, for use at speeds up to 160 mph, and the others for California.

As of now, the plan is to announce on Thursday, January 17, a request for information (RFI) to get input from builders of such trains, almost all of them headquartered in Europe and Asia. Amtrak President Joe  Boardman said earlier this week that Amtrak expects to put the project up for bids by the end of this year.

The surprise in all this is the partnership between Amtrak and CHSRA. The common denominator between the two organizations is Frank Vacca, who left as Amtrak’s chief engineer last October to become the California organization’s chief program manager, effectively managing the build out of the new bullet train line between the San Francisco and Los Angeles areas. With his ties to Amtrak's Boardman, Vacca is believed to be instrumental in getting Amtrak and CHSRA together to make a combined purchase.

Obviously, the two organizations hope to achieve economies of scale by combining their orders. Amtrak’s new train sets would be limited to 160 mph because of safety factors, including intermingling with far slower freight and commuter trains, and the loss of capacity created by overtakes of slower trains at speeds above 160. But the equipment can possibly be configured to run at higher speeds. California has said its trains would operate at up to 220 mph. Amtrak currently limits its Acela Express trains to 135 mph except for two 150 mph sections in Rhode Island and Massachusetts, where modern, constant-tension catenary is in place. A major rebuilding of a 24-mile segment between New Brunswick and Trenton, N.J., is expected to permit 160 mph speeds by 2017.

A drawback of the existing 20 Acela train sets, delivered more than a dozen years ago by Bombardier and Alstom is that they weigh significantly more than high-speed trains in other countries, due to Federal Railroad Administration rules aimed at making them more crashworthy. In fact, some experts question whether it is possible to achieve 220 mph speeds given the weight of this equipment. Amtrak has not formally requested that FRA modify its rules, and it is not likely that FRA will do so before the bidding process begins.

The request for information process is a preliminary step to actually ordering equipment. Amtrak and CHSRA will invite builders to talk with them, explain the state of the art for such train sets, define critical engineering issues, and discuss the possible cost. This week, Amtrak had this to say about the RFI: “In considering options for this equipment, Amtrak will be assessing how high-speed equipment technology has evolved since present Acela equipment was built, how modern equipment can provide an equivalent or better level of safety while also meeting business needs, and opportunities to use this equipment acquisition to serve as a catalyst to encourage expanded domestic manufacturing of high value modern technology.”

The Acela equipment has proven a powerful revenue source for Amtrak in recent years. In fiscal 2012, which ended last September 30, Acela service generated $510 million in sales, or almost one-fourth of all Amtrak revenues. Moreover, after deducting fully allocated costs, they contributed $179 million toward Northeast Corridor capital improvements. Therefore, either expanding their size or ordering new-generation train sets became a top priority in 2012. When it was discovered that the plan to add two coaches to each of the existing train sets would be prohibitively expensive—imagine trying to replicate a 1999 model automobile—Amtrak moved up its plan to order new train sets and increase the number of those train sets from 22 to 32. The present Acela train sets would continue in service until approximately 2025, according to Amtrak planning documents. The new train sets would allow Amtrak to begin half-hour instead of hourly headways between New York and Washington.

For the California High Speed Rail Authority, the equipment-buying process being initiated is the most visible sign yet that the controversial project will actually come into being. Last summer the California legislature approved financing for the first 130 miles, between Madera and Bakersfield in the Central Valley, along with funds to electrify the Caltrains commuter territory between San Francisco and San Jose and make improvements on Metrolink rights of way it will use in the Los Angeles area. Construction in the valley could begin this summer. The political perils of this undertaking became easier in 2012 when the authority “rescoped” the project and knocked out $30 billion in estimated costs.—Fred W. Frailey

  • This has to be viewed as an auspicious development.  The more cooperation the better.  Now let's see Amtrak begin to plan to bypass the most notorious bottlenecks , i.e. the 70 mile segment of the old New Haven Shore Line between New Haven and Westerly, R.I.  This segment is loaded with 60 mph curves and even 25 mph through New London, Conn.  Can you imagine that, 25 mph on a so called high speed line between New York and Boston.  Also, the Metro North segment between New Rochelle and New Haven is notorious.  Given the high degree of development in this populous area perhaps tunneling under New Rochelle Jct.,  Port  Chester N.Y., Norwalk and Bridgeport,Conn. and the speed restrictions at each of these points could be considered as more practical than new construction.  Let's hope.

  • A most interesting blog, Fred, and thoughtful comment by WJM2223.  Those of us who are old enough to remember the Budd Co. Metroliners operated by Penn Central on he NE Corridor also know that 150 mph was consistently delivered and 160 mph was in certain locations such as Trenton-New Brunswick.  At least one round-trip was to a 2-hour, 29 minute schedule.  At that speed, Metroliner actually was competitive with the Eastern Air Lines shuttle by the time you figured in central city to central city time and cost by rail vs. the same standard by air.

    Will the cretins in Congress stand for bid solicitation and purchase of trainsets from European or Asian manufacturers?  Or, will they posture and pander about "buy American" and other silliness, not even knowing that no U.S. manufacturer is in this business at present.  Posturing and pandering is one of the very few things members of Congress seem able to do consistently.

  • Pundit - Those Metroliners only seemed like they were doing 150!  They were really operated at 120 in regular service - even the 2:29 non-stop.  In the early Amtrak years, is was reduced to 110.  

  • A great idea to team up for an equipment order. Also a great idea to see about safety vis-a-vis FRA standards.  For the NEC and CAHSR, what would be so terrible about just buying the latest ICE or TGV equipment?

    I'd also like to know how much of a premium we (us taxpayers) will wind up paying for the "build American" requirement.  Why is it so important to try to create a US supply base for this equipment?

  • Never understood why extending by two coaches the Acela Express trainsets would be 'prohibitely expensive', when (for example) Richard Branson's Virgin is busy extending their Pendolino electric multiple units from 9-car to 11-car sets.

    And the majority of the first generation French TGV PSE train sets are still running (in 2011, one set had surpassed 11.7 million km), and now these are preparing for their third interior rebuilding...


  • oltmannd:  Thanks for the comment.  As I recall the Penn Central back in the day, it's not at all surprising that Metroliner service had a hard time adhering to published schedules.  As I also recall, the original contract between PRR and Budd called for 150 mph capability.

    As for "buy America," come on, give the Congress a break!  Those clowns cannot do the public's business on the simplest of issues, so they come up with silliness like "buy American" to try to convince the public that they are serving our interests.  Sorry if this appears as a political rant, but I hold both Democrats and Republicans in Congress in complete contempt, so while this deals with public policy, it really is not political. I don't know how much of a lug we taxpayers will be forced to absorb, but considering there is no domestic HSR equipment capabaility at present, you can figure that once the lobbyists have earned their bonuses, we will be providing much if not all of the capital to create the domestic capacity, plus the obscenely higher cost of its output since there won't be any competition for the equipment.  But, we continue to send our own Congresspersons back year after year, while condemning the other 534.  We rationalize irrational behavior by condemning everyone else's senator and represent but giving our own a pass.  Recent studies say there were no more than 35 House seats that really were in play this past November.  Oligarchy lives.

  • N.F.

    ...because the Talgo folk have been churning out a steady stream of equipment for Europe for quite a while.  I'd guess Virgin's pretty much is ordering equipment from a facility has had a fairly full order book over the years.  The Acela equipment fairly specialized and was built at a "one and done" factory. You want new cars, you have to re-create the factory, complete with skilled workers.

    Now, if Amtrak had just purchase ICE equipment off the shelf, assembled in Europe, they'd be in the same boat as Virgin.

  • This is good in many ways ---  

    1.  a larger order should reduce unit prices

    2.  the order will strech out the deliveries to meet demand especially in california

    3.  even though on the 2 coasts --  spare parts can be interchanged

    4.  Only one FRA certification at Pubelo will be needed savig a lot of money

    5.  If one system RIDERSHIP DEMANDS are higher than anticipated then more train sets to the higher demand location.

    6.  Starts a national standard for other HSR lines ----  VIP  IMHO

    7.  There may be seasonal demands that are different enabling interchange of the train sets.

    8.  Identical train sets may have some multiplier effect for passengers from other coast.

    9.  Training schools could be joint as well.

    a downer would be that if a major problem is discovered after trains begin service then both systems might have shut downs of the trains at the same  will goeso

  • This was a great post! Mr. Frailey’s blog often breaks news in addition to informed commentary, making it a valuable source of rail information.

    Today I had a long talk with the Amtrak station master at Saratoga Springs about the Acela, and he told me that he has heard thru the grape vine that President Boardman and Amtrak where moving to ask the FRA to loosen up its safety requirements to allow for overseas “off the shelf” high-speed train-set designs.

    An Amtrak news memo he showed me stated that it was proving to be too expensive to reset up the Acela production line for only a small number of coaches.

    Virgin Trains has far more Pendolino sets than Amtrak has Acela sets, so it’s a much larger car order, and Pendolino is an international design by Alstom that with modifications is used in several European countries including Italy and Switzerland. The Acela design is unique to Amtrak.

    It seems that with both the CHSRA and Amtrak aiming for 220-mph (350 km per hr) service in California and the Northeast Corridor, they are both aiming at acquiring a train-set design that is far less modified than the Acela, a light-weight design unlike the Acela.

    That’s why teaming up makes sense.

    However I think that this 220-mph business is a big mistake, there is no railway in the world with a top commercial speed that high, the fastest service is the TGV Est with a top speed of 200-mph.

    Some of Spain’s newest high-speed lines where design for 220-mph running, but to date they remain at 186-mph, which seems to indicate that there is little to gain commercially from such super high speeds that have the cost of stressing both infrastructure and rolling stock maintenance.  

    China had a few lines including the new Beijing-Shanghai link at 220-mph, but a new limit of 186-mph was imposed after several problems, including a fatal crash in southern China from a major signal failure.

    Unfortunately during construction illegal skimping on the quality of building materials means that the new concrete slab track cannot support the very high speeds due to the use of poor quality concrete. This of course was a major scandal that led to arrest of the railways minster that had become a multimillionaire thru corruption and had a mistress in every city.

    But even if all the construction had been up to par, 220-mph running has won plenty of critics in China, the cost of construction and maintenance increases exponential for speed, there comes a point in the relationship between speed, costs, and revenue where the operational costs of additional speed overtake the additional revenue that can be generated from shorter travel times.

    Consider the following piece from a Chinese news article…

    Speed-obsessed policymakers including Liu (China’s ex-rail minister) have been blamed for opaque financing and higher-than-necessary costs. Feasibility studies in 2003 for high-speed lines between Beijing and Tianjin, Wuhan and Guangzhou, and Zhengzhou and Xi'an called for 200 kph (125mph) trains. But after construction began, the ministry ordered boosting each project to accommodate 350 kph (220mph) trains, which "immediately increased costs," said Zhao.

    Caixin learned that the National Development and Reform Commission, the government's economic planning agency, originally approved a 12.3 billion yuan investment for the Beijing-Tianjin project. But by the time the 115-kilometer line opened in 2008, the cost had risen to roughly 21.5 billion yuan, or 185 million yuan per kilometer of track – the most expensive in China.

    True, the Beijing-Tianjin train can travel up to 350 kph instead of 200 kph as originally planned. But the higher speed saves less than 10 minutes per trip.

    From: Caixin Online: All Aboard China's Fast Trains to Trouble

    So China tried these super high speeds and fell back to 186-mph (300 km per hr) which seems to be the world’s standard from Spain to Korea. The faster a train goes the more problems there are of ride stability, wheel adhesion, energy consumption, and noise; and while these difficulties are not insurmountable technically, they increasingly seemed to be commercially.

    Thus if you are for example connecting two city-pairs that are roughly 500-mph apart and you want a 3-hour travel time, then maglev is the way to go from the stand point of technology.

    But maglev is expensive and is completely incompatible with existing rail infrastructure, so you have to build the entire line at once, and maglev services can’t branch out on to older slower rail lines to reach cities beyond the new infrastructure, as with the TGV and Mini-Shinkansen today.

    This likely limits maglev to city pairs of many tens of millions of people, and in many places better commercial airlines may make more sense.

    So there you have it, 220 mph seems to be beyond the commercial limits of modern high-speed rail; yet this recent development of a seemingly “maximum commercial speed” for traditional steel wheel on steel rail transport has seeming yet to be noticed by either Amtrak or the CHSRA.

    I have recently reread the CHSRA’s California High-Speed Rail Program Revised 2012 Business Plan and while a great improvement with its “blended system” plan, it still is troubling to me.

    For example it still has the top speed of 220-mph, when I believe that a drop to 186-mph, or even initially to 160-mph would be a better move. One reason why the proposed system costs so much I believe is due to accommodating this super speed, curves must kept to an absolute minimum, really you can’t have much of a curve at all.

    Legally the bill which authorized the system mandates a 2hr 40min travel time LA-SF which works out for the need of an average speed of 162-mph. This needs to be change. A final travel time of 3hrs 25mins at full system build out would work just as well I believe (have a nicer café car and good comfortable reclining seating) and require a far more realistic average speed of 125-mph.

    My other problem is the new plan still delays the start of thru one-seat high-speed service Bay Area to Los Angeles till the completion of step 3 in 2029, even after the next section over the Tehachapi Mountains to Mojave is completed in 2022.

    They are apparently upgrading Metrolink’s Antelope Valley Line (double tracking, grade separation/improvements, signaling) so that future intercity passengers can transfer to faster commuter trains at Palmdale and then later at San Fernando to reach LA’s Union Station.

    This is nuts! The goal should be Bay Area-LA service as soon as possible, the extra capacity that upgrading the Antelope Valley Line creates should be reserve for thru high-speed trains (both CHSRA and DesertXpress) to Union Station.

    On the northern end the CHSRA talks about upgrading existing Amtrak and Altamont Commuter Express tracks to bring service to San Jose and Sacramento sooner. This is a good move, despite the large amount of single track running, the rail lines have very wide right-of-ways, plenty of room for a second and in some places third mainline track.

    I would suggest that the CHSRA/Caltrans work with BNSF and Union Pacific on upgrading the existing rail lines used by Amtrak and the Altamont Commuter Express.

    First the Initial Operating Section should be extended northward from Merced to Modesto. From there the existing San Joaquin route thru Stockton to Sacramento and the Altamont Commuter Express route from Stockton to San Jose should be double and tripled-tracked, grade crossings upgraded, and signaling enhanced, with a modest increase in top speeds from 79 to 90-mph.

    Electrification should also be done, with the Antelope Valley Line also being electrified to Union Station. A dual mode turbo-electric train-set is a alternative possibility, after the TGV was switch from the turbo prototype to the all-electric final model, the SNCF did consider a dual-mode to reach some cities, but in the end electrification of these older lines was seen as being more economical.

    Thus for billions less and a lot sooner, you would have a one-seat ride from LA Union Station to downtown San Jose and Sacramento. For San Francisco and Oakland the planned BART extension south to San Jose provide good access. Perhaps at a short extension of BART to the Pleasanton Rail Station could be undertaken.

    After this the next step would be building a new dedicated high-speed line over the coastal mountains from Merced to San Jose. From San Jose the electrification of the Caltrains line to San Francisco gets you to SF’s Fourth & King Street Station or the Transbay Terminal.

    That’s why for future rolling stock I think that the CHSRA should be looking at an “Acela type” heavy-weight high-speed train, perhaps an EMU with a with a tilt system. It’s possible that an EMU would have a good enough power-to-weight ratio to allow for a top speed of 186-mph despite all the extra steel it would need to meet FRA crash standards.

    A tilt system could further cut travel times over the Altamont Pass and thru the Antelope Valley. The tilt system could also be a 2 or 3-degree passive tilt system like those developed in Japan. For example Nippon Sharyo new design is completely reliant on onboard equipment (no track side transponders) and the maintenance is for more simple than active tilt systems. A 2 degree tilt gets you an extra 25-kph (15mph) around a curve with a radius of radius of 600 meters (2000 ft).

    The train won’t be a Pendolino (8-degree tilt) but should give performance around curves about equal to the current Acela (4-degree tilt).

    As for Amtrak, I believe the same logic applies. What we need is more tunnels into Manhattan and more sections of track where high-speed trains can run over 125-mph, let alone 220-mph. The top commercial speed should be no more than 186-mph, and 160-mph should be acceptable.

    Let’s learn from others and not repeat their mistakes. A slightly slower high-speed train while be a lot more affordable, and thus more likely to actually be built!

  • @oltmannd:

    As others noticed already, Talgo has nothing to do with the Virgin Pendolino:


    In Europe, the highest speeds being operated regularly are 320 km/h (205 mph) in France LGV lines, while Spain is slowly rising the top speed above 300 km/h. Officially, the ICE-3 has 330 km/h maximum speed (you can see the UIC numbers on its side), but this is not the typical operating speed, as the energy consumption rises steeply beyond 300 km/h, and the time gained was not worth it.

    The train driver has the option of going faster than 300 km/h in order to recover a delay.

    Chinese have a different problem, though, than Europe: vast distances (compared to Europe, that is).

    One line inaugurated recently has a length of 2200+ km, so speeds like 380 km/h are deemed necessary for extending the train operation radius (see , especially the "Track Network" part, and ).


  • Also, I would suggest you study this blog entry regarding China HSR (he has lots of entries in his blog about railroads, but I feel this entry is a good chunk of material):


  • If they can't figure out how to keep the weight down, we could wind up with HSR trains that are less energy efficient than future airliners and automobiles.  CAFE is going to double in the next decade or so, and current Amtrak trains are only about 20% better than driving.

  • N.F.

    Oops.  Point remains, however.  Virgin's additional cars won't be constructed in a new facility with new workers.  Acela was.  And the Acela replacements likely will be too, unless we get smart.

  • I guess I am a skeptic / cynic.  How will CHSRA link the electrification of Caltrain to the Madera - Bakersfield new build section to operate these new trainsets, much less finish a through route between the Bay and L.A.?  I hope this doesn't really just end up screwing up Amtrak's N.E. Corridor equipment plan because CHSRA cannot deliver on their part of the deal.  

    If someone told me Amtrak and the FEC were going together on an RFI / RFP for new trainsets (Talgo anyone?) for the Chicago-St. Louis / Kansas City route and the Chicago - Detroit route and the new Florida trains, I might believe it.  But I still think CHSRA is a fantasy, and I worry that it will just hold back progress where real progress is needed on the NE Corridor.

    I am happy Amtrak scrapped the notion of expanding the white elephant Acela sets, but I just fear now the evolution of the NE Corridor will be held hostage to a fantasy in California.  I would love to see CHSRA build out the California line, but I just don't see how the financing falls together yet.  CHSRA has made a lot of progress towards rationality with the decision to use the Antelope Valley Line, but there are still big questions of how the funding ultimately comes together.  I think counting on Federal and State funding in the next decade is nothing but a fiscal pipe dream and I have yet to see the private sector offering to fund it based on the current Federal and State seed money.

  • Amtrak needs the units long before CHSRA will be ready to operate the trains. If they are ready in even five years for Amtrak, they will probably be 10 years old before they ever run in California.

Amtrak and CHSRA team up for new high speed trains