The High Speed Trainset Makers and America's Passenger Network

A 6-part series by the Infrastructurist:
Alstom - Bombardier - Talgo - The Japanese - Siemens - The Chinese
(this inherently takes a somewhat skewed look at the worldwide train manufacturers and mushes a lot of the middle-speed train makers but is still a decent look at the scene)


And now for a brain-dump of just a few of my thoughts on the matter:
(some epic run on sentences and multi-layered parenthetical thoughts ahead - you have been warned!)

The Acela's were a Alstom-Bombardier consortium with domestic manufacturing to satisfy the politicians (ironic given our not insignificant diesel loco export business for GE and EMD to those same countries...). They were reasonably-successful, but had their usual share of teething problems inherent to all new designs (which is why standardization of these designs should be a high priority - whatever you choose you should stick with as long as you can, something the American freight RRs know well - Union Pacific has over 1000 SD70M's which helps reliability and cost-savings). But the Pacific Northwest (Vancouver to Portland) has the Cascades Talgo trainsets. These sets have been outrageously-successfull and have something close to a 97% availability record - a reliability record any company would kill for despite there only being 5 of them in the continent. They use what I believe to be the superior tilt-train technology - the passive system that uses inertia, not computers to control tilting (and therefore is inherently more reliable). They also have some of the safest designs and remarkably low weight designs (which maximizes efficiency, acceleration, and speed). As in the Cascades corridor, the Talgos often have separate conventional locomotives power the sets - this is a huge efficiency, reliability, and logistics boon (for a number of reasons that would require far more knowledge of the rail network structure than most of you care to know, but I'll gladly elaborate if you wish). I think that Amtrak should have returned to Talgo for the Acelas and gotten more conventional electric locos like the AEM-7 to power them - they have more than sufficient maximum speeds, greater reliability, and far greater operational flexibility and future-proofing.

One significant shortcoming of all of these highspeed trainset designs is the fact that they are integrated, semi-permanently-conected units, not individual cars. This is required in part to maintain stability and structural integrity as such speeds, but is not an inherent requirement, particularly in the mid-speed ranges (80-125mph) where most of America's new "high speed" ventures actually fall. For this speed range either conventional locomotive-hauled cars or building-block style short MUs ("multiple units" - self propelled mini trainsets) are ideal from an operations standpoint. However it may be harder to get tilting trains into this form factor since I know of no examples elsewhere in the world.

One thing is certain though, we can't simply copy foreign designs wholesale - it's been done before and ended up costing far more in retrofits and overhauls to meet the different domestic operating needs. Also American reliability and maintenance standards are far different from those of the rest of the western world - American diesel locomotives cost less than half as much to maintain as most foreign designs and require far less specialized training or expensive equipment. Japan has the most reliable passenger network in the world not because of superior designs, but because Japan spends WAY more on maintenance, infrastructure, and redundancy than anyone else (even the French and Germans). It is simply an entirely different operational culture that leads to very different design priorities - standardization and reliability matter more than tweaking out the last 10% of performance in America, and anything that is designed with the expectations of French-stlye maintenance will break down constantly here. On the other hand, for electric locomotive designs we have to go foreign - there is simply no technical experience domestically. Sweden was the source of the Northeast Corridor's AEM-7 locos and they have an impeccable record. The key is focusing on using existing proven reliable designs and adapting them to American conditions, not starting from scratch.



If I had the power to choose the passenger equipment to standardize on for America, it would look something like this (all operated and owned by Amtrak and Amtrak/state cooperatives - none of this privatized or regionalized bullshit, well, unless we wanted to make a serious investment and policy shift - those options are valid, but only if you're willing to make some radical politically-charged changes and put up serious dough):

-None of this lowest bidder or most politically-convenient decision crap - this is a serious longterm investment in an entire manufacturer set for at least the next 20-30 years and the reliability, flexibility, and upgradeability of this equipment is far more important and economical.

-Screw China - none of their technology is actually theirs, and they have't got the design or manufacturing experience for reliability - history has proven that locomotives and trainsets made by newcomers to the business are ALWAYS retired within the first 15 years of use (most locos/cars have 20-45 year lifespans)

-for true high speed (150-250mph) go with Bombardier, Alstom, or Japan. Period. They have the technical know how and track record. These will have to be electrics, nothing else will go this fast reliably and affordably, so it will only be possible in the very highest-density corridors that can support such massive investments - you need much higher train frequencies in the US than overseas to make electrification affordable and it will always be nonstandard in the US which shortens the lifespan of the equipment significantly (nonstandard always costs more to maintain and therefore is the first to be cut).

-but for the majority of the network, mid-range speed will have to do (79-125mph) - it's all we can afford politically speaking in most areas, and can be implemented for a tenth the cost of true high speed. What matters far more than speed in day to day use is reliability and a practical schedule (high train frequencies, logical timing, good connections, etc). These are all things Amtrak already knows how to do if only it has the money it needs. (The host freight railroads will cooperate just fine if Amtrak pays it's fair share of the track investments and maintenance - history shows this). Conventional equipment is more than adequate for many of these routes, but tilt trains will noticeably-improve service. I'd use conventional locos (mostly diesels with electrification in those few affordable key corridors - electrification is more expensive in the US than europe, and we don't have the local technical know how so diesels are a much safer bet for most areas given the relatively low traffic frequencies) and Talgo trainsets for the faster routes. Conventional Viewliner, Amfleet, and Superliner-style cars will be best for the other routes where reliability, versatility, and familiarity are key. (Obviously the long-overdue new cars Amtrak has been trying to get for over a decade and a half would need to be bought to meet demand since Amtrak is already desperately short of cars - I'd stick to the already developed and prototyped excellent Morrison Knudsen Viewliner designs for much of it.)

-The more expensive alternative/adjunct approach is short standardized trainsets (DMU and EMUs). These should be no more than 3 'cars' long and designed to be used building-block style to meet the variable traffic needs of a given route. England, and well, most of the westernized nations use these extensively for their commuter and slower intercity traffic, and they work particularly well in metro shorthaul situations, but are also quite effective in mid-hauls. These have distinct efficiency and power to weight ratio advantages, but have serious logistical limitations and are best used in certain specialized services. Ideally these are part of a larger design approach, not the main strategy.

-The big key is not the equipment - the Acela's are capable of 150 mph but trudge along at 79 for much of the route - but the infrastructure. Dedicated right of ways are necessary for true high speed, but are so expensive that they can only be justified in a couple locations for now (one day perhaps...). For the most part it is more important to upgrade the existing routes incrementally to remove the numerous bottlenecks. This will mean working cooperatively with the host freight railroads, and they have been very receptive to this for the past decade-plus as long as Amtrak or the states pay their share.



We need true high speed rail, but it needs to be part of a larger passenger network investment. Bullet trains make great public works spectacles, but the money for one high speed corridor could triple the service levels of all the other routes combined. The rest of the world backs up their high speed trains with frequent mid-speed and local feeder services that make an effective transportation network actually work in the real world. For every bullet train there are at least a dozen less glamorous trains making the bullet train possible. We're talking hundreds of billions of investment that we need to make here, not the pathetic 500 million a year Amtrak typically struggles to survive on.