In 2017, an Amtrak Cascades train derailed outside Seattle. The train driver sped on a curve and the heavy locomotive derailed, dragging the trains with it, as had happened in 2013 in New York and in 2015 in Philadelphia. The primary culprit was the tardy installation of automatic train protection (“positive train control”), which would have prevented overspeed: the Philadelphia accident happened shortly before that section of track was scheduled to get PTC, the New York accident happened on a line with weaker protection against running red signals but not against overspeed, and the Seattle accident happened on a line not-yet equipped with PTC but with ongoing installation.
Despite the similarity, the National Transportation Safety Board’s recommendations, just released after a year and a half’s investigation, are different: it demands that Amtrak withdraw the passenger cars used. Those cars are tilting trains manufactured by Talgo, running on a waiver from federal regulations that have since changed. They are not what led to the derailment, and evidently the crash was less deadly than the New York crash (which killed 4) as well as the Philadelphia one (which killed 8). And yet, the NTSB wants these coaches removed where it said no such thing about the coaches used at previous accidents.
Rolling stock that is designed to avoid derailments at curves should focus on getting its center of gravity under control. Diesel locomotives have high centers of gravity, limiting their speed on curves. The standard measure for how fast a train can go on curves is called cant deficiency, and de facto has the conversion rate of 150 mm = 1 m/s^2 of lateral acceleration in the plane of the tracks (which may be banked, so the acceleration in the horizontal plane is larger). Top-of-the-line tilting trains in Europe can safely take curves at 300 mm cant deficiency, at which point the limiting factor isn’t passenger safety, but the complexity of maintaining the equipment’s tilt system. The Amtrak Talgos are less powerful – their tilt system allows only about 180 mm, but their center of gravity is so low that they could go much faster without derailing, just with a lot of passenger discomfort over the centrifugal force.
In contrast, the heavy diesel locomotive is limited to 6″ of cant deficiency, as its center of gravity is so high that going much faster could risk derailment, as indeed happened. Electric trains have no such problem – the Acelas do 7″ every day, and could have done 9″, but as with the Talgos the limit is passenger comfort and not train toppling. In fact, some non-tilting trains in Europe run about as fast on curves as the Amtrak Cascades service does, and FRA regulations from 2010 even suggest that modern non-tilting trains could do about 6″ of cant deficiency.
I bring all of this up to explain that the Talgos on their own, with a typical European locomotive, would not have derailed. Moreover, after the derailment, they stayed upright, unlike the Amtrak coaches in Philadelphia or the Metro-North ones in New York. The reason people died is that the train fell from a bridge. As far as factors that are controllable by the coach builder go, the Talgos performed well.
What’s more, the regulators who compared European and American designs for rolling stock have come to the same conclusion. This underlay the tests Caltrain used to argue in front of the FRA for its waiver to get rolling not compliant with the FRA’s then-current regulations demanding heavy trains designed around static buff strength; this would later underlie the FRA’s own rule change, now permitting lightly-modified European imports on American tracks.
So why is the NTSB so dead set against them? In three words: not invented here. The Talgos were designed and built in Europe. They are designed around European ideas of crash avoidance. Trains here have buff strength requirements too (and are too heavy as a result), but they’re much laxer than those of last generation’s American regulations, because at the end of the day lighter trains are no less safe than American tanks on rails. Lighter trains, designed to brake more quickly and not to derail in the first place, underlie the superior train safety of Europe to that of the United States – and Europe is downright dangerous compared with Japan, whose ultralight trains kill passengers in crashes at maybe 1/15th the per-passenger-km rate of American ones.
Foreign ideas threaten Americans, especially incurious Americans. Americans are not used to not being in the center of the world. They have many reactions to the fact that when it comes to transportation, the global centers of innovation are elsewhere. Some believe private cars are just superior. Others do not but still deny the need for trains, for example elements in tech media who cover Elon Musk as if he were God. And yet others assert that the US actually has amazing rail service on a variety of specious grounds – freight, safety, unique history. All of these are excuses made to avoid learning from foreigners.
The United States should treat its mainline passenger rail as a tabula rasa. Its ridership is a rounding error by European and Asian standards, and this is in large part due to the failure of management to modernize over the last few generations. An American regulator, investigator, or manager who neither knows how to adapt best industry practices, none of which is domestic, nor has any interest in learning, has no business working in the industry. Moreover, to reinforce the need to learn from the best, the NTSB should begin an exchange program with European and East Asian regulators; exchange programs among railroads in Western Europe, Russia, and Japan already exist, but to my knowledge American railroads do not participate in them.
The lessons from the Cascades crash are the same as those from the New York and Philadelphia crash, and drawing different lessons is prima facie evidence of irrational prejudice against foreign design elements that in fact work better than domestic ones. The people at the NTSB who authored the recommendation to get rid of the Talgos have just shown themselves to be incurious about practices in parts of the world with better rail safety. Thus, they should all be immediately dismissed from their jobs and replaced by people who are more informed. Any day they keep their jobs is a signal that the NTSB will be an obstacle to any program to make American passenger rail not only faster, more convenient, and more reliable, but also safer for passengers and employees.