My thoughts are with the victims of the tragedy in Philadelphia this morning. I’ve been following the derailment of Amtrak’s Northeast Regional 188 at Frankford Junction north of Philadelphia 30th Street Station since shortly after it happened the night of Tuesday, May 12. It hits close to home- on my way to work that night I passed over the same tracks that 188 had been traveling north along a couple hours before just prior to the accident. My father worked for Amtrak for many years and I’ve ridden Amtrak frequently since I was a child.
With at least
five eight people reported dead (update 3:16pm EDT May 14- death toll now eight) and many others hurt, it’s Amtrak’s worst accident in many years. As of 7:00am, there haven’t been many updates in the story since around midnight last night. Although I avidly ride trains and prefer it as a way of medium-distance travel, I have no expertise that allows me to speculate on what happened based on overhead helicopter footage of the wreck. The news media is already speculating on the cause based on the presence of a curve (update 4:34pm EDT- seems they were right, as NTSB is reporting the train took a 50mph curve at twice that speed) and the fact that a disaster happened at almost the exact same spot…in 1943. Of course, the 1943 accident was blamed on a hot box. I can’t recall the last time hearing that a passenger train crashed because of a hot box. Do they even have that problem any more? Did Amtrak’s form of Positive Train Control, the Advanced Civil Speed Enforcement System (ACSES), function as designed or was it not a situation in which it could have prevented the accident?
Among Amtrak’s accidents, very few have been due to failures on Amtrak’s part. Among the incidents of the last 25 years, there is plenty of blame to go around- barges hitting bridges, tractor trailers running crossings, host railroad track maintenance failures, rails warped by heat, engineers on other railroads under the influence of drugs ignoring signals… Only in a few incidents did blame fall at the feet of Amtrak engineers or maintenance. With this incident involving the Amtrak owned Northeast Corridor (NEC) and no other trains apparently involved, one would imagine Amtrak will bear primary responsibility in this particular incident.
Regardless of who is to blame, traveling by Amtrak is very safe. Deaths per passenger mile are about 15 times higher by automobile than on Amtrak. I can only imagine the extra traffic deaths that might result if the public reacts with irrational fear to the news of this accident. To put these generalizations into perspective, prior to this incident, Amtrak passengers were killed in accidents only twice in the last fifteen years- 2002 when Auto Train derailed due to weather and CSX track maintenance issues, and 2011, when a tractor trailer crashed into the California Zephyr. On the NEC itself, the busiest line in the Amtrak system (currently with over 40 trains in each direction each weekday), between 2004 and 2014 some 113,807,000 passengers traveled on Amtrak without a fatality. In fact, before 2015 the last time an Amtrak passenger died in an accident on the NEC was 1987. I’ve ridden this route many times and I would board a Northeast Regional tomorrow (or whenever the NEC reopens) with no hesitation.
I don’t have the numbers to support my impression but despite well publicized accidents, I suspect rail travel is now safer than it’s been at any time in the past two centuries. My salute to the first responders who managed the crisis from the outset and all night. Mass casualty incidents are a true test of an emergency response system.
UPDATE May 14, 12:37am EDT
So it seems speed and crew error are emerging as the primary causes of the accident. NTSB is saying the train entered a curve with a 50mph speed limit at 106mph. It’s not only the curve speed that is at issue. Even in the straightaway leading into the curve, a speed of 106mph in an 80mph is an egregious violation. Driving in to work on I-95, I was being passed by cars going 80 in a 55. That’s not particularly uncommon and the penalties if Delaware State Police were enforcing would be a mere traffic ticket. The same speed violation on a railway is liable to get an engineer’s license revoked. Supposedly, the engineer lawyered up after brief questioning by police. It’s with good reason- losing his job is assured…manslaughter or similar charges, virtually certain.
In regard to my earlier inquiry, Amtrak’s ACSES Positive Train Control system (which variously seems to be pronounced “access” or “aces”) was not installed in that particular stretch of the NEC, although eventually the entire NEC should be equipped, ideally by December 2015. According to the Jan/Feb Amtrak employee newsletter:
Currently, ACSES is operational on 400 miles of track. These service lines include the New England Line from Boston, to New Haven, Connecticut; the New York Line from New Brunswick to Trenton, New Jersey; and the Mid-Atlantic Line from Perryville, Maryland to Wilmington. On the horizon, Amtrak looks to implement ACSES on 1,200 more miles of track. Target lines include the remainder of the Northeast Corridor from Washington, D.C., to New Rochelle, New York; the Harrisburg Line from Philadelphia to Harrisburg, Pennsylvania; the Springfield Line from Springfield, Massachusetts to New Haven, Connecticut; and the Empire Line from Penn Station to Spuyten Duyvil in New York City and Albany to Poughkeepsie, New York.
Speaking of Spuyten Duyvil, this accident appears to have a lot in common with the December 2013 crash on Metro-North’s Hudson Line. Error by the engineer that he attributed to fatigue caused the commuter train to derail with four fatalities after entering a 30mph curve at approximately 82mph. It remains to be seen what explanation if any the Amtrak engineer will provide.
UPDATE May 14, 3:33pm EDT
The death toll is now at eight and with all passengers reportedly accounted for, will hopefully not rise any more. I suspect most if not all fatalities were in the shredded first car behind the locomotive. It’s a pity it wasn’t a long distance train, since that car would have been a baggage car. Some questions have been raised about whether seatbelts would have saved lives. I can’t think of a train anywhere in the world that provides them (and I bet like buses most people wouldn’t wear them if they were present), though in principle I acknowledge that especially in a rollover following a derailment, seatbelts would probably make a huge difference, as would having hatch-like luggage storage like on aircraft and the Acela. Bad as it is, the death toll is astoundingly low considering the high speed of the derailment, a testament to the safety of modern rail car design.
Now that we have some idea about the cause of the accident, the question on everybody’s mind now is “Why?” It’s probably safe to speculate that the engineer did not make such a dangerous and career-ending move intentionally, although following the recent Germanwings crash I suppose we can’t take anything for granted. According to the current Amtrak timetable, which shows 188 due to depart at Philadelphia at 9:09pm, the train was on time. At any rate, Casey Jones-style cannonballing to make up time is a thing of the past.
UPDATE May 15, 11:39pm EDT
A lot of articles about train safety have been swirling since the accidents. An surprising number quote without critique a biased “report” on train safety from the American Enterprise Institute, a right wing think tank that has been trying to dissolve Amtrak for years. My father, a retired railroader, recommended a New York Times article, Amtrak Says Shortfalls and Rules Delayed Its Safety System as one of the best accounts of the issues over Positive Train Control implementation.
In an interesting coincidence, prosecutors just announced they do not plan to pursue criminal charges against the Metro-North engineer responsible for the crash I mentioned above. His “undiagnosed sleep apnea” and Metro-North scheduling issues apparently made them think it was not in the interests of justice to pursue manslaughter charges or such. It remains to be said what explanation will emerge regarding the Amtrak engineer’s actions, which are especially puzzling since the NTSB indicated that he only started speeding very shortly before the accident occurred.
UPDATE May 21, 2:35am EDT
So the NEC has reopened and the NTSB investigation may not be complete for a year. The media has been lapping up every morsel of the investigation they hear about. The latest is whether or not the engineer’s phone was really off and in his bag as he stated. I suspect the end result is going to be a finding of simple human error. The engineer began speeding only just before the accident. A moment of inattention perhaps that in his mind he was already around the curve in the 100mph+ zone. The engineers are expected to know their territories by heart. The fact that nobody has derailed at that curve in the last few decades of regular operation indicates that is usually enough.
I recently asked my father why locomotives aren’t mandated to have a two person cab just like commercial flights carrying a similar number of people require two pilots. He answered that this is a requirement on long-haul routes over 4 hours, but with a crew change in New York this doesn’t apply to the NEC. Even freight railroads, which typically have fewer lives at risk, often have the benefit of a two person cab since the conductor rides in the locomotive and calls signals, whereas conductors on passenger trains do not. Certainly, a two person cab may have prevented this accident or the Metro-North one mentioned above. Interestingly, my father informed me that the engineers’ union has long been opposed to one person cabs. Indeed, on May 19 the Brotherhood of Locomotive Engineers and Trainmen released a statement that included the following:
In 1981, Congress passed legislation (the Northeast Rail Service Act of 1981) that ended the previous Conrail requirement that there be a second crew member in the control cab of commuter rail trains on the Northeast Corridor. Armed with that legislative precedent — and mindful of where its funding originates — Amtrak has since 1983 refused to crew Northeast Corridor trains with more than one employee in the cab – the locomotive engineer. Although BLET and SMART–TD have steadfastly maintained that there should be two crew members in the cab of all trains to ensure public safety, only Congress can change the 1981 legislation that reduced crew size on the Northeast Corridor. But this is only one piece of a very large, complex puzzle.
Clearly the union’s position would serve the interests of its members (or at least add to the ranks of its members!) but that doesn’t mean it is wrong. As a practical matter, of course, with well over a hundred trains per day operating on the NEC’s various railroads, the costs of adding a second crewmember to each cab would be staggering. At the same time, the victims and their families would probably not be swayed by such financial concerns. It comes back to a frustrating basic question- how many millions of dollars a year should be spent on something that might prevent a fatal accident every few years or every few decades? How much money is a human life worth? I don’t have an answer for this point of contention.