Transport: Chatsworth Collision

TORONTO, ONTARIO - A number of people have asked me questions about the collision between a Metrolink commuter train and a Union Pacific freight train in the Los Angeles basin on Friday, killing at least 25 people. I generally do not find it constructive to comment on such events until all the facts come out, but some ideas do transcend the details so I will present them here.

Some facts have come out and are clear, gleaned from discussions on the trainorders.com web site and from looking at the latest Altamont Press California Region Railfan Timetable. Westbound Metrolink train #111, the first afternoon run out of Los Angeles on the Ventura County Line, was bound for Moorpark, California when it made its station stop at Chatsworth, California on Friday, 12-September-2008 at about 16:15. It was running locomotive-first, with three passenger cars, and carried about 350 people. From an audio recording of radio conversations, the engineer had seen an "advanced approach" or flashing yellow signal at CP (Control Point) Bernsen, two signals before the station, had called this out on the radio, and the conductor had acknowledged the signal on the radio. An intermediate signal near Lassen Street in Chatsworth, before the station, was not acknowledged on the radio but likely displayed an "approach" or solid yellow signal. Both these signals would be consistent with the next signal at CP Topanga, about a mile beyond the Chatsworth station, being a "stop" or red signal, indicating that the train should stop.

The line is dispatched out of Los Angeles by the Metrolink Valley Dispatcher, and his intention on this day was that the Metrolink train would wait at CP Topanga for the eastbound Union Pacific Leesdale Local (LOF65), returning to the Van Nuys yard after its day's work at industries as far west as Oxnard. The Leesdale Local had two state-of-the-art freight locomotives and at least twenty freight cars. The line has two tracks between CP Bernsen and CP Topanga through the Chatsworth station, but is single track on either side of those locations, so meets in this section of track are common. There is even a YouTube video of the very same trains, Metrolink #111 and the Leesdale Local, meeting at this location posted back on 30-April-2008.

After making the station stop at Chatsworth, the Metrolink train would have been under "delayed in block" rules that require the train to be prepared to stop at the next signal regardless of previous signal indications until the next signal is visible and indicates that the train may proceed. The signal at CP Topanga is readily visible from Chatsworth station, so if the engineer saw a "proceed" or green signal, the "delayed in block" rules would have been irrelevant and the train could have accelerated to track speed, which is 70 miles per hour for passenger trains until the curve that starts just west of CP Topanga, on which all trains are restricted to 40 mph.

Whether the engineer ignored a red signal or the signal was actually green, the bottom is line that the Metrolink train ran through a switch apparently lined against it (which is not necessarily perceptible) at CP Topanga and entered the single track after the Leesdale Local had already entered the same block of track. The dispatcher actually noticed this on his display board and immediately got on the radio to contact the Metrolink train, but it was too late. The conductor responded on the radio that a collision had already occurred.

Most of the impact of the collision was taken by the first passenger car on the Metrolink train and the second locomotive and first freight car on the freight train. That passenger car ended up on top of the passenger locomotive, and was where most of the fatalities occurred.

While the collision took place in a bad location, on a six degree curve such that neither engineer would have had a chance to see the other train until way too late, it could have been worse. Less than a mile west of this location is the 537-foot long tunnel #28.

Many people have asked if it is really possible that an engineer can run past a red signal without anything happening before a collision. Unfortunately, on the vast majority of the trackage in North America, this is the case. While systems that would stop a train after passing a red signal exist (most notably the Advanced Civil Speed Enforcement System or ACSES used on the Northeast Corridor), some primitive ones for nearly one hundred years, they have been deemed cost-prohibitive on a rail system that is dominated by freight traffic. Passenger service, especially urban commuter rail systems, has been an after-thought since the 1950's.

While basic systems such as Automatic Train Stop (which, ironically, is in use on Metrolink's Oceanside Line) would not have prevented the collision completely, they would have at least slowed the passenger train. Even the implementation of a full Positive Train Control (PTC) system such as ACSES on the passenger train would have similarly at best stopped the passenger train and not slowed the freight train. The only way to actually prevent such tragedies is a full implementation of a PTC system everywhere that passenger trains run, and this would mean not only adding to the wayside signaling system but adding such safety equipment to at least the majority of freight locomotives. The prospect is so dead-on-arrival in the United States, costing in the billions, that nobody appears to have even done a full cost estimate.

In fact, the whole strategy in the United States has not been to prevent collisions, but make the collisions survivable. Crash standards for passenger railroad equipment in the United States are far more stringent than in the rest of the world, such that off-the-shelf high speed equipment from other nations cannot be used without significant re-design. Europe, where freight traffic is generally secondary to passenger service, has taken the opposite approach, trying to prevent collisions and requiring less of its equipment. Perhaps this crash, in which passenger equipment that met Federal Railroad Association requirements still failed to protect passengers, will cause a re-thinking of that strategy.

When the remainder of the facts come out, perhaps there will be more to say.