Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
Sign in to follow this  

The DC Metro's lowest points--The June 2009 Red Line collision and the 2015 L'Enfant Plaza smoke incident



Shit, I missed the 11 year anniversary of the Red Line collision by 2 days. Rather than waiting until next year, I'm just going to roll the June 2009 and January 2015 incidents into one as they lay the groundwork for some post ideas I have and I want to stop beating around the bush with them.

So, here we go.


The evening rush on June 22nd, 2009 was typical for the Red Line, at least until 5:02 PM. Red Line train 112 left the Takoma station at 4:57 PM en route to Shady Grove station, located in a Maryland suburb. Ahead of it, train 214 had entered a faulty track circuit which gave the ATO system a speed command of 0MPH. So far, trains had enough momentum to make it through the track circuit without coming to a full stop, preventing events similar to what was about to unfold. The thing is, train 214 was running at a lower speed than usual due to it being operated in manual mode for whatever reason. This meant that the train did not have enough momentum to pass through the track circuit when the speed command of 0 MPH was received. Because train 214 was holding in a faulty circuit, it was invisible to the ATO/ATP system (the latter is still used today, as the system is still using ATC which is a less autonomous form of ATO). Train 112 was given a 55MPH speed command (the highest allowed for that stretch of track), and train 214 was behind a blind curve. Once train 214 became visible, the T/O on train 112 (the DC Metro kept operators around for door closings (well, not originally) and in case of emergency) activated the emergency brakes. They had little effect, and IIRC, train 112's railcars were supposed to receive break service a couple months before the incident, but it was differed. (I say IIRC because I remember reading that but I can't find a source). At 5:02PM, the lead railcar, a 1000 series delivered in the late '70s when the system opened, telescoped over railcar 5066 at the end of train 214, creating this image that is famous in the "fix the DC Metro circle":

June 22, 2009 WMATA Collision - NTSB accident photo 422860.jpg

Ouch. Railcar 5066 took this pretty well, but it and it's mate were retired and used for spare parts:


This crash wasn't the first incident on this system, nor was it the first of it's kind (more on that later in this post). But it was the first to cause deaths. 9 of them.

So what caused it?

Well, as I said, one of the causes was a faulty track circuit. It had been suffering from parasitic oscillations which caused it to unreliably report if a train was occupying the circuit. This wasn't the first time this was an issue, either. A series of near-collisions occurred in 2005 between Foggy Bottom and Rosslyn stations, due to the same type of failure. A new testing procedure meant to detect these issues after installation was imposed. Unfortunately, by 2009, Metro's engineered were unaware of these incidents or the tests developed to prevent them. But the issue that really added salt to the wound was the poor design of the 1000 series that allowed them to telescope in a collision.

In 2004, this was first seen when a train that was out of service began to roll (due to no anti rollback protection) backwards into Woodley Park station in DC. It collided with a train servicing the station, and guess what, the end railcar was from the 1000 series and telescoped over the lead car of the other train. Luckily nobody died, but when the NTSB published their report on the incident in 2006, they had recommended the accelerated retirement of the 1000 series, or at least a rehab program that would bring crashworthyness up to the level of the then new 6000 series railcars. Of course, WMATA did nothing to make this happen, and in 2009 it came back to bite them.

The aftermath of this lead to the 1000 series being placed into the middle of trainsets (though this didn't really do much to help as in a collision the energy could still be transferred to it, an NTSB official even said that they would not ride in them at all) and the deactivation of ATO. Thing is, the NTSB didn't blame them for the track circuit failure, as Metro had replaced all 20,000 of the relays in 1999 since they began to fail after only 25 years of the specified 70 years. WMATA inspected all circuits in the system after the crash, finding 6 others with unusual behavior. They disabled circuits that couldn't immediately be fixed and created an online circuit tracking system similar to their elevator outage tracker. It was also found that the track circuit where the collision occurred had been malfunctioning since 2007.

Overall, this collision could've been prevented if WMATA focused more on safety. Unfortunately, this crash did not cause WMATA to learn any lessons, as during the evening rush of January 12, 2015, a southbound Yellow Line train (#302) departing L'Enfant Plaza began to fill with smoke. The cause of the event was an arcing insulator that had begun producing smoke due to degraded covering of the wires allowing for contact between a pool of water (caused by tunnel leaks) and said wires. 

L'Enfant Plaza station had been evacuated, but the following events showed the dysfunction of WMATA.

First off, under platform fans were put into "exhaust mode" which blew the smoke in the station towards the train. Adding salt to the wound, ROCC took 8 minutes to activate ventalation shaft FL-1 on the other side of the train. It was also put into exhaust mode, so no fresh air was introduced into the tunnel. Two of the fans in the shaft were not working due to falsified maintenance logs (and yes, the union tried to get the douche who signed off on the "maintenance" re-hired a couple years later.)

Second, there was no published procedure to disable the ventilation system on all railcars, making it so there was smoke constantly being introduced into the railcars.

Third, the train couldn't be moved back into the station due to the fact another train had pulled in and was evacuated (ATO could've helped here, allowing the train to be moved even without an operator).

Fourth, it took over 30 minutes for first response to arrive to disconnect 3rd rail power and to evacuate passengers. It had been 44 minutes since the insulator started arcing in the first place. Some passengers evacuated themselves because of the long wait, and unfortunately, somebody died.

After the incident the NTSB said that WMATA's safety advisory lacked the resources, technical capacity and authority needed to provide proper oversight of the system. They also said that the FTA lacked the authority, expertise, and resources to assume temporary, direct safety oversight  of rail transit agenesis.



Holy shit, the reason Yellow Line train 302 left L'Enfant Plaza was because of this:



If that doesn't show the lack of safety culture at WMATA, I don't know what would.

We now return to your regularly scheduled programming:


In March 2016, the entire system was shut down for a day to allow for inspections of the 3rd rail system. In May 2016, there was an explosion at the Federal Center station (before service started AFAIK, I can confirm there were no injuries or deaths) that nearly lead to another shutdown (though WMATA didn't suspend service on this section of track until another track fire occurred later in the day). Later that year, WMATA launched the "SafeTrack" initiatve, where sections of track were fully shut down for weeks or even months to allow for much needed maintenance. it ended in 2017, and had mixed results. In 2017, track fires were very common, though they did slow down after the last SafeTrack phase was completed. I remember the section of track near my house being shut down, restricting access to the biggest yard in the system (Greenbelt). Ironically, on the reopening day of the section, a switch (that had been worked on as part of the shutdown) at College Park failed.

I must say that overall, the system is safer than it was in it's darkest years, but it still has a long way to go.



Regarding accuracy:

  Hide contents

I'm definitely not 100% correct with any of these posts. If you find an accuracy issue with one, please let me know.



Recommended Comments

There are no comments to display.