On Friday [January 13th, possibly the coldest day of the year with a low of just 11 degrees], a MAX train derailed at Rose Quarter while going from the Blue Line to the Yellow Line. On KOIN 6 News, Mary Fetsch noted how "it has never happened before". However, looking at news media photos and videos, it seems clear that the piece of track was unusable and a derailment was likely to occur:
KOIN 6: Ice-covered tracks derails MAX train in Rose Quarter
KATU 2: Complete MAX derailment on icy track shuts down multiple lines; seek alternate means
KATU 2:MAX train derailment crippled morning commute, TriMet investigating
KPTV 12: TriMet crews restore service after getting derailed MAX train back on track
The Oregonian: Oregon transportation department investigating after MAX train derailment
Moreover, it seems that there were many reasons that TriMet should have expected that something could happen where the derailment occurred and checked to see if the track was OK:
- While most stretches of track see trains frequently, and there had even been some service overnight since the snow and ice fell, that track does not. Specifically, it is scheduled to be used just 3 times a day, at 4:17 AM, 4:47 AM and 5:02 AM, by Yellow Line trains going into service. So normally the last one through would have been 23 hours before.
- The track and adjacent switch are in the middle of a heavily used street intersection, with lots and lots of vehicles crossing that can push ice, snow, gravel and all manner of other debris into the grooves that the track consists of.
- TriMet added to the traffic by rerouting busy Line 8 and Line 77 buses through there in order to make room for a bike path through Rose Quarter.
- Nowhere else do switches or occasionally-used track get pounded by traffic. They are generally in areas where only trains travel, and many have no pavement around them, allowing debris to fall away.
- Even under normal conditions, that junction is problematic. Every evening when the opposite move is done to send Yellow Line trains home, a supervisor is dispatched in advance to the area in case the switches do not set correctly after not being used all day.
- Due to the fluctuating temperatures and heating from the sun, a layer of the snow/ice everywhere had been freezing hard, even on sidewalks where traffic is light weight.
- I recall a recent news article mentioning that MAX trains were spinning their wheels, which shows that it was already known that problems where happening.
- Seen on Twitter:
It seems that this was comparable to Halloween 2015 when MAX trains were sent into a pool of water deep enough that the insides were getting flooded.
Also, it should be noted that TriMet could have used an alternative route to get trains onto the Yellow Line, by either having them run reverse on the southbound-to-westbound track, or by sending them across the river and back to get to the eastbound-to-northbound track, both of which see regular use.
Lastly, while some may fault the operator, controller or maintainers for not checking on the track, the real responsibility lies with those above who's job it is to ensure everything gets done. Plus, many front-line employees feel demoralized, belittled and under pressure to keep going even without a storm, discouraging them from speaking up. And given that it was dark out, it may have been hard for the operator to closely see the tracks.
I have found a couple radio clips from just before the incident:
At 3:14 AM, about an hour before, maintainers were working right around where the train derailed, "to make sure these switches are ready to go in the morning".
At 4:15 AM, the operator of the incident train reported that workers were there, and asked if it was OK to go past them.
This shows that the operator was not the only one there, and raises various questions:
Did the maintainers ever look at or mention anything about the section of unusable track immediately next to the switches they were at?
Was the operator led to believe everything was checked and OK, or perhaps instructed to try going through?
What did the controller think, and was there confusion between the three parties involved?
It would be good to have a recording of the rest of the communications, as that would show what the operator, maintainers and controller had thought.