UPDATE: TriMet has complained that audio clips of communications don't tell the whole story and get taken out of context, but has responded by only releasing clips themselves--nothing better. I was expecting complete communications leading up to the service suspension that would give a good picture of what was going on and what operators felt they were supposed to do.
TriMet releases audio of driver, controller communications during rain storm (Portland Tribune)
TriMet releases audio from day of MAX flooding (KGW TV)
FURTHER UPDATE: MAX dispatch radio indicates problems before flooded train from The Oregonian has at least some real audio, and in it there doesn't seem to be any questioning about whether to continue or not. The article text ends with an operator being the one raising the issue and not wanting to go.
ALSO: KGW says it was either the operator's or manger's decision that the trains to go through the high water. This presumes that one of them made the decision on their own, which may not be the case.
In addition, there does not seem to be a clear, written protocol about when to shut down service due to high water; it seems reasonable to have one given that an inundation from heavy rains, broken water/sewer pipes, etc could easily occur.
My view is that I'm not sure it was clearly understood how high the water had become, and that controllers (and, by extension, managers) could/should have proactively discussed water heights. The train order could have easily included details, and Portland Streetcar started suspending service at last 15 minutes before the last MAX train went through the underpass. Instead, there seemed to be an underlying push to continue on, and operators believed they were to do so. Also, it's not healthy trying to assign blame.
-There's two separate issues: the trains going through the high water in the Morrison Bridge underpass, and 13 "trains" being out of service a few days later. It's unclear if these two are linked--whether the 13 "trains" were ones that had gone through that pool. I'm also unsure if they were damaged or held for inspection as a precaution.
-It is also unclear whether it is actually 13 2-car trains or 13 cars. It's confusing because they could easily decide to hold entire trains, and of course the newer trains have to have two cars. But the older cars can run individually. I don't think 13 trains were taken under the Morrison Bridge between the time the water got high and when they stopped going through.
-It would be good to know what the train order and other procedures say regarding water height. There definitely was a push to continue service, and I don't recall controllers asking about specific water heights. This article says:
"The operator's manual states that trains should stop if rails are invisible under standing water. If the rails are visible, proceed at no more than 10 miles per hour.Note that a controller's or supervisor's command could be seen as overriding that, and it doesn't say to never continue after stopping.
KGW obtained a copy of a directive, sent out that Saturday that says "Due to heavy rain through out the entire alignment all trains need to reduce their speed to 10 MPH in areas where there is noticeable standing water." There is no reference to whether tracks are visible or not."
-Regarding trains going through the Morrison Bridge underpass specifically:
--My recollection is that a supervisor had checked on the underpass and declared it OK, then was called somewhere else after which the water presumably had risen. Later operators may have questioned the water level, but I don't think the controller realized just how high the water now was (I believe TriMet has made a statement saying the water rose quickly). The supervisor did come back to it and "not recommend" that trains go through, at which time operations were suspended.
--Operators may have also assumed because the train ahead of theirs went through, they should too.
--It is not true to say that "managers ordered" trains to go through the water level as it was, as that sounds like a manager was on scene or at least clearly understood the specifics of the situation. Any OK from them was issued generally, and they may have been assuming there would only be a few inches of water. There may be some confusion in the media/union between "mangers" and controllers/supervisors.
--I do think the operator(s) should have stood up and resisted going through it (before one may have), but I do believe that they thought they were authorized or instructed to do so. It is not as if they decided to proceed all on their own.
--Operators should be familiar with the underpass, and able to grasp how high the water was. Watching this video, it was clearly way over the curbs and the tracks were far from visible. In addition, there is a switch/crossover at the bottom of the underpass, and there was no way to tell if it had become out of alignment (this stretch does not have signal protection, so the switch would need to be checked visually).
--The tracks are set in cement there, so there is little chance of them shifting. Also, I don't remember the underpass flooding before, and neither does TriMet.
--Besides the possible train damage and safety issues, passengers should expect the inside to stay dry.
-Portland Streetcar did shut down, and the restroom at their operator break room in the Pearl District was out of commission afterwards. Operators were told to use one at an adjacent business. Customer Service was also reporting standing water in the break room building at Rose Quarter.
-Saturday was not a good day for the deluge, as in addition to it being Halloween, there was a football game at Providence Park and a Blazers game. The MAX disruption affected attendees of both of those events.
-At the end of this article TriMet says "We will be releasing radio transmissions from Saturday afternoon as soon as they are available." It seems that they should be available immediately. It is not hard or expensive for an individual to set up a software-defined radio that can record all transmissions coming from their main radio site into time- and channel-stamped files, and a professional system should be able to do the same.
-Last (Tuesday) night, an operator got on the radio with one who had been gone and said the operator had to be "coaxed" to go through the flood waters. The controller got on the air and shut down the conversation.