Something Went Really Wrong

By on December 19, 2005

I read this weekend about the reservoir accident at the Taum Sauk Proffit Mountain Reservoir in Missouri. A section of the reservoir wall washed away early on December 14th, allowing a billion gallons of water to cascade down a hillside, washing away a house and nearly killing its occupants. Several cars & trucks on a highway that happened to be in the wrong place at the wrong time were also washed off the road. Nobody was killed, but still; a dang scary deal.

The Taum Sauk reservoir is used in generating electricity by provider AmerenUE; water is pumped into the reservoir at non-peak times (at night) and released to drive turbines & provide additional capacity when needed. The reservoir holds 1.5 billion gallons of water, and AmerenEU officials say that malfunctioning overflow sensors allowed too much water to be pumped into the reservoir, which overflowed and quickly eroded a 700 foot wide section of the reservoir wall. It also turns out that the section that washed away contained substandard fill and dirt instead of the granite that was specified by the engineers who designed it. Uh oh.

It seems strange to me that malfunctioning sensors are at fault in this accident. For one, there had to be more than one system watching over the water level. Something that critical ought to have several levels of redundancy, shouldn’t it?

Now, I have no intimate knowledge of how the water level maintenance system at Taum Sauk was designed or how it’s supposed to operate, but I would think that if one sensor malfunctioned, another sensor with a contradictory reading should have raised some kind of alert to inform an operator that something was up. And if it were me designing a system like that, contradictory readings from two sensors would shut down the pumps until a real live human being could step in and determine what the problem was.

But with this incedent, maybe the operators have seen these sensors malfunction before and have learned that the sensors can be ignored. It’s also possible that an overflow wasn’t considered a critical issue; that nobody envisioned an overflow situation would wash away a huge section of the levee and allow it to drain in 12 minutes. Who knows.

One thing is sure; sensors were in place that should have told someone that things were going wrong, and for whatever reason somebody either didn’t get the message or ignored it. And somebody else is going to be in the hot seat answering some difficult questions about why this happened, and in the end spending a lot of money getting things back to where they were on December 13.



  1. you read the RISKS digest, right?

    all the issues you mentioned are regularly ignored by engineers the world over, and the RISKS digest is there to document it…

  2. I live near and had adored the site of Taum Sauk. Even as a little child, I noted in my mind that there was water trickling down the side of the ‘hill’ on our way to climb the stairs and look over the top of the reservoir. I can remember thinking, ‘What if this thing ever broke open.’ I was in school when Ameren put the new turbines in and I was in high school when the new liner was put in for the leaks. Again, I thought about the possibility of a break. Three years later.Our ambulance crew went out on a ‘run’. I thought the medic said a broken hand at Johnson Shut Ins, so I went back inside and turned the weather on for a patient and learned that it wasn’t a broken hand but a broken ‘dam’. Expecting my first child whose grandmother and great-grandmother doesn’t live five miles from Johnson Shut Ins, I froze. I knew that it would be horrible, I remembered instantly the thoughts that I had had as a teen and a young child about the ‘trickles’ of water running down the side of the hill. I just knew that the wall finally gave and I was lost. Not being directly affected still didn’t mean that it didn’t have an effect on me.
    As more and more details came to light about the causes, althought Ameren did not directly cause the breech, they are still responsible in more ways than one and ignorance of mechanics is unacceptable. A family nearly lost thier lives and a county suffers from the taxes that help the economy.

  3. Ameren did directly cause the breach.

    They ignored some gauges, knew some were broken, and intentionally disabled others. Then they pumped water right over the wall until it failed.

    The only way it could have been more direct is if they had blown it up.

  4. After the breach, I was sent with the first drill team to work on site at Taum Sauk. It hasn’ been stated here, but I know some facts that haven’t been told. When Hurricane Katrina hit in Sept 2006, the on-site maintenance person called in with a report of water lieterally blowing over the top in the area that washed away. There were inspectors from FERC that visited and were concerned about the erosion caused from that situation. It was as simple as time ran out before FERC had the ability to shut it down so repairs could be made. It was a terrible accident, yes, but FERC has as much blame as Ameren in this.

Comments are closed. If you have something you really want to say, tweet @gadgetopia.