It's [1] an embankment dam, so basically just a pile of dirt/rocks. When the water goes over, the material of the damn goes with it [2]:

> Even a small sustained overtopping flow can remove thousands of tons of overburden soil from the mass of the dam within hours.

[1] https://en.wikipedia.org/wiki/Oroville_Dam

[2] https://en.wikipedia.org/wiki/Embankment_dam#Safety

Just too add: Water need not literally overtop the visible structure. The top few feet of the dam are not totally waterproof. Water pressure will get through material which is pourous enough to deal with rainwater and such. Rather than waves overtop, failure will star with small flows bleeding through near the top. These may run down inside the dam a little, only appearing on the surface a ways down the far side.

Here's a small-scale, but somewhat-satisfying view of what happens when an embankment dam is compromised.

Where?

Must be very small-scale, less than 1 pixel.

Here's one video showing the danger.

https://youtu.be/tkOBmiyTfEQ?t=8m20s

Basically, when the dam over-tops the water will cut a channel into which a little more water will flow and cut a little deeper into which more flows...and on and on.

It's a run away failure and cannot be stopped.

Here's a quicker example of how just a tiny breach at the top will drain the entire dam:

https://www.youtube.com/watch?v=hmKLxoAPU1c

Note in the first video it's the over-topping that causes the initial erosion/breach rather then the artificial one in the second video, but the mechanism of further failure is essentially the same.

I'm not sure the exact construction of the Oroville dam but it's clear the emergency spillway was concrete lined at the top likely with this sort of failure in mind.

I'd be surprised if the main bulk of the dam didn't have at least some measures to stop this kind of failure but someone with knowledge of it's construction can chime in on this point.

In 1986, water flowed over the top of the Auburn Cofferdam and resulted in complete failure of the dam. See this video of that incident [1] for an example of how the overtopping of an earthen dam leads to failure.

[1] https://www.youtube.com/watch?v=tDmwo5nsWfQ

This seems to show that it was the "spillway plug" that eroded and failed. I don't know what a spillway plug is, but the narration seems to imply that it is designed to fail in emergency situations.

In the beginning of the video, the narration talks about the spillway plug being a feature designed to erode in a fashion that delayed/slowed the inevitable failure of the entire dam. In this case, the spillway plug helped the dam release water in a semi-controlled fashion over a series of hours, instead of an uncontrolled fashion in a matter of minutes.

If a embankment dam is overtopped, it is going to fail. They designed it with a slightly lower section on one end so they could be sure if it failed, it failed in a predictable way.

Interesting footage. There's a good before/after comparison at the end:

https://youtu.be/tDmwo5nsWfQ?t=9m4s

Woah. Could you give me some more VCR, please?

This was the coffer dam for the construction of Auburn Dam, work on which was halted, but the coffer dam never removed?

Excellent video!

The main dam structure isn't built to withstand the water flow.

Wikimedia has a nice picture of the area:

https://upload.wikimedia.org/wikipedia/commons/d/d5/Oroville...

The emergency spillway is on the far left, it is lower than the rest of the dam. So if water is flowing over the main structure, it's anyway already failed.

It's an earthen dam. If water goes over the top, it will start to erode the dam, resulting in more water going over the top, resulting in catastrophic feedback.

Because it's an earthen dam, not concrete. Water going over the top will erode the dam, lowering the height, causing more water to go over, which causes more erosion...

My understanding is that water flowing rapidly down the vertical surface of the dam and impacting its base rapidly erodes the face and base of the dam, quickly reducing its ability to withstand the pressure behind it (which also must be at its maximum level for the dam to be overtopped.)

For an earthen dam, erosion of the top of the dam when overtopping is also, I believe, a significant issue, but overtopping is a catastrophic failure issue for any dam.

Water isn't passing over the dam, it's passing over a much shorter wall to the side of the dam with unprotected earth around and below it. Large volumes of water tend to wash away dirt and could potentially undermine this side wall. So there is a potential for release of a very large volume of water, without compromising the main dam itself, because the erosion cuts a new exit channel for the lake.

People have been saying that if the emergency spillway fails, it will release the top 30' (?) of the lake. What determines that height? Is there bedrock or something else expected to stop erosion 30' below the level of the emergency spillway or something?

Good question! According to this article, the bottom of the emergency spillway is indeed bedrock: http://www.drroyspencer.com/2017/02/why-the-oroville-dam-won...

Yes, it's basically on bedrock.

This image of the main spillway gives you an idea of what they're building on: http://i.dailymail.co.uk/i/pix/2017/02/13/08/3D252AE20000057...

You can see that they blasted into the rock to make the spillway.

However, the idea that erosion will undercut the emergency spillway and cause it fail implies that it's not built directly on the rock. This suggests that there could be more than 30' of draining.

The emergency spillway is a 30 ft concrete wall in the side of the lake.

If you look at the photos, some on Wikipedia, you can see the spillways are around a bend in the river from the main dam and pass over a high bedrock canyon wall.

>>Why does water passing over the dam destroy the dam?

>Water isn't passing over the dam

The question is about what happens when water passes over the dam.

If reverse Superman flew up and dropped a mountain in the spillways so the dam could be overtopped, the dam would begin to erode.

Whichever part of the dam was lowest would see the fastest flowing water and would erode the most quickly. As a chasm was cut down through the dam, the water would flow faster and faster through it - if the height of the water above the bottom of the cut is 1 foot, only the top foot of water is "pushing" the water over; if the cut is 10 feet deep, now the top 10 feet of water is "pushing" water out the hole, and it will be pushed much faster.

This will turn into a feedback loop, and the water will rapidly cut a channel through to the bottom of the dam. As water flows through it, faster and faster, it will weaken the sides of the chasm, and more of the dam will collapse. This will let more of the water flow through, but slower, eventually limiting further erosion.

This is essentially the same thing happening in the spillways right now, but slower, because the spillways are designed to resist erosion. Once the vicious cycle gets started, though, the end result is the same.

The dam is 922', but the emergency spillway is 901 feet and very, very wide. It's designed to make the dam impossible to overtop. Even the worst case scenario this year would not lead to overtopping the dam.

The question is about why water flowing over the dam would damage it. Clearly they have engineered the spillway to prevent that, so there must be a reason, which is why the person asked.

If someone asks you what would happen to the people inside of a submarine if they let it fill with water, you don't respond by saying they design submarines to prevent them from filling up with water.

erosion