-
1
- #1
msquared48
Structural
Member Spartan: Stage storage flow data here for those interested:
Mike McCann, PE, SE (WA)
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations JAE on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Oroville Dam Spillway Concrete Failure (Feather River Flooding, CA) 36
- Thread starter msquared48
- Start date
- Status
Member Spartan: Stage storage flow data here for those interested:
Mike McCann, PE, SE (WA)
Looking at the whole picture, I put my money on cavitation and some sloppy engineering design.
Let us keep Bernouli on the side for the moment, and bring Uncle Newton in. That area is below the curve in the spilway. The water want to continue straight, but eventually gravity will win,slamming that water mass down on the concrete.
Now the big problem. One never ever pour high load bearing concrete directly on bedrock,unless you dowl deep or pile support it. That concrete starts to "hop" on the bedrock and the cavitation hammers it away from the top.the result, a total failure of the concrete itself. Now the demolition continue on the bedrock.
That bedrock should have been removed way back, and layer filled that base foundation,compact each layer to 98 % mod, built it up for about 3 meters. Only then cast the concrete and build the spillway. Secondly, add synthetic fiber as additional re-inforcement in the concrete at a dose of about 5kg/m3,will reduce the amount of expansion joints and give the air bubbles less bite on the edges of the individual concrete slaps.
My 2 pennies for what it is worth.
Let us keep Bernouli on the side for the moment, and bring Uncle Newton in. That area is below the curve in the spilway. The water want to continue straight, but eventually gravity will win,slamming that water mass down on the concrete.
Now the big problem. One never ever pour high load bearing concrete directly on bedrock,unless you dowl deep or pile support it. That concrete starts to "hop" on the bedrock and the cavitation hammers it away from the top.the result, a total failure of the concrete itself. Now the demolition continue on the bedrock.
That bedrock should have been removed way back, and layer filled that base foundation,compact each layer to 98 % mod, built it up for about 3 meters. Only then cast the concrete and build the spillway. Secondly, add synthetic fiber as additional re-inforcement in the concrete at a dose of about 5kg/m3,will reduce the amount of expansion joints and give the air bubbles less bite on the edges of the individual concrete slaps.
My 2 pennies for what it is worth.
"Secondly, add synthetic fiber as additional re-inforcement in the concrete at a dose of about 5kg/m3,will reduce the amount of expansion joints and give the air bubbles less bite on the edges of the individual concrete slaps."
Not a structural guy or engineering historian but would that technique have been available when the spill way was built?
Not a structural guy or engineering historian but would that technique have been available when the spill way was built?
Posting guidelines faq731-376 (probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484
What is Engineering anyway: faq1088-1484
-
1
- #263
![URL]](https://res.cloudinary.com/engtips/image/fetch/w_800,c_lfill,q_auto,f_auto,g_faces:center/[URL unfurl="true"]http://www.infomercialscams.org/prod/flexseal-liquid-sealer.jpg[/URL])
![URL]](https://res.cloudinary.com/engtips/image/fetch/w_800,c_lfill,q_auto,f_auto,g_faces:center/[URL unfurl="true"]http://cdn1.thecomeback.com/wp-content/uploads/sites/72/2015/05/Untitlhed.png[/URL])
Hey RockLove your comments are interesting. Since when is a slab of concrete with even 8 feet of water a heavy load? Also tell us where compacted soil at a high percentage of Proctor is less likely to erode away than even the weaker rock at this site? My bet is the deep erosion areas will have a fill of well cemented material, likely similar to concrete. That would be much less likely to erode than your suggestion. With an estimated "design period" for the repair of three months, I hope all these possible deficiencies will be covered. Certainly the original design and construction really "missed the boat".
btrueblood
Mechanical
Didn't see your earlier post, sorry John. But, I'm also about 100 years off (and probably more) in my estimate for the development of energy dissipation and aeration designs in spillways...see:
Designed and built in 1880, but apparently still standing 100+ years later. People (not just our US Dept. of Reclamation - see Glen Canyon dam spillway erosion problems recurring as late as the 1983 floods) are still underestimating cavitation and erosion rates on spillways. Arrogance or ignorance, dunno which.
Designed and built in 1880, but apparently still standing 100+ years later. People (not just our US Dept. of Reclamation - see Glen Canyon dam spillway erosion problems recurring as late as the 1983 floods) are still underestimating cavitation and erosion rates on spillways. Arrogance or ignorance, dunno which.
To Oldest, the base would only corrode if the concrete on top is damaged,but it provide a much more stable foundation as what bedrock is. its not the 8 feet of water thats the problem. its the tiny air bubbles in the water that cause cavitation. In the past many believe cavitation is a minor only to realize now that it is still a nightmare if not taken in consideration.
No argument from my side. Well,whatever, going to be a very costly 3 months.I wonder what the cost is up to now since all this started this year.
No argument from my side. Well,whatever, going to be a very costly 3 months.I wonder what the cost is up to now since all this started this year.
Rocklve: Notice any of the photos of the spillway walls when it was operating. That water spouting out of the side walls is from under the slab, getting there from a herringbone layout of perforated clay tile pipes (shown in a few damage photos) under the slab FULL LENGTH. The whole length of spillway had leaks getting to those drain pipes. It seeped downward via MANY MANY shrinkage and joint cracks, not likely cavitation. Once under the slab and moving there, the ultra weak sections of the bedrock easily eroded, causing the failures. Any repair that does not address the erodability of the stuff under the slab is just half the job needed. In my view the entire remaining spillway needs to be torn out and these weak zones replaced with hard stuff, such as concrete. Fixing the cavities is a good idea, but not doing a complete replacement of crap under the slab will only result in another situation like this some day. Looking at the emergency spillway you see the rotten nature of the bedrock in this area.
HamburgerHelper
Electrical
I thought cavitation was caused by fast moving fluid that that has low pressure pockets that cause the liquid to turn to a gas. I thought the damage was done when the bubbles collapse back into a fluid when they move back into a higher pressure region. I didn't think air bubbles ,not cavitated water, caused damage.
rocklove... the problem comes from the water surface wanting to lift off the concrete, creating the negative pressure required for cavitation. Water slapping the concrete (other than the wave created) is not normally associated with cavitation.
Dik
Dik
racookpe1978
Nuclear
What troubles me is the large number of spillway failures. And the implications of those failures. And the root cause of those number of failures.
First, clearly the problem (failure of the concrete surface and substructure) ON and ABOVE the surface of the spillway walls at high flows is not being solved now, nor was it correctly approximated and solved in the past assumptions and construction. The combination of huge (but constantly, instantaneously varying) turbulent impact forces, combined with small but also continuously varying cavitation forces (the vacuum formation, then rapid re-collapse and micro-shocks as the bubbles re-form and re-collapse) is immense - and not be solvable at all even with today's higher-speed computational fluid dynamics FEA models. But even that is telling: If we cannot even determine consistently with the "pure theory" of these massive water flows, how do you design a system to withstand them under emergency conditions? If the models are not "accurate" to a sufficient degree to predict the specific multiple failures we see worldwide when the spillways are used at high flows, how do you re-design existing spillways to prevent a future loss? If "testing" breaks the spillways that are tested, you have to rely on the CFD models to go forward.
Second, the failure inside the walls between the impact and cavitation are not being absorbed reliably in what was the original (and repaired) concrete, rebar, and steel liner designs. Here at Oroville, if the mid-span conrete and rebar were adequate, the spillway would not have torn in half between the upper curve (a lower flow pressure area), and the lower straight discharge region with its spray breaks. Instead, the spillway floor tore out at the concrete expansion joint/construction joint.
Third, the failure BELOW the walls and floors between the spillway concrete and the rock below. Maybe that is inevitable: After all, if you don't know the underlaying material and so cannot estimate how much strength is needed to support the open "spans" between good support and weak support areas under the concrete being pounded by unknown forces from above, you can't design and pour the concrete around the any rebar network strong enough to absorb the blows. But, how many existing dams worldwide have been properly surveyed and prepared underneath? Almost as soon as the Hoover Dam was filling, the poured "joint" between the dam and the rock underneath began needing repairs and grout to seal leaks, strengthen the rock against tearing forces.
The tunnel walls under Hoover used steel liners to define the flow lines and reduce turbulence in the original bypass tunnels, which were then plugged at the mid-point and adapted as spillway tunnels. Expensive, but those liners might be needed many places.
Here, bare concrete over bad rock failed even before significant emergency flows were needed in a spillway - for a dam designed 30+ years after Hoover. Doesn't speak well for "lessons learned" and "better design analysis" for the dam industry over the slide rule era. Now, in the computer era, what else do we "not know we don't know"?
First, clearly the problem (failure of the concrete surface and substructure) ON and ABOVE the surface of the spillway walls at high flows is not being solved now, nor was it correctly approximated and solved in the past assumptions and construction. The combination of huge (but constantly, instantaneously varying) turbulent impact forces, combined with small but also continuously varying cavitation forces (the vacuum formation, then rapid re-collapse and micro-shocks as the bubbles re-form and re-collapse) is immense - and not be solvable at all even with today's higher-speed computational fluid dynamics FEA models. But even that is telling: If we cannot even determine consistently with the "pure theory" of these massive water flows, how do you design a system to withstand them under emergency conditions? If the models are not "accurate" to a sufficient degree to predict the specific multiple failures we see worldwide when the spillways are used at high flows, how do you re-design existing spillways to prevent a future loss? If "testing" breaks the spillways that are tested, you have to rely on the CFD models to go forward.
Second, the failure inside the walls between the impact and cavitation are not being absorbed reliably in what was the original (and repaired) concrete, rebar, and steel liner designs. Here at Oroville, if the mid-span conrete and rebar were adequate, the spillway would not have torn in half between the upper curve (a lower flow pressure area), and the lower straight discharge region with its spray breaks. Instead, the spillway floor tore out at the concrete expansion joint/construction joint.
Third, the failure BELOW the walls and floors between the spillway concrete and the rock below. Maybe that is inevitable: After all, if you don't know the underlaying material and so cannot estimate how much strength is needed to support the open "spans" between good support and weak support areas under the concrete being pounded by unknown forces from above, you can't design and pour the concrete around the any rebar network strong enough to absorb the blows. But, how many existing dams worldwide have been properly surveyed and prepared underneath? Almost as soon as the Hoover Dam was filling, the poured "joint" between the dam and the rock underneath began needing repairs and grout to seal leaks, strengthen the rock against tearing forces.
The tunnel walls under Hoover used steel liners to define the flow lines and reduce turbulence in the original bypass tunnels, which were then plugged at the mid-point and adapted as spillway tunnels. Expensive, but those liners might be needed many places.
Here, bare concrete over bad rock failed even before significant emergency flows were needed in a spillway - for a dam designed 30+ years after Hoover. Doesn't speak well for "lessons learned" and "better design analysis" for the dam industry over the slide rule era. Now, in the computer era, what else do we "not know we don't know"?
- Moderator
- #274
btrueblood
Mechanical
Air bubbles, purposely injected, help mitigate cavitation by limiting the minimum bubble size during collapse, and thus the peak pressure in collapsing bubbles. I.e., if you have air bubbles before cavitation onset, then the original bubbles grow under low pressure (cavitation) conditions, and upon collapse don't get much smaller than they started - and the air that was originally in the bubble acts as a spring, limiting peak collapse pressure. Conversely, in bubble-free flow, cavitation bubbles are filled with water vapor only, and upon collapse, the water vapor returns to liquid, the bubble becomes infinitely small and the peak pressure ridiculously high. See and its references, for starters.
As for your comments, racooke, agree mostly - except that the use of stepped (aerated)/baffled/energy-dissipating spillways has been a thing since 1880 at least. One can only surmise arrogance and ignorance of the designers precluded such designs at Oroville.
As for your comments, racooke, agree mostly - except that the use of stepped (aerated)/baffled/energy-dissipating spillways has been a thing since 1880 at least. One can only surmise arrogance and ignorance of the designers precluded such designs at Oroville.
MikeHalloran
Mechanical
Following the stepped spillway link, one can arrive at an entry for the Gold Creek Dam, which bears eerie similarities to the extant dam, and was first built with a simple chute spillway, which failed when actually used, in 1890.
I don't think the root cause is repeating history that you didn't study, or even arrogance; it's simple economics. A straightish chute is certainly cheaper to build than any sort of stepped structure, and the MBAs who typically end up steering projects would openly scoff at any assertion that tiny bubbles could eat concrete, and publicly deride any engineers who spoke up, while secretly figuring that:
- thanks to drought, the original design would 'function' long enough to deflect blame.
- once the used food hit the fan, the problem would be big enough to justify emptying bigger pockets.
Mike Halloran
Pembroke Pines, FL, USA
I don't think the root cause is repeating history that you didn't study, or even arrogance; it's simple economics. A straightish chute is certainly cheaper to build than any sort of stepped structure, and the MBAs who typically end up steering projects would openly scoff at any assertion that tiny bubbles could eat concrete, and publicly deride any engineers who spoke up, while secretly figuring that:
- thanks to drought, the original design would 'function' long enough to deflect blame.
- once the used food hit the fan, the problem would be big enough to justify emptying bigger pockets.
Mike Halloran
Pembroke Pines, FL, USA
btrueblood
Mechanical
Yeah, probably right Mike, except an 1890 failure is fairly easy to describe to pesky dis-believing MBA's, one would think.
... MBA's again, where's Snorgy?
... MBA's again, where's Snorgy?
I think way back people at projects from a different angle as what we do today. Seen it in some dams built by Even large foreign companies in AFRICA. The project was kept within the budget, no matter what come up. Then 20/30/40/or like in the case of KARIBA, 60 years later, it become somebody else's problem. Same story there. Cavitation was the main culprit and now to repair will cost 100,s of millions dollars. And in some cases, it was a case of lets take the money and walk. Again, making it somebody else's problem. Do anyone know what is the cost to date more or less on this episode on the spillway(excavations on soil washed into the river below). Then not to talk about the repairs. As I can see, this will not be cheap if they want to do it properly and repairs that last. I still try to work myself around the drainpipes under the spillway. My design might be much more expensive,but I guarantee that it would have been still cheaper then all this repairs to date.
Link to the KARIBA project. Problem described on page 16
Link to the KARIBA project. Problem described on page 16
racookpe1978
Nuclear
Construction projects have been falling down since the "Bent Pyramid" began falling apart because it was sloped too strongly.
But... The "theory" of massive, unconstrained random and turbulent (but I repeat myself, don't I ) water flows down a "simple" spillway MUST BE better evaluated, or we will lose the dams right at the very point when they are at extremely high levels with uncontrollable waters going down the spillway.
Compare it to a high pressure boiler or pressure vessel. Two relief valves are mandated by law, and by common sense. By law and by common sense, we do not operate pressure vessels above their maximum design pressure. BUT.
We cannot "regulate" water flows during flood conditions into a dam. We can, at best, make estimates of the maximum expected water flow and maximum expected snow pack levels and maximum allowed water levels before the rain begins and before the snow pack begins accumulating. We cannot "stop" the flood before, during over after the rainfall, nor prevent more water from coming downstream into an overloaded dam. We have a "pressure vessel" that - basically - is uncontrollable.
But that "pressure vessel" has been built with a relief valve that not only has never been tested under real world conditions, but cannot be tested before it is needed, and consistently, when that "approved officially designed" relief valve HAS BEEN used, it fails more often than not in most of the dams around the world!
But... The "theory" of massive, unconstrained random and turbulent (but I repeat myself, don't I ) water flows down a "simple" spillway MUST BE better evaluated, or we will lose the dams right at the very point when they are at extremely high levels with uncontrollable waters going down the spillway.
Compare it to a high pressure boiler or pressure vessel. Two relief valves are mandated by law, and by common sense. By law and by common sense, we do not operate pressure vessels above their maximum design pressure. BUT.
We cannot "regulate" water flows during flood conditions into a dam. We can, at best, make estimates of the maximum expected water flow and maximum expected snow pack levels and maximum allowed water levels before the rain begins and before the snow pack begins accumulating. We cannot "stop" the flood before, during over after the rainfall, nor prevent more water from coming downstream into an overloaded dam. We have a "pressure vessel" that - basically - is uncontrollable.
But that "pressure vessel" has been built with a relief valve that not only has never been tested under real world conditions, but cannot be tested before it is needed, and consistently, when that "approved officially designed" relief valve HAS BEEN used, it fails more often than not in most of the dams around the world!
There's been a lot of supposition that "cavitation" is the problem, which remains to be proven. Note that when the spillway was used c2006, with high flow rates, there presumably was not a specific problem associated with the spillway, since it was then used in 2011 with no obvious deleterious effects. The initial hole development occurred with only 40-60 kcft/s flow for less than a day on 2/6, and that was a fairly substantial failure of the spillway surface, which then caused a shutdown of the spillway. Prior to that, the flow had been 10-30 kcft/s, on and off, for almost a month, but it was mostly at or under 20 kcft/s until 2/3.
It seems to me that the damage had to have been from the drought and the heavy rains this year that undermined the spillway, just simply because there hadn't been any sort of flow that would have experienced much cavitation prior to 2/7, going all the way back to 2006
TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! faq731-376 forum1529 Entire Forum list
It seems to me that the damage had to have been from the drought and the heavy rains this year that undermined the spillway, just simply because there hadn't been any sort of flow that would have experienced much cavitation prior to 2/7, going all the way back to 2006
TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! faq731-376 forum1529 Entire Forum list
- Status
Similar threads
