Oroville Dam Spillway Failure

[Jody Bourgeois]

Hoover Dam spillway at work (in the lower left corner of the picture.)

visited Hoover Dam last year with students; my understanding, from displays, is that the only time this spillway was "at work" was during the 1983 floods?

 

[Scott Gates]

Hi, new here. Lowland archaeologist without real knowledge of reservoirs. Reading and learning a lot here.

When I saw post #883, I wondered if the big concrete dents would still be there. In some pictures, the lowest part of the spillway seems not completely destroyed. I see a smooth flow, indicating a bit of concrete bed, in the upper right part of the picture in post #911, and what looks like the leftmost dent. If they're still there, would they still help lessening the impact of the water on the outflow pond and its other bank? Not contibuting to more erosion there, that might add to the build-up that's a problem for the power plant?
Or is the general water flow so chaotic that any surviving dent doesn't make a difference there anymore?

I can't put the image of post #911 here (obsolete mac with only partially functional browsers). Sorry about that.


Edit to add: https://www.metabunk.org/oroville-dam-main-spillway-waterfall-erosion-watch.t8402/page-3, post #85, new picture shows that at least three dents are still there, and it looks like they do spray the part of the wather that flows at the left side well enough that it doesn't slam into the far bank of the pond. Nor, looks like, the water from the right side. So at least that doesn't add to the problems for the power plant.

Thanks mod, for adding the picture from #911!

mod add: photo from 911
View attachment 24856

The "new flow" down the hillside some were worried about earlier today ... simply a flow that has been running for days thru the breech in main spillway wall on emergency spillway side ... has become more pronounced with the rain, some likely additional opening downslope with embankment sloughing, and potentially because they have SLOWED the rate from 100,000 down to 70,000 cfs - which makes it easier for water to accumulate at the wall break than at higher flows ...

 

[deirdre]

It's a challenge to keep up with this site! Perhaps this topic lies among all the pages. What role might cavitation have played in the original spillway failure? I know that cavitation was implicated in the near-failure of the Lake Powell Dam system in the 1983 floods.

try typing 'cavitation' into the search bar. as people might not want to repeat the whole conversation.

 

[sushi]

Water flows downhill. Water flowing downhill cannot get to the dam from the spillway damage location.

And pictures of the underlying ground conditions in the damage area show the left side of the spillway is blue-green bedrock.

I'm not concerned about the water reaching the dam. I do not think that is possible. I am trying to determine if a belt of weak rock runs across the face of the slope. I am not sure if these are called tectonic pipes or fault layers, or veins, but they are clearly evident at various locations across the site and appear to be associated with less competent rock. A poster with experience rock climbing in the area stated that this rock is friable and will not support the weight of a human being. Knowing the stratigraphy of the site would help identify other possible areas of weakness.

 

[Twominds]

The "new flow" down the hillside some were worried about earlier today ... simply a flow that has been running for days thru the breech in main spillway wall on emergency spillway side ... has become more pronounced with the rain, some likely additional opening downslope with embankment sloughing, and potentially because they have SLOWED the rate from 100,000 down to 70,000 cfs - which makes it easier for water to accumulate at the wall break than at higher flows ...

View attachment 24868

Thanks for this excellent picture. It's even better than the one in the Spillway Erosion thread.

I was wondering about the big concrete dents at the bottom of the spillway, and the possibility of erosion at the opposite bank of the Feather River where the spillway water ends. I can see two black dots to the left of the big new island in the waterfall, that look like dents, and they seem to do their work of spraying the water so the opposite bank won't be pounded by it and adding to the erosion.

These photo's have answered my original question.

 

[KenMH]

I'm not concerned about the water reaching the dam. I do not think that is possible. I am trying to determine if a belt of weak rock runs across the face of the slope. I am not sure if these are called tectonic pipes or fault layers, or veins, but they are clearly evident at various locations across the site and appear to be associated with less competent rock. A poster with experience rock climbing in the area stated that this rock is friable and will not support the weight of a human being. Knowing the stratigraphy of the site would help identify other possible areas of weakness.

I was the poster that has been to the site before and compared what I saw to my rock climbing experience, but I can only say the exposed weathered rock is bad, the blue grey bedrock appears to be solid. As i said earlier I went there but I did not dig down to bedrock or establish the rock's resistance to hydraulic separation and transport. Check Rock Whisperer's post #922. that post has a link to a geologist that visited the site and makes educated assessments with photographic examples. From what I understand of geology yes there could be more areas of weakness, but without a thorough geophysical survey including ground penetrating radar, geomagnetic mapping, and induced seismicity mapping, it's hard to tell. None of those methods was available when the dam was built, cores were drilled and men in white button down shirts with pocket protectors examined them meticulously, but they can ether reveal or deceive based on if the core pattern drilled over an area was lucky to hit relevant geology or unlucky and missed something.

 

[deirdre]

I'm not concerned about the water reaching the dam. I do not think that is possible. I am trying to determine if a belt of weak rock runs across the face of the slope. I am not sure if these are called tectonic pipes or fault layers, or veins, but they are clearly evident at various locations across the site and appear to be associated with less competent rock. A poster with experience rock climbing in the area stated that this rock is friable and will not support the weight of a human being. Knowing the stratigraphy of the site would help identify other possible areas of weakness.

Dont paraphrase other posters. It is against posting guidelines. What he said is:

That is all amautuer opinion, I have been there but not dug down to bedrock

 

[Junkie]

I am with you here. I am fascinated by dams and their engineering (bridges too), and maybe it's possible to fulfill this criterion If you think you may need it, you have to test it for some (certainly not all) of those cases. I am thinking about earthquake engineering and also tsunami engineering such as sometimes worked and sometimes failed in 2011 northern Honshu. Or what happened in Christchurch. That is, how do you test for an earthquake or a tsunami? Granted you can do lab tests and also design in factors of safety, but seems like engineering for disasters requires design beyond what can in actuality be "tested"

I'm not aware of any full scale testing on modern bridges. Obviously it's nearly impossible to do full scale seismic testing (the energy involved is enormous), but even live load testing just doesn't happen (at least not often). Modern steel and concrete is very consistent, which may be why they don't bother.

Geotechnical engineering is a much less exact science, though, which is probably why they're more likely to do testing (destructive pull-out testing of anchors and the like).

 

[wrorke]

I was the poster that has been to the site before and compared what I saw to my rock climbing experience, but I can only say the exposed weathered rock is bad, the blue grey bedrock appears to be solid. As i said earlier I went there but I did not dig down to bedrock or establish the rock's resistance to hydraulic separation and transport. Check Rock Whisperer's post #922. that post has a link to a geologist that visited the site and makes educated assessments with photographic examples. From what I understand of geology yes there could be more areas of weakness, but without a thorough geophysical survey including ground penetrating radar, geomagnetic mapping, and induced seismicity mapping, it's hard to tell. None of those methods was available when the dam was built, cores were drilled and men in white button down shirts with pocket protectors examined them meticulously, but they can ether reveal or deceive based on if the core pattern drilled over an area was lucky to hit relevant geology or unlucky and missed something.

Apparently not an overly dense pattern of core drills near the spillway

Note the "EXPLANATION" in the bottom right that shows the symbols for the core samples.

https://ia800302.us.archive.org/3/i...3calirich/zh9californiastatew2003calirich.pdf
Document page 94 (pg 146 of 546 in the .pdf)

Looks like there were only three core samples near the failure

 

[Ethan O'Connor]

Apparently not an overly dense pattern of core drills near the spillway
Looks like there were only three core samples near the failure
View attachment 24883

Here's this graphic overlaid on the 4/14/2015 Google Earth Imagery:

 

[Tom W]

Not sure if this came up before, but the google Earth image from 4/14/2015 shows a lot of patches on the join at the top of the original failed area.
View attachment 24867
They actually look like rectangular patches, like someone sawed out spalling concrete to make a better patch. Like
View attachment 24872
You can verify this is the join by the pattern of white repairs on the join above.
View attachment 24869

Just an observation - the failed area is just downhill from the point where the chute turns downward. I wonder if there may be extra load placed on that area from the falling water where it might hit the panel with some downward velocity after it goes over the hill where the "bottom falls out". Just a thought.

 

[wrorke]

It looks like the bottom slabs used step joints where one slab overlapped the other. The depth of the step was only about six inches.

More worrisome -- where the slabs are broken, the rebar appears pristine and there is much damage to the concrete -- the rebar did its job. But at the joints there is virtually no sign of any reinforcing spanning the joints.


Joint -- no rebar

Joint -- no rebar (but lots of apparently empty dowel holes)

Joint -- no rebar

Joint -- no rebar

Broken slab -- lots of like-new rebar.

In some places there appears to be rebar ties into the ground

Other than these widely spaced anchors, it appears the slabs were poured on fill. There is no evidence of concrete adhered to the rock.

No evidence of sealing, although the 100k cfs rinse cycle might have cleared any evidence of that.

I do not understand the lack of ties between the slabs. And so few anchors on a slope.

 

[Herb Snart]

New member here to this interesting site. I've followed this thread from the start and have had numerous questions, some have been answered, some haven't as far I know. The thread goes faster than I have had time to read, I'm only to page 16 and it's up to 24 now.

Back on Monday, in post 345 concerning 'operating rules' of the dam, https://www.metabunk.org/oroville-dam-spillway-failure.t8381/page-9 under ''Limitation on releases" it says "C. Releases from Oroville dam are not to be increased more than 10,000 c.f.s nor decreased more than 5,000 c.f.s in any 2-hour period." (I'm assuming it's the main spillway they are referring to)

An increase from 0 cfs to 100,000 cfs should take about 20 hours and a decrease back to 0 should take about 40 hours according to that. I don't believe they followed a schedule even close to that, but I could be wrong.

The rules didn't state why that schedule should be followed. I wonder if it's to allow the temperature of the spillway to gradually adjust to avoid creating cracks.

Severe problems develop in massive structures where heat cannot be dissipated. Thermal contraction on the concrete's surface without a corresponding change in its interior temperature will cause a thermal differential and potentially lead to cracking. Temperature changes that result in shortening will crack concrete members that are held in place or restrained by another part of the structure, internal reinforcement or by the ground. For example, a long restrained concrete section is allowed to drop in temperature. As the temperature drops, the concrete tends to shorten, but cannot as it is restrained along its base length. This causes the concrete to be stressed, and eventually crack.

http://www.engr.psu.edu/ce/courses/...ermalexpansioncontraction/thermalexpcontr.htm

It seems possible the spillway problem could have been caused by human error more than anything to do with the underlying soil. The post about the accident caused by opening those valves to 100% (post #447) https://www.dir.ca.gov/dosh/citations/CA Water Resources 313228637Summary.pdf showcases that the DWR didn't much care about the regulations concerning operating the dam and had, (and maybe still have) little knowledge of the possible effects of whatever they did to it would have.

Another thing is that big rock at the bottom of the spillway wasn't there to begin with and was washed out and rolled there within a matter of hours after reopening the spillway gates. That seems to show the mountain isn't all that homogeneous and is laced with easily erodible veins.

 

[Michael G]

visited Hoover Dam last year with students; my understanding, from displays, is that the only time this spillway was "at work" was during the 1983 floods?

Correct. It was tested in 1941 by limiting outflows and allowing the level to rise to maximum before opening the spillways. Erosion problems due to cavitation were encountered in the tunnels after the first test and were repaired. Nonetheless, the 1983 floods caused additional damage which also had to be repaired. The performance was deemed acceptable though, because the structures were never at risk, even in that unusual event. There are acceptable levels of damage to the spillways (and other structures in the system) during extreme events. In theory, they could be eliminated but in practice there's always an economic tradeoff: spend a billion dollars to make it perfect, or spend $200m every couple of decades to repair something that's imperfect? And keep in mind that "perfect" is only as perfect as you imagine it needs to be and mother nature may disagree with your judgement.

Glen Canyon, on the other hand, could have easily failed and the spillways required significant upgrades (including an air slot to eliminate cavitation risk). That's the one I have least confidence in, but unless the water situation on the Colorado changes significantly, it's not going to be much of an issue.

 

[wrorke]

In some places there appears to be rebar ties into the ground
View attachment 24892

Other than these widely spaced anchors, ..

I want to retract my assertion regarding the rebar ties into the rock. I only see the survivors. I have no basis for knowing how many were used -- they may have all gotten pulled out with the departing concrete.

 

[mtchap]

What might this pipe be?
View attachment 24572

New to the site...this has been a fantastic and mesmerizing thread to follow so far, well done to all! I don't have a whole lot to add at this point (mechanical engineer, not a civil engineer), but there are two posts from earlier that I do have some input on.

For the first post, quoted above, I believe the "not a light pole" to be a short drain. In the picture below (http://pixel-ca-dwr.photoshelter.co...wA/DK-Oro-Spillway-damage-3955-02-15-2017-jpg), you can see that the pipe continues downhill beneath the closest roadway, travels on for a short distance beyond the road (maybe 20 feet?) and then abruptly ends...

This picture from 2-11 (http://pixel-ca-dwr.photoshelter.co...pDAW7xUDs/ZC-Oro-Spillway-0070-02-11-2017-jpg) appears to show the drain in place and in action (water can be seen exiting the pipe when zoomed way in)...

An earlier picture from drier days (no date) provided by J.W. Wolf in post #208 shows the same pipe...

And finally, the original 1968 Dedication picture (http://pixel-ca-dwr.photoshelter.co...WU/Oroville-Dedication-3574-37-05-04-1968-jpg) provides some additional details that might shed some light on the original design intent. If you look closely at the base of the wier in the below picture, you can see what looks to me to be a small, dished bowl that the opening of the drain pipe is located in. I believe the intent of this drain is to capture small flows making their way downhill along the base of the wier and divert them away from the wier instead of allowing them to cascade over the edge and down into the lower level. Perhaps the thought was that even small flows over this edge over an extended period of time would cause erosion? This would seem to be kind of an afterthought fix considering the grand engineering that went into the rest of the dam, but I can picture someone looking at this hill at the last minute and deciding that a drain was probably needed there...

Two final observations concerning my theory that this was a last minute hack fix... 1) if you click on the quoted picture in Mick West's post, you can see that when the pipe first goes beneath the roadway, it enters what looks to me to be plain square box steel conduit. This would've likely been to keep the roadway from crushing the drain pipe, but there's nothing at all professional about the installation, and doesn't seem likely to me that a decent engineer would've designed it that way, and 2) if this drain were a part of the original plans for the dam, I feel it's likely that the "dished bowl" would've been engineered, concreted into place and would've been integrated much more seamlessly into the design of the base of the wier.

 

[SFX]

Returning to an earlier point.

Looks the same to me. Perhaps you are reading too much into a photocopied photo.

I don't think so. The postcard photo clearly shows the original rock at the base of the spillway.

The spillway rocks in question are clearly pretty robust, or they would have eroded away years ago.

They eroded quite a bit, and this was from just the occasional use of the spillway, not a constant battering at 100,00 cfs. It's pretty obvious the extreme erosion back to the concrete (large crevasse on the left) is actual erosion of the rock from water.

Not sure if this goes here or the other thread. If so, please move it to the appropriate thread. (the discussion started in this thread, based on the old photocopy of a photo Mick found in an old document).

I noted the rock does not look anything like that now.

"Now" meaning before the spillway failure. There has obviously been erosion of the rock at the end of the spillway.

Logic says the rock at the base of the spillway was assumed to be able to handle the erosion, since if it couldn't it would undermine the end of the concrete spillway, allowing failure there. (my opinion of course)

If this goes in the other thread, I apologize in advance,.

 

[KenMH]

Here is my current hypothesis of the conditions and events that led to this situation. All just speculation based on preliminary evidence and vaguely legitimate looking links to documents that seem complicated enough to be government agency public policy.

I suspect long term seepage during years of operation combined with geologic conditions, weaker chemically weathered rock ether from ancient hydrothermal venting, intrusion of geothermal features during a later time period, or recent weathering from groundwater intrusion, caused the undermining of the main spillway.

I believe the drain system under the main spillway is in question and likely contributed to the event. My reasoning is influenced by photographic evidence presented on MetaBunk of the drains not functioning in key areas, as well as descriptions of spillway failures in the Bureau of Reclamation dam spillway design document here (PDF page 105-106, document page 3-88 3-89).

I suspect 'Stagnation Pressure' as described by the Bureau of Reclamation,

Stagnation pressure refers to two conditions that can result in damage and/or failure of the spillway: (1) High velocity, high pressure flows enter cracks or open joints in the spillway flow surface (such as a chute), which results in uplift pressure that lifts (displaces) portions of the spillway conveyance feature; and (2) High velocity, high pressure flows enter the foundation through cracks or open joints in the spillway flow surface, which results in internal erosion of the foundation and loss of support of portions of the spillway conveyance feature [26].

is the accelerator of the failure, with the underlying geologic conditions initiating the erosion under the spillway.

A likely series of events I conceptualize is as follows:

Long term seepage caused weaker rock to slump, causing cracks and gaps and slight height variations in some of the spillway concrete slabs. Thermal changes caused normal expansion and contraction but the underlying geology caused slabs to shift slightly vertically during these.

Water was able to create a high pressure flow under the spillway through a gap and even a minor a height offset of spillway concrete slabs.

For some unknown reason the drainage pipes failed in the area eroded by stagnation pressure and caused further erosion of the sections those pipes served. I believe this is the real culprit that caused a large area to fail instead of just a disruption of flow and a small spout that would indicate a problem to be fixed by the DWR. I suspect DWR expected to see the edge of a slab fail and cause a minor spouting or surface wave which they could keep running for a time with and then fix, not whole sections of the spillway floor to drop away.

The unsupported floor of the spillway failed to support the water pressure as the water drops from the flatter portion to the steeper portion of the spillway and collapsed.

After that point it's all on video, so I will stop speculating there.

I reserve the right to change my mind! But that's the way it's looking from this perspective as of now.

 

[Shadowwalker]

If they are damming the backflow in the pond then that must be in case of a quite large release of water. But is there some other explanation for these dozers?[/QUOTE]

Can anyone tell me what those 45 degree cracks are below the dozers? They seem to go through rock and would intersect the dozers right in the middle of the trenches they are working on. They don't look like they came from drainage

 

[sushi]

Another thing is that big rock at the bottom of the spillway wasn't there to begin with and was washed out and rolled there within a matter of hours after reopening the spillway gates. That seems to show the mountain isn't all that homogeneous and is laced with easily erodible veins.

Herb - I think that rock that you mention at the bottom of the Flood Control Spillway didn't roll there but emerged when the surrounding soil was eroded away.

The first erosion channel was along the right spillway wall (facing upslope toward the sluice gates). After they increased the release to 100 k CFS then the additional scour removed more soil and created a new channel to the right of the rock. That big rock appears to be composed of the hard blue / grey bedrock associated with the site. The material around it appears to be composed of a red / orange weathered rock which erodes very easily.

I agree with your observation that the site appears to be laced with easily erodible veins of poor quality rock.

 

[CRM114]

When I first started learning about the Oroville spillway damage, I had heard the spillway discharge had been higher in the past from an article like this which said the releases in 1998 and 2006 were higher (up to 150,000):

http://www.orovillemr.com/article/zz/20110321/NEWS/110323052

The data for the spillway does not bear that out:

In fact it shows this recent release was higher. This is relevant to potential cavitation, since cavitation risk goes up with flowrate. There could have been events pre-1998 where the flow was higher, but I am unable to locate. Anyone have any info.

Update: I think they just misspoke; the release of 1986 was 150,000:

http://pixel-ca-dwr.photoshelter.co...gQ/NH-Oroville-86-Flood-6590-2-02-21-1986-jpg

 

[EricL]

Which page?. I can explain how to extract images.

Okay, here's some info regarding that site that seemed to have built-in methods for blocking the ways in which I'd normally snatch photos or copy links.

First, here's the main page:

http://pixel-ca-dwr.photoshelter.com/index

Once there, go to "Galleries", then go to the subsection called "Oroville Spillway Damage"

Many of the photos in this gallery have nothing to do with spillway damage, but instead show the spillway and other parts of the dam in good working order.

There are too many photos in this gallery to direct a person to any one of them by counting through the list, and it appears they are adding photos frequently so that would not be a reliable method in any case. In addition, the numbering/naming system for the photos is inconsistent as well, so the name is of partial use. The method I use here identify photos is to first give its position in the list (as indicated by the position of the scroll bar), followed by the photo name.

Near the center of the current collection of photos in this gallery are two that are called FL-Oroville-1187.jpg and FL-Oroville-1192.jpg, which provide a nice perspective of the first few drain outlets at the top end of the spillway, showing that they clearly are higher than the surrounding terrain. This is especially evident on the right side of the spillway. At the time I was examining these photos, this was the kind of thing I was looking for to support the idea that the drains remove water from beneath the spillway itself, not the adjacent terrain as some people said were very certain was the case. On that note, see that the very first drain is quite far from the start of the spillway, as it would have to be if carrying water from beneath the floor on a section having such slight slope.

Near the end of the collection (about 80-percent of the way through, going by the position of the scroll bar), there's a shot called BB-Oroville-0458 that also shows the nature of those first few drain outlets quite well.

There are also many good shots of the damage at various stages of progression. About 90 percent of the way through the list is one called "KG_oro-spillway-damage-10448" (the last part of the file name does not show, but it's the last photo in a list of a few with "KG_oroville-spillway-damage..." showing in the title). This shot would likely be helpful in answering the question here:

Notice on the right side how much soil has sloughed off the embankment. Did water splash out of the spillway and cause that erosion, or was it from a faulty drain?

Okay, I'll be interested in seeing what anyone comes up with, for snatching info from this site to post here. The screenshot method I saw mentioned earlier sounds promising, but I'm wondering if there's something better.

There are several other galleries, some with historic photos of construction which are quite interesting.

 

[MortarBoarder]

An increase from 0 cfs to 100,000 cfs should take about 20 hours and a decrease back to 0 should take about 40 hours according to that. I don't believe they followed a schedule even close to that, but I could be wrong.

Apparently during the emergency they didn't follow that procedure. Went from 55,000 cfs to 100,000 cfs in less than 3 hours, maybe 1 or 2.

02/12/2017 15:00 54904
02/12/2017 16:00 65117
02/12/2017 17:00 0
02/12/2017 18:00 99969


http://cdec.water.ca.gov/cgi-progs/queryF?s=ORO&d=12-Feb-2017+16:07&span=25hours
http://cdec.water.ca.gov/cgi-progs/queryF?s=ORO&d=13-Feb-2017+17:07&span=25hours

 

[651pe]

When I first started learning about the Oroville spillway damage, I had heard the spillway discharge had been higher in the past from an article like this which said the releases in 1998 and 2006 were higher (up to 150,000):

http://www.orovillemr.com/article/zz/20110321/NEWS/110323052

The data for the spillway does not bear that out:
View attachment 24899

In fact it shows this recent release was higher. This is relevant to potential cavitation, since cavitation risk goes up with flowrate. There could have been events pre-1998 where the flow was higher, but I am unable to locate. Anyone have any info.

Update: I think they just misspoke; the release of 1986 was 150,000:

http://pixel-ca-dwr.photoshelter.co...gQ/NH-Oroville-86-Flood-6590-2-02-21-1986-jpg

The graph you show... are the datapoints daily or hourly averages?

 

[Tom W]

Apparently during the emergency they didn't follow that procedure. Went from 55,000 cfs to 100,000 cfs in less than 3 hours, maybe 1 or 2.

I don't think that there was much choice at that juncture. It became obvious through the course on Sunday that the "Auxiliary/Emergency" spillway was in trouble rather quickly.

 

[EricL]

I suggest an unproven hypothesis -- over years, slow water permeation under the slab slowly eroded support eventually leading to enough shift to break one or more drain pipes. Flow through that drain pipe accelerated the erosion and undermining. This is consistent with the seepage at the sinkhole origin.

I agree with the logic here. I've seen other posts suggesting a broken drain must have been responsible, but that begs the question, why would a drain pipe break? The usual answer would be that it would break due to stress from movement of the soil and overlying structure, and that could happen once the initially slow process of erosion had progressed for a time. Another reason this makes sense is that this was a location where the subgrade was soft enough to erode easily. If initial shifting of soil and the slab led to a broken pipe, that would then have multiplied the severity of an already-existing problem.

 

[EricL]

I believe the drain system under the main spillway is in question and likely contributed to the event. My reasoning is influenced by photographic evidence presented on MetaBunk of the drains not functioning in key areas, as well as descriptions of spillway failures in the Bureau of Reclamation dam spillway design document here (PDF page 105-106, document page 3-88 3-89).

Could you explain whether or not it seems most likely to you that the drains stopped functioning at that location instead of the simpler idea that the once erosion due to massive downhill seepage beneath the slab provided an easier route for the water to follow, that's the route that it followed so it simply bypassed those particular drains? Water will take the easiest route, given various choices. I'm willing to believe either reason is possible, but I'm not ready to dismiss the one that seems to be the simplest in principle.

Here it appears that the weathered rock extended deeply enough, and with a downhill slope to the right (remember, the weathered rock is already largely missing at the time this photo was taken, so there's a bit of a 3-D profile of the weathered zone in which new eroded pathways could have been established), that once some erosion or piping had occurred, the exit path could have been downhill to the right, leaving the structure entirely (not the only possible option of course). Water would not follow a gently-sloped drain pipe if it had an easier downhill option such as what seems possible in this case.

 

[CRM114]

The graph you show... are the datapoints daily or hourly averages?

Daily.

 

[deirdre]

About 90 percent of the way through the list is one called "KG_oro-spillway-damage-10448"

you just click on the picture, then copy the url and paste it here.
http://pixel-ca-dwr.photoshelter.co...4/KG-oro-spillway-damage-10448-02-08-2017-jpg

 

[Chris Cothrun]

Near the center of the current collection of photos in this gallery are two that are called FL-Oroville-1187.jpg and FL-Oroville-1192.jpg, which provide a nice perspective of the first few drain outlets at the top end of the spillway, showing that they clearly are higher than the surrounding terrain.

Near the end of the collection (about 80-percent of the way through, going by the position of the scroll bar), there's a shot called BB-Oroville-0458 that also shows the nature of those first few drain outlets quite well.

There are also many good shots of the damage at various stages of progression. About 90 percent of the way through the list is one called "KG_oro-spillway-damage-10448" (the last part of the file name does not show, but it's the last photo in a list of a few with "KG_oroville-spillway-damage..." showing in the title). This shot would likely be helpful in answering the question here:

Notice on the right side how much soil has sloughed off the embankment. Did water splash out of the spillway and cause that erosion, or was it from a faulty drain?

Quick instructions for grabbing images if their signup and download system isn't working for you:

1. Using the Chrome browser, navigate to the image you're interested in.
2. Press F12 to open the developer tools.
3. Press Shift-Ctrl-C to activate the element inspector. Chrome will start highlighting parts of the page as you mouse over them.
4. Click the image you're interested in. That portion of the HTML should be highlighted in the developer tools.
5. Near the end of the <div> containing the image is the image URL. Doubleclick on that to put the Chrome developer tools in selection mode.
6. Carefully reselect just the image url at photoshelter.com. Get everything inside the double-quotes.
7. Paste this URL into a new tab and adjust the number at the end after the /fit=#### to some large value.
8. Press enter and a higher resolution image will appear that you'll be able to download and save for posting here.

 

[Mick West]

Quick instructions for grabbing images if their signup and download system isn't working for you:

1. Using the Chrome browser, navigate to the image you're interested in

....

That works, but it only gives you a 2040 pixel wide image at most. Unfortunately to get the full resolution images you need to sign up. As I noted earlier it took me a few hours before I could download anything. So try signing up now and see tomorrow.

While a 2040 is a good screen resolution, the full images can be over 7,000 pixels wide, which make all the difference when you zoom in. For example:

 

[Cocos]

Just a lurker here
But can anyone answer what this is?
Thanks

 

[Shadowwalker]

https://archive.org/stream/august11975orovi20378cali#page/292/mode/2up
Don't know if this is the start of the underground pipe system but since it's at the spillway gate floor it might be (Pg282)

 

[Mick West]

Just a lurker here
But can anyone answer what this is?
Thanks
View attachment 24908

Seems to be one end of a drain.
http://pixel-ca-dwr.photoshelter.co...uk/DK-Oro-Spillway-damage-3990-02-15-2017-jpg

This one:
http://pixel-ca-dwr.photoshelter.co...g/KG-oro-spillway-damage-10576-02-09-2017-jpg

 

[sushi]

[Broken External Image]:http://pixel-ca-dwr.photoshelter.co...4/KG-oro-spillway-damage-10448-02-08-2017-jpg

Attempting to insert the image URL as described at https://www.metabunk.org/oroville-dam-spillway-failure.t8381/page-25#post-201218

Not sure that the image is going to load. It normally shows a preview when editing the post prior to saving it to the site. All I am getting is a glyph with an icon and

The image in the post is very interesting for the following reasons:

1 - On the left side of the spillway above the area of the damage there appear to be boils of white water erupting on the spillway surface.

2 - On the right side of the spillway in the area of damage there appear to be brown boils of water and the entire right side of the spillway is stained the reddish brown associated with the loose soil or weathered rock.

The white boils suggest a lifted slab edge with the spillway flow hitting the lifted edge and some quantity of this water entering beneath the slab. This same water may be exiting down-slop of the right. Some of the water creates the brown boil on the slab surface and some of the water is exiting under the sidewall and creating a new drainage channel to the right of the spillway.

EDIT
When I try and load the image using the image URL shown in the post I get a message indicating a DNS resolution issue.

 

[Mick West]

Interesting article from McClatchy on Wednesday, Feb. 15th:
Oroville puts focus on dam spillways – aging and some never tested

Mentions New Don Pedro emergency spillway use in 1997
Content from external source ..During that flood, more than 70,000 cubic feet of water per second poured out of the reservoir, including down the unlined emergency spillway. The torrent scoured out dirt and trees, destroyed a road and dumped debris in the river channel downstream. Afterward, the district had to remove "thousands and thousands and thousands of cubic yard of dirt and debris from the river," said Curtain...

Don Pedro might use its spillway again in a few days (first time since 1997):
https://ww2.kqed.org/science/2017/02/17/another-california-dam-grapples-with-flood-danger/

Operators are releasing as much water as possible to make room for anticipated storm runoff. The lake level is hovering around 826 feet in elevation, close to the 830-foot maximum.

But with almost five inches of rain expected on Monday and Tuesday, officials say if the forecast pans out, they might need to open their "controlled spillway."

The last time the spillway was used in 1997, parts of Modesto were flooded.

Shasta Pretty full too:
http://www.redding.com/story/news/l...-ahead-north-state-braces-more-rain/98049690/

With more rain and runoff on its way, the U.S. Bureau of Reclamation is planning to hold water releases from Shasta Dam to about 64,000 cubic foot per second (cfs) through Saturday.

But It will begin reducing flows on Sunday to around 34,000 cfs due to the expected heavier rainfall Sunday night and into Monday.

After the storms pass, however, the river flows will be gradually increased back up to 79,000 cfs to make more room for future storms and runoff.

"Once they (the storms) clear out we will be increasing flows," Don Bader, the bureau's area manager, said Friday.

Lake Shasta was only six feet from the top of Shasta Dam on Monday when water releases were upped to 79,000 cfs and the lake, which is about 91 percent full, is now about 14.5 feet from the top of the dam, Bader said.

Basically the entire system is saturated, and other dams are being stressed.

 

[sushi]

Still learning workarounds for image uploading.

Thumbnail above, Full image below.

The black square shows the area of water boil. This may be hard to see in the original so I have tried to make it evident.
On the left side are a set of arrows. Each arrow points to a sidewall drain discharge. The two yellow arrows show roughly the same amount of discharge. The red arrow shows much greater discharge. The red arrow discharge plume extends out from the spillway wall by several feet. This suggests this discharge line is under higher pressure than the other two discharge lines. It is believed the drain line feeding the red discharge port extends into the area just above the white boil.

The blue box at the top is an area of spillway wall in which there should be other discharge plumes visible. I can see none even when I inspect the image in Photoshop or when I manipulate image contrast.
The yellow line along the right sidewall is in the area in which a discharge plume should be visible. None are seen. Nor are any seen above this position but this may be due to them being obscured by water spray / mist.

 

[sushi]

Don Pedro might use its spillway again in a few days (first time since 1997):

Basically the entire system is saturated, and other dams are being stressed.

And there is roughly 30" of water bound up in the snowpack. A few days of warm winds, rain, and soil fully saturated with water, and the inflows to all the dams is likely to be high.

 

[rock whisperer]

Freshly excavated hillside in 1968 showing the two main bedrock units, likely sheeted dikes(between blue lines) and the metavolcanics. Note the varying composition within the metavolcanics, light yellow-orange, dark orange, and very light grays, extending deep underground. Apparent dip to the layers is to the west. What are the circles on peoples heads?

 

[KenMH]

Could you explain whether or not it seems most likely to you that the drains stopped functioning at that location instead of the simpler idea that the once erosion due to massive downhill seepage beneath the slab provided an easier route for the water to follow, that's the route that it followed so it simply bypassed those particular drains? Water will take the easiest route, given various choices. I'm willing to believe either reason is possible, but I'm not ready to dismiss the one that seems to be the simplest in principle.

Here it appears that the weathered rock extended deeply enough, and with a downhill slope to the right (remember, the weathered rock is already largely missing at the time this photo was taken, so there's a bit of a 3-D profile of the weathered zone in which new eroded pathways could have been established), that once some erosion or piping had occurred, the exit path could have been downhill to the right, leaving the structure entirely (not the only possible option of course). Water would not follow a gently-sloped drain pipe if it had an easier downhill option such as what seems possible in this case.

I would not argue against that theory, I think it's highly possible, so water bypassed the drains channeling down hill and to the south side of the spillway along the weaker orange rock, instead of broken drains those are unfed drains in the images. But that is essentially the same thing just delineating between the drain 'system' including any grading and grouting of the graded cut to channel seepage to drains, and the pipes that carry the water. Either way, bypassed or broken pipes, the drain system is what we both are talking about.

I tend to want to give credit to the original builders as changes in excavation depth due to bad rock are mentioned in the 1974 CSWP report pdf we have been reviewing (PDF page 185, document page 133), so they knew about some of the bad rock at the time. So I postulate that until proven otherwise the grade of the cut and treatment of the surface were up to the standards of the day and those standards are still valid. I was one of the first to raise concerns about the local rock having been to the site, I strongly suspect the geology and that angled seam of weak material. But I don't want to speculate on where stuff moved underground out of sight so I am going with the drains flow as a primary indicator for now, since it is clearly visible and has a time sequence of images from not only event but from years past. We have learned from various documents that the drains are to control seepage and leaking of the spillway, we can see them in operation, so they are in my opinion a very good place to look for clues to this situations development.

here is the primary 'drains might be broken before the incident' image, from Scott Gates post #14 in the 'spillway drains- how they work' thread.

and his other image from that post (thanks Scott! great images!)

The BOR spillway design document mentions spillway drainage as a cause of damage in several instances (PDF p106, doc p3-89 & 3-90), both of those however were partially attributed to freezing conditions and lack of insulation. Something that all drains do that I know, and I know you know, is just clog up. That is a fact, drains get clogged eventually if they dont have enough prefiltering. So the drains in the image don't need to be broken to not be functioning. But there is evidence that these drains have been an issue in the past.

Mick posted these:
2012

and 2007

(Thanks Mick!)
Both show the same drains may have been clogged or broken for a long time period.

You may very well be right, I find that totally resonable that erosion under the spillway angled in the way you describe and caused the undermining without any additional flow from a broken drain, but there is evidence that there was a broken drain. Possible not as I described in the timeline I conceptualized but far earlier.

Glad I reserved the right to change my mind!