Thread Status:
Not open for further replies.
  1. So were the drains vitrified clay right up to the discharge on the wall? If so, it takes corrosion out of the equation, but it allows flow, both into and out of the drainage system. It is difficult to ensure integrity for obvious reasons (brittle, joints,etc) . If a drain is pressurized, the path of least resistance isn't necessarily back to the deck?
    Last edited: Feb 20, 2017
  2. Pozzolith

    Pozzolith Member

    That would be logical. A clay pipe system would have literally dozens of joints on each branch, and then more joints at each inlet/outlet to the collection drain. When settling occurs, leaks follow.
  3. wrorke

    wrorke New Member

    Because the herringbone is at an angle to the wall, the notch distance on the wall would be larger than the distance between the pipes (perpendicular to the herringbone). I have not done the trig but the pipe spacing could be 20 ft. The side wall distance could be used to estimate the angle of the herringbone.
    • Like Like x 1
  4. sushi

    sushi Member

    This is the image of what appears to be a damaged wood form such as may have been used to protect the clay pipe when the slipway walls were cast.

    • Like Like x 1
  5. AlmostaCE

    AlmostaCE New Member

    From the full resolution image

    It appears to be just stones.
    • Agree Agree x 3
    • Informative Informative x 1
  6. sushi

    sushi Member


    Had this memory of seeing what appeared to be wood form remnants in an image. Could not remember the image or where I had seen it. Have just been reviewing the thread and found your post, likely for the second time. In the interim went and dug through several hundred images looking for what I had thought I had seen, found one, and incorporated it in a new post that essentially duplicates yours.

    I am not sure what is worse. Inadvertently borrowing from another poster, or being debunked by that same poster. :)
    • Funny Funny x 2
  7. deirdre

    deirdre Moderator Staff Member

    that's called Pareidolia.
    Something we have to be careful about when looking at photos.

    but thanks for pointing out where the 'square' was.. it was driving me nuts trying to find it :)
    • Like Like x 1
    • Funny Funny x 1
    • Informative Informative x 1
  8. wrorke

    wrorke New Member

    The herringbone spacing would be different if measured in horizontal plane (plan view) vs in local plane of spillway. That prompted me to try to extract data from the drawings.


    Some of the numbers are hard to read. I typed the readable values into a spreadsheet and plotted them, then adjusted the unreadable values to get a smooth profile


    I fudged the values in gray cells to fit.


    5.5% grade is about 3.15 deg slope. The hypotenuse of 3.15 deg is only a fraction of a percent longer than the horizontal run. In the upper section, the drain pipe spacing would be very close to the same whether measure in plan view or along the slope.

    The bottom slope of about 25% is a 14 deg slope. The hypotenuse of 14 deg is about 3% longer than the horizontal run. 20 ft spacing in plan view would measure about 7 inches wider when measured along the slope. Probably not an important correction to our analysis.

    When I measure the notch spacing in the bottom of the sidewall footing, I get about 28 ft spacing. If the pipes are spaced at 20 ft in plan view, an angle of 42 deg from perpendicular to centerline would produce 28 ft spacing at the footing (accounting for both the herringbone angle and the spillway slope). I would not be surprised if the herringbone legs were each 45 deg from crosswise, with a 90 deg included angle between them.

    Possibly of greater import is the concept that the slope of the herringbone drains ("laterals") will be less than the slope of the spillway. Defining the angle of the lateral in plan view where 0 deg the lateral runs across the spillway and 90 deg the lateral runs directly down the spillway, I calculate slopes vs possible angles for both the upper spillway (~5.5% slope) and lower spillway (~25% slope).


    Could lack of slope (and related lack of flow rate capacity) on the upper slope be a contributor?
    • Informative Informative x 1
  9. wrorke

    wrorke New Member

    It appears to me that the sections in the upper right show the collector drain along the sidewall.
  10. wrorke

    wrorke New Member

    Can anyone help decode the Notes on that drawing?

    1. See A-??? for concrete notes
    See A-??? for general reinforcement notes
    2. See A-??? for excavation details
    3. Sections shown on this sheet are typical showing ???
    ?? of different situations
    4. Formed wall contr??? and ??? ??? ??? not
    shown are to be spaced 50' ??? ??? ???
    5. All ??? joints are to be sealed with specified material
    6. Concrete ??? ??? ??? compressive strength of ?000 psi
    7. Lateral expansion joints ??? ??? ??? ???
  11. EricL

    EricL Member

    Water within the drain pipes won't become "pressurized" in any way that I can imagine, other than that which results from the combination of the force of gravity and resistance to flow (leading to the head pressure which causes the water to spurt from the outlets). The collection pipes simply provide an easy route for gravitational flow to the main pipe just outside the wall, and the flow in that pipe also operates by gravity. If water beneath the slab has a pressure somewhat greater than atmospheric, and that is possible since the head pressure within the stream of water above is the most basic reason the water comes through cracks from above in the first place, that would only encourage under-floor water to find an easy path of escape to an area of lower pressure. That easy path of escape would be into the collection pipes. The fact that water can run out of the collection pipes as easily as in makes no difference as long as the net gravitational flow ends up in the non-perforated pipe just outside the wall. Any water lost from one collection pipe, if that water runs far enough downslope beneath the slab that it can't return, will simply end up encountering another collection pipe during its downhill journey soon enough. If the water runs downhill (and it will), and if other void spaces beneath the floor are not excessive (which is a whole other topic), it will end up being taken out from beneath the floor by the drain system as fast as it enters (assuming the rate of entry is not excessive, which again is a whole other topic).

    If you have an idea how the water within the collection pipes could somehow have a pressure that's greater than that of their surroundings, maybe you could elaborate on that.
    Last edited: Feb 20, 2017
    • Agree Agree x 2
  12. wrorke

    wrorke New Member

    Drain spacing purportedly at about 20 ft. I previously estimated slab length at 52 ft. These spacings seem approximately consistent. Possible variation in spacing between drainlines. Drain channel shape consistent with prior photos. Appears to be simple box form over drainline. There was a prior comment regarding gravel backfill.

    Set side forms, add gravel bed, lay drainpipe, fill to top of form w gravel, pour slab on top?

    Attached Files:

    Last edited: Feb 20, 2017
  13. sushi

    sushi Member

    Great wrorke!!
    You beat me to it as I have been working along similar lines. The image below gives a plan view of the spillway. The red dots represent the position of the discharge ports.

    Question - Do each of the points in the Slope of Lateral Drains graph represent a unique discharge port?

    I am trying to develop a profile such that it becomes possible to identify the potential for the development of a harmonic response within the slab structure. We are not allowed to speculate on this site so I will not mention my hypothesis that such an harmonic may have lead to drainage system failure and / or to slab failure.

    Since the harmonic will vary depending on the volume of flow there exists the possibility of the structure being at increased risk under specific discharge scenarios. This is very common in many engineered structures where a specific harmonic frequency will lead to the destruction of the structure. Marching soldiers always break step when passing over bridges for this reason. Knowledge of harmonic stress is relatively recent and I suspect it was not broadly known, or well understood in the pre 1967 engineering design phase.

  14. sushi

    sushi Member

    I do not understand this.
    I understand covering the drain and the gravel blanket with polyethylene sheet to protect them when the concrete slab is poured.

    But this would result in an impermeable film running the full width of the spillway channel and this film would inhibit any water percolating between the slab joints to reach either the gravel bed, or the drain pipe. Any weeping through the joint seam would run down slope to ??

    This would effectively defeat the entire purpose of the drainage system.

    Polyethylene sheet will degrade over time but the key factor in that degradation is UV exposure and the sheet would be protected from this by the poured slab.
  15. EricL

    EricL Member

    My thoughts on this are far more simple. We already know that there was a sizable pocket of very soft, erosion-prone material immediately beneath the slab at the location of the failure. To my way of thinking, it's a little early to start looking for extremely complex models to explain the failure when such a simple idea that long-term erosion created a void beneath the floor which was too great a dimension for the slab to span under load. Your method of thinking is far beyond my abilities, so I won't try to shoot it down as implausible, but I think we all need to be careful about too-easily considering ideas in a context that suggests the designers probably missed something, when there's little doubt that they knew more about this stuff than any of us.

    I do worry about this new bit of information that the collection drains were bedded in gravel. Gravel is not a filter, and perhaps the designers expected that to be okay if they considered the whole subgrade to be non-erodible bedrock. But when erosion-prone material is present, as seems to have been the case, the decision to use gravel bedding instead of filter sand would literally open the door for moving water to carry material out from beneath the slab.
    Last edited: Feb 20, 2017
    • Like Like x 1
  16. EricL

    EricL Member

    You are quite correct. When I read that, I naturally assumed that the sheet was placed ONLY over the locations of the pipe and associated bedding. Water dropping through the slab onto a sheet that's say, three feet wide, would simply run off that sheet and downhill to the next collection pipe. The same would happen to any water that entered just inches downhill of the pipe if there were no plastic sheeting at all, so the net difference between this and a no-plastic-sheeting situation would be minor.
    Last edited: Feb 20, 2017
  17. No I don't disagree with what you are saying. However, for emphasis, there is pressure on the drain pipes (backpressure / friction loss causing visible jets at the operating drain discharges), and if the drains go thru voids, or get broken creating voids, they can empty into the voids, and potentially increase erosion in the voids, making them bigger. Just sayin...
  18. EricL

    EricL Member

    Very true, and I've posted that very thing before. I was strictly talking about the system "as designed".
    • Like Like x 1
  19. sushi

    sushi Member

    1. See A-3B5-4 for concrete notes
    See A-3B6-1 for general reinforcement notes
    2. See A-3K1-7 for excavation details
    3. Sections shown on this sheet are typical showing HANDLING
    of different situations
    shown are to be spaced 50' APART BETWEEN EXPANSION JOINTS
    5. All INVERT joints are to be sealed with specified material
    6. Concrete DESIGN BASED ON compressive strength of 3000 psi
    7. Lateral expansion joints IS ABBREVIATED AS LAT EXP JT

    I'm making up for the bad pun earlier.

    • Like Like x 1
  20. Pozzolith

    Pozzolith Member


    This photo seems to show keyways at the longitudinal slab joints. Also, there is no evidence of rebar or dowel bars connecting to the adjacent slab.

  21. wrorke

    wrorke New Member

    In general I agree.

    Over the past few days I have been regretting my prior statements about those high outflows must be pressure driven. It is possible (likely?) those high exit velocities are merely due to high velocities -- free stream, not full pipe section. (Also my earlier comments re pressurization were prior to me seeing description of the drain system - herringbone w perforated pipe, etc.) They were based merely on seeing some plumes that appeared similar to fire hoses without nozzles.

    That said, if there were no flow out from under the slab, the pressure at the bottom of the approx. 550 ft spillway drop would be roughly 250-300 psi (assuming 1/2 psi/foot). Obviously there is leakage between the slab and rocks and through the drain system which greatly reduces that pressure. The pressure builds up to where the driving pressure is sufficient to vent the incoming water. It is a very complex and distributed network. It depends on the flowrates and the drain rates (both natural and manmade) everywhere under the slab. We know the pressure is zero at the sidewall openings and zero at the top of the sidewall backfill. Everywhere else it depends on the pressure required to drive the flow through the system. The backfill looks like that should be a very low pressure, so the pressure at the sidewall footing should be close to zero.

    Upstream of that sidewall (ie as you move under the slab toward spillway centerline), the natural and man-made flow channels to the sidewalls may have low capacity relative to the inflows.

    The system is unstable because erosion creates preferential flow channels which results in increased local flow and local erosion. Lather, rinse, repeat.

    If all the perforations were taking away all their share of the local water, underslab erosion would be minimal, but if preferential flows developed, look out. Eventually any flow at all will remove the fines and reduce the resistance to flow, developing preferential flow channels. French drains typically are wrapped in Geotechnical fabric which acts as a filter to keep the fines from being carried away.

    Question on different issue: Would upflow through the slab seams wash away the interslab seals and thereby open the seams to later leakage of spillway water to beneath the slab? (Why/where would there be underslab pressure in one scenario and later top side pressure at same location in different scenario?)
    • Like Like x 1
  22. AlmostaCE

    AlmostaCE New Member

    • Like Like x 1
    • Informative Informative x 1
  23. CRM114

    CRM114 Member


    There seem to be cracks associated with these features, some of them look to be patched, so they probably originated prior to damage.
  24. CRM114

    CRM114 Member

    An interesting article:

    • Informative Informative x 2
  25. wrorke

    wrorke New Member

    The dots on the curve are the Stations and Elevations from the spillway profile. They are not related to discharge port locations.
  26. Scott Gates

    Scott Gates Active Member

    What is the source for this graphic?
  27. Scott Gates

    Scott Gates Active Member

    I agree as noted in a comment above ... the poly would likely only cover the width of the pipe and adjacent gravel blanket - would seem water would simply drain down slope, past the poly and to the next drain
    Last edited: Feb 20, 2017
  28. Scott Gates

    Scott Gates Active Member

    There would be seemingly little point to a drain located IN the slab, or even inset partway in the slab. It weakens the slab (less thickness) and doesn't allow drainage ... perforated pipe is perforated on top side with solid bottom to allow flow in the pipe.

    And the point is to draw moisture away from the slab ... which would mean they should be in a gravel trench below the slab.
    Last edited: Feb 20, 2017
  29. Scott Gates

    Scott Gates Active Member

    This was addressed (page 93-94 of book or 145-146 of PDF):
    • Informative Informative x 1
  30. CRM114

    CRM114 Member

    Yet there they are in the photo. There is always a gap between design and constructed versions. Engineers draw neat lines and contractors try to cut rock to a neat line, and an inspector splits the difference. A trench for a 6 inch pipe is mighty hard to cut in some rock without making it twice as big as designed.

    There could be multiple explanations - an example would be the contractor was allowed to over thicken the slab on either side in trade for not so deep trenches, and the engineer approved because the trenches would end up more gravel than bedrock otherwise.
    Last edited: Feb 20, 2017
  31. Scott Gates

    Scott Gates Active Member

    Except that a perforated drain pipe, with the even the top half embedded in concrete, would not let water get into it, let alone drain ...

    And yes the trench would need to be bigger than designed to accommodate gravel around the perforated pipe.

    Even if you mounded gravel over the pipe to maintain a permeable space to the pipe perforations, the water would have to build enough pressure to get to the top part of the pipe where the perforations are
    • Like Like x 1
  32. sushi

    sushi Member

    I love counter intuitive evidence.

    DRAIN IN SILL v2 21-02-2017 3-38-46 AM.

    There is another deviation from the enginnering specs found in

    FCS PLAN zh9californiastatew2003calirich_0152a.jpg

    The above FCS PLAN suggests that from centreline to the sidewall there should be 4 slabs each 20 feet wide then a fill in slab of 9.33 feet (9.4 inches) for a width from centreline to sidewall of 89 feet 4 inches. Multiply by 2 and we have the full internal width of the spillway of 178 feet 8 inches. This dimension matches the reported width contained in the Hydraulic modelling document which had to be exact as they sought to build an accurate replica of the structure for flow tests.

    Looking at the current spillway images it appears that there are only 4 main slabs plus the two sidewall slabs rather than an 8 plus 2 configuration.

    If they did cast sills or beams integral to the cast slab then this stiffening may help explain the cantilever of the otherwise unsupported slab seen in the straight on images of the erosion void.

    I suspect they first cast the slab over a section of the drainage field and left the 9.33 foot slab at the sidewall margin for later. This would give them some degree of "wiggle room" to align the drain lines with the collector running outside the sidewall. I suspect they then erected the form work for the sidewall and cast that as a unit at the same time as they cast the 9.33 foot sidewall margin.

    In post

    wrorke identified a significant variation in the width of the cast sidewall footer.

    Trying to puzzle out this structure is almost as bad as trying to figure out someone else's sphagetti code.
    • Like Like x 3
    • Agree Agree x 1
  33. CRM114

    CRM114 Member

    I think they still would have had gravel over the pipe allowing water to reach all perforations. The trench would be designed for the gravel and the pipe. I have designed these before and specify a minimum clearance of gravel around the pipe. The contractor is allowed to overexcavate rock up to a point if that is what it takes to get that clearance.

    These are tough to design. The geotechnical engineer wants lots of gravel to get the water to the pipe and would put gravel under the whole slab if possible, the hydraulics and/or structural engineer wants concrete connected to rock as much as possible - somewhere in the middle they compromise. The project manager wants to cut costs, and lots of gravel / rock excavation is expensive. They all compromise and then it gets built differently, when they uncover actual conditions, and hopefully these are documented in a back and forth between engineer and contractor.

    If these features are what they seem to be - that is what they ended up with.
    • Informative Informative x 1
  34. Pozzolith

    Pozzolith Member


    This is from a do-it-yourself website, but I would think the principle would be the same. Basically, a network of 'French Drains' beneath the slab that drain into collection pipes on either side of the spillway.
  35. CRM114

    CRM114 Member

    • Like Like x 2
  36. CRM114

    CRM114 Member

    A French Drain is the general idea, yes. Dimensions on the gravel for the Oroville spillway drains would be bigger, I would guess.
  37. aczlan

    aczlan Member

  38. CRM114

    CRM114 Member

  39. sushi

    sushi Member

    This image suggests significant variation in the depth of the concrete pour. If correct, this has implications for the structural rigidity and load bearing capacity of the spillway slabs.

    UNDERCUT SLAB v02  21-02-2017 1-28-30 PM.
  40. aczlan

    aczlan Member

    If the slab is thick enough at the thinnest part, the variations in thickness should be immaterial (unless the shape of the under slab voids caused stress risers).
    The thin slab seemed to have held up relatively well on the far side until the middle fell out, so theoretically the thin sections of the slab were thick enough for "normal use" (when properly supported from underneath ;))

    Aaron Z
    Last edited: Feb 21, 2017
Thread Status:
Not open for further replies.