Piping/internal erosion 1

draw a flow net.
determine the exit gradient (let's call that ie).
determine the critical gradient (let's call that ic).
f=ie/ic

The are many references on this topic. I'd suggest you do some research and let us know what part of your studies have you confused.

f-d

¡papá gordo ain’t no madre flaca!

To fattdad's post, let me add:

Make damned sure that you are applying the concept of critical gradient properly, as it really only applies to boiling and heave, not to initiation of internal erosion on non-vertical paths.

Beyond that, you must use filtering criteria such as have been developed by Terzaghi, Sherard, and plenty of others. Sherard's criteria from the late 1980s (ASCE JGE ~1988 or 1989) are probably the most widely used and accepted.

to the OP. Others bring up critical design details pertaining to penetrations or conduits through an embankment. Most embankment failures occur along such penetrations. The other advice is correct and very important. Your original question pertained to the calculation of piping through an uninterupted soil zone. The steps I outlined represent the solution. Even with extreme care on this single issue, however, your dam can still fail from piping. Please study safeguards to prevent piping along interfaces and penetrations.

f-d

¡papá gordo ain’t no madre flaca!

And, make sure you don't have dispersive clays (pinhole test). All it takes is a pinhole and . . . see the Teton Dam failure. You can link from here:

I have been watching this thread as I have a lot of interest in the subject. Piping, oftentimes vertical, is a significant problem in alluvial/colluvial soils in Western Colorado. Unfortunately, it is not well recognized as problems with compressible and collapsible alluvial/colluvial soils, possibly overlying expansive formations tends to keep the mind focused.

The issue of potential dispersive properties complicates the issue. As BigH notes, the examples can be spectacular. A note about the Pinhole test, ASTM D 4647:
It is a good test but, a review of the Standard, notably Sections 5 (Significance and Use) & 6 (Limitations) provides a reality check.

Dispersive Soils can be identified but no computation methods are produced. To my knowledge, you will continue to be on the 'Frontiers of Science" when dealing with soil piping. The above posting by fattdad to 'draw a flow net' is appropriate.

The Pinhole test, along with the Double Hydrometer (ASTM D 4221) and the Crumb (ASTM D 6572) methods can usually (note usually, not always) identify a Dispersive Soil. The problem comes when you have a soil with Dispersive-Like characteristics.

Regarding soil with Dispersive-Like characteristics, which has apparently been the condition I usually encounter, I have an old reference from my Father's Library.
ASTM STP 623, 1977, Dispersive Clays, Related Piping and Erosion in Geotechnical Projects, Sherard/Decker, editors.

The chapter by Daniel Resendiz, Relevance of Atterberg Limits in Evaluating Piping and Breaching Potential, has been helpful in my circumstances. The discussion section describes observations by Resendiz relating observed piping in small dams to values of clay activity (Skemptons Activity, [plasticity index / percent clay, 0.02mm]). Resendiz observed that "...all the dams whose failure was attributed to one of these mechanisms {piping and breaching} have clay activity between 0.3 and 1.1".

As my Father insisted on the Hydrometer analysis being required, I have 35 years of data in Western Colorado, Southwest Wyoning & Eastern Utah. In all cases of soil piping in soils with low water flows (at least initially) I find the Resendiz criteria of clay activity between 0.3 and 1.1 to be applicable for identifications of soil with Dispersive-Like characteristics.

Does anyone have any other published references???????

Robin Fell and his students at Univ. of New South Wales have been investigating erosion of dispersive material along with almost every other part of internal erosion of dams. He is now "emeritus" but is still involved in the work. UNSW's website can probably lead you to contact info. One of his doctoral students, Mark Foster, who may be with URS in Sydney or somewhere now, compiled a tremendous amount of info on seepage performance and case histories world wide, and therefore probably knows more than anyone else in the world about actual seepage performance.

I don't believe the Teton core was actually dispersive, and I couldn't find mention of that in the report of the independent review panel, which included R. Peck, A. Casagrande, H. Seed, and a few others. It didn't really need to be dispersive. It was aeolian silt, fine grained, mostly nonplastic, therefore very easily eroded and in need of a filter with D15 < 0.7 mm by Sherard's criteria. The volcanic bedrock at the site has wide open fissures, shrinkage cracks, etc., and the fill in the cutoff trench was probably exposed to high head, high velocity flow in rock that may have scoured the core material, high gradients, etc.

dgillette, so rather than "piping", Teton Dam could be better classified as a failure caused by "internal erosion" with removal of the eroded material through the abutment and foundation contacts?

For me to stay interested in this thread, I guess we need to agree whether: 1) there is some distinction between "piping" and "internal erosion" and 2) what that distinction is.

Mike Duncan (my professor) lectured on the Teton Dam failure and if I recall correctly, there were big seepage forces in the rock fractures of the abutments. This resulted in a very high gradient from the rock fracture through the embankment soil and toward the open air. I'd think this defines "piping" as the exit gradient exceeded the critical gradient. Now, how does "internal erosion" differ from "piping" in this scenario?

f-d

¡papá gordo ain’t no madre flaca!

Semantically, I consider piping to be a subset of internal erosion, although the two are sometimes used interchangeably.

Any of several things could have occurred at Teton. It's clear that there was erodible fill adjacent to very large discontinuities in the rock. What isn't known for sure is whether the internal erosion initiated as "classic piping" where there was v. high exit gradient out of core material in the cutoff into rock, or due to erosion by flow through a crack in the core (maybe associated with arching in the cutoff trench), or whether relatively high-velocity flow in the rock fractures adjacent to the fill started the erosion as scour. The evidence of exactly how it started was all washed away. JMD is correct - because the rock was so pervious, the gradient over the top of the grout cap was huge. You couldn't make things much worse if you tried. If ever in eastern Idaho (between Idaho Falls and Yellowstone), stop and look at the site, just a few miles off US 20 near Rexburg.

What do you mean by "critical gradient"? The only definition I know relates to upward flow causing blowout or boiling at the downstream toe of a dam or levee, not to anything like the more-or-less flow through Teton's cutoff trench.

According to Graham's ASDSO presentation in 2005, failure cause was "Piping of dam core in foundation key trench". Attached photo from his presentation doesn't really square with that and shows the failure near the abutment and 100 feet or more above the base of the dam. I would have expected the higher exit gradient causing piping in the key trench to be farther down near the toe. To me it looks very likely that there was seepage through the abutment.

cvg: The cutoff trench was cut deep into the bedrock all the way up the right abutment, where some of the worst rock was. I think there is general agreement that it occurred in the right abutment, where the cutoff trench was deep and narrow with near-vertical sides. I don't have any digital pictures of the trench handy, or I would post them for you. If you have access to Engineering Geology, see 1987 vol. 24 pp. 239-256, "Lessons from the Teton Dam Failure" by J.L. Sherard. At the maximum section, the embankment was on alluvium, and the cutoff went through it and a little into rock, as I recall.

Graham is Wayne Graham?

DRG

yes - Wayne Graham. I believe I have some photos, I will try and post them tomorrow

see attached file with photos from a presentation given in 2007 by Chris Veesaert (USBR) of numerous dam failures. Page 4 shows the Teton Dam key trench and what appears to be abutment contact area. If these holes in the rock were there when the core was placed against it, than internal erosion which eventually made its way through to the downstream side of the dam (perhaps by piping) is very likely the cause of failure! My notes also indicate that it was grouted, but there is no grout evidenced in the photos. It is not clear when the photos were taken and unfortunately, I did not take good enough notes...

It really was that bad, cvg. There was a grout cap that ran along the center of the key trench, but the downstream side of the trench did have big open joints that were never grouted. They cut the key trench into bedrock because the upper bedrock was too pervious to grout effectively. As it was, something like half a million cubic feet of grout went into a 3-row grout curtain, parts of which did not achieve closure and apparently weren't ever expected to.

The middle right picture on p. 12 (big fractures in rock) was probably taken after the failure. The two bottom pictures were, I think, taken during construction. The bottom left looks like a guy with a blowpipe cleaning the bedrock surface.

Remind me next week when I have more time, and I will scan some pictures from the review panel's report. I won't scan the whole report for you, as it is 1.3" thick, mostly double-sided.

Thanks for the snaps and notes! - my way for snaps

Here is some more detail - finally remembered. I think it came from a retired geologist who was involved in the forensic investigation.

I was wrong about the guy with a blowpipe. It was actually a compressed-air pogo stick compactor.

Regards,
DRG

BTW, the fissure alluded to in slide 7 is in the wall on the left. I'm fairly sure the big trench running down the invert of the trench is for the grout cap.

I'm glad you clarified that, I was looking at the trench. Slides 9 & 10 show the rock fissure close up and to me it seems like a very likely failure point.