Subsurface Drainage in Low-permeability Material 1

[mraye77]

I am looking for information re: subsurface drainage of low-permeability materials. An overthrust has developed on a cut slope (7H:1V) of glacial overburden over clay/shale. I am investigating subsurface drainage for slope stabilization. Can anyone suggest any good texts/articles/professionals with information specific to drainage of low-permeability materials or other potential rehab methods?

 

[Focht3]

Could you walk us through your problem? I don't understand "what" moved "where." Is the structure just a slope, or a dam/levee embankment? Why did it fail? Any relevant factors that might influence our comments?

If the clay shale moved and there are no natural drainage layers within that material, drainage may not be a viable option because of the low permeability of these materials. You may have to construct a substantial toe berm to improve stability. (Waco Dam (Texas) experienced a horizontal sliding failure on a thin clay shale seam. They fixed it with a toe berm.)

Anyway, that is all speculation at this point: we need details.

 

[mraye77]

Ok. More info. needed, not a problem. The slope is a dam embankment slope constructed circa 1960. A minor overthrust was detected in the slope in 1994 and associated upland cracking in 1995. The cracking was sealed and clay capped and has not reappeared. The current deformations appear to correspond to the northern limits of a massive overthrust that developed in 1962 during dam construction. Excavation of the slope (in 1962) is believed to have reactivated an ancient shear zone seated at the base of a softened shale. (stratigraphy in the slope is glacial till overlying clay/shale, which can be further broken down into a softened shale overlying a harder shale) A berm was constructed over the area at the time, but the slope continues to creep (as evidenced by the reoccurrence of the overthrust).

A recent drilling program in the movement area revealed no continuous granular seams, although several holes showed thin (1 to 2 inch) granular seams within 15 ft of the slope surface. Slope movements do not endanger the operation or safety of the dam but present more of a nuisance than an immanent threat. Current thinking is to drain the overburden to reduce the driving force on the slope, possibly with parallel trench drains or vertical sand drains in combination with manufactured drains. Are we looking in the right direction? Does anyone have any experience with this type of situation?

 

[Focht3]

Hmmm, still not enough data. For me, anyway.

What kind of instrumentation is installed? (I assume that inclinometers were / are used to identify and monitor the shear zone.) Any piezometers in the suspect zone? What types? (I hope that more than one type was used.) What does the phreatic surface look like in the shearing zone? If you have any standpipe piezometers, how big (ID) are they? For the standpipe piezometers: Have you performed slug tests? Falling head / rising head permeability tests? These are a few ways to get a feel for whether drainage will work. (Hopefully others will chime in with additional comments here.)

Discontinuous granular seams...my gut says that using drainage alone to stabilize the dam is wishful thinking.

It's been moving since 1962? Wow. What's downstream?

 

[mraye77]

...as to instrumentation, there are several SIs in the movement area - movements are around 5 to 12 mm a year (although more significant movements occur during years with high snowmelt/runoff). Piezometers were installed in November 2001: 13 standpipes (ID = 6" if I remember correctly) and 3 pneumatics. Standpipes were installed into granular seams encountered in the overburden - i.e. shallow installations. No in-situ permeability has been done so far - only 3 of the standpipes maintained a readable water level throughout the past year, most were installed dry, or dried up within the first month. The pneumatics were installed adjacent to the SIs showing the most significant movements and tips were placed into the assumed shear zone (depths of 9.4 m, 10.4m and 16.8 m). Piezometric head in the shear zone appears to be between 1 and 3 m, and increased winter slope movements are accompanied by slight increases in piezometric pressure; this is all preliminary because the piezos have only been in place for one year - additional readings are definitely in order. We are also in the midst of a drought here and we have had three extremely dry years in a row.

The movement is actually occurring on an abutment slope. The movements are causing compression of and damage to a concrete drainage flume on the slope surface. At the toe of the slope is the dam spillway (slope movement has not yet impacted on the spillway structure). As I said before, movement is more of a nuisance than an immediate threat as repeated repair of the flume structure has been necessary. I understand that drainage will not completely stabilize the slope, but we are trying to manage slope movement to eliminate the surges that accompany wet years. We don't know what will happen if we get a significant snowmelt after a prolonged dry period, or if we get several extremely wet years in a row, so we want to have something in place to minimize the stress placed on the slope.

Well, I hope I have covered everything this time!

 

[BigH]

This would be a nice problem to really be able to sit down and kick around. I will comment as follows:

1. If I remember Chuck Brawner (Prof at University of British Columbia) correctly, drainage usually has the greatest potential for increasing the stability of a slope.
2. In low permeable soils, gravity might not be the most appropriate way in many situations where short term results are required. I have used vacuum horizontal drains previously (see some earlier threads on horizontal drainage). Again, there is a system that was patented by Pakinis and Brawner (if I have Pakinis') name correctly. It has been used in California and elsewhere (circa., 1984).

Hope this helps -

 

[jdmm]

I don't deal with dams but with landslides it is can very often be details that cause the problem. Specifically, sand or gravel layers transmitting high pore pressures. I would suggest using CPT with pore pressure measurements to detect sand layers. Once the sand layers are identified then stop and carry out a dissipation test. This can find layers as thin as a couple of inches and the dissipation tests will help to evaluate the groundwater regime. BigH is right that pore pressures are usually the cause of the problem. Rarely are we too far wrong on the strength parameters and there affect on design is not as great as pore pressures. I just finished an evaluation of a failure on a 4H 1V slope in firm clay. Several zones of sand carried artesion pressures up to 20 feet above the ground surface. We used the CPT to find them.

 

[Focht3]

Okay, still missing some details - but I'll wade in with the information at hand. Static CPT is a great tool for identifying granular seams within moderate strength clays, BUT...

I've used CPT before (I worked for Fugro many moons ago) and I suspect that it won't work. Re-read the original question and note the "clay/shale" and "glacial till." I'm afraid that the cone will max out before reaching through the entire embankment, and will certainly max out in the first shale or glacial till strata it encounters. The embankment itself should be fairly uniform in material make-up, so CPT is unlikely to be of much help.

Your standpipe piezometers are entirely too large in diameter. They need to have an ID of 25 mm (and preferably less.) The pipe ID directly affects the piezometer's response. Put in new ones; keep the old ones for reference.

With dams, minor nuisances have a nasty habit of becoming major problems. Note that a movement of "only" 5 mm (about 0.2 inch) per year represents a total movement of 210 mm (about 8.3 inches) over 42 years, and a movement of "only" 12 mm (about 0.5 inch) per year represents a total movement of 504 mm (about 21.3 inches) in the life of the structure! The presence of the flume may have helped avoid a catastrophic failure - regular inspections picked up the concrete damage. Check the maintenance records to see how long the dam operator has been making repairs -

The previous comments regarding controlling the piezometric forces are absolutely correct; if possible, reduce the piezometric pressures first. Then you can look at other options if that doesn't work as desired.

Another word of caution. A total movement of 200 to 500 mm in a clay/shale can (will?) result in a degradation of shear strength along failure surface(s). A fairly rapid failure during a peak load event can occur under these circumstances...take advantage of the drought conditions to install instrumentation where it will be difficult to accomplish in wet periods.

Prepare an action plan to install mechanical dewatering to reduce the piezometric head - in case the weather (and dam) turn against you before you are entirely sure what is happening. You may have to install piezometers and dewatering wells "on the fly" in order to evaluate and reduce the excess piezometric head if the movement begins at an unexpected rate. Be sure to log the piezometer holes and wells - have an engineer or engineering geologist do this, not a technician.

Emergency repairs will be very stressful for both the field and office staff, but it will also be very rewarding professionally and personally. If emergency repairs become necessary, take advantage and use the time to build staff relationships and train your younger engineers. (I have some "fond" memories during repairs to Morris Sheppard Dam [Lake Possum Kingdom, Texas] - an Amberson-type dam. Imagine 300 mm holes drilled into a fractured clay shale beneath that dam that erupted into a lake-fed fountain over 3 meters high! This always seemed to happen late on Fridays...)

Let us know what you decide to do. Good luck!