Sheet Pile Cofferdam Installed in Embankment Dam 2

[Baldie]

I am concerned about removing sheet piling that will be installed in an existing embankment dam to function as a cofferdam while the intake is replaced. The main concern is that the soils will be disturbed when the sheet piles would be removed after construction and that seepage through the embankment could be increased due to the loosened soils, possibly adversely affecting the stability of the dam. (Stability analyses have not been done yet, but are likely to yield low safety factors for a drawdown condition of the downstream reservoir--the downstream face is currently partially submerged from an adacent reservoir.) It also might just be bad practice to drive sheet piles into an embankment dam and then remove them. On the other hand, I don't want to be overly conservative resulting in a costly design.

Generally, the foundation soils are fairly clean sands, typically with less than 10% passing #200. Embankment fill soils are also granular, but range from clean to clayey sands. Interlensed claystone and sandstone bedrock is about 40 feet deep. The dam is considered low hazard. Seismicity in the area is very low. A seep exists at one groin of the dam and is believed to be due to a cobble layer in the foundation located just below the embankment fill.

I have been unable to find any literature on the amount of disturbance to expect if sheet piles are removed. Of course, we could recommend that the piles be left in place, but they would need to be cutoff near the bottom of the reservoir which could be costly to do underwater. Another possibility would be to grout as the piles are removed. Yet another possibility is to make the contractor responsible for restoring the integrity of any loosened embankment soils. (This makes sense because the type of cofferdam and installation procedures will affect the amount of disturbance, but it goes back to the same question of how much disturbance would there be.) I would appreciate anybody else's similar experiences or suggestions. Following are some more details of the project:

As part of a repair to an outlet pipe through an existing embankment dam, a contractor will be required to design a cofferdam to replace the inlet so that work can proceed in the dry. It is not possible to completely drain the reservoir because of operations.

My task as the geotechnical design engineer is to comment on the feasibility of different cofferdam designs. Cellular, sheet pile, or double sheet pile walls all appear to be feasible. The intake is a morning glory type (vertical CMP) located near the upstream toe that will be replaced. A contractor will probably elect to use some type of sheet pile system because the construction limits are relatively small. We envision that a u-shaped configuration will be used with the open end in the upstream embankment slope and the end of the piling near the upstream crest shoulder. The parallel walls would be on the order of 20 to 25 feet wide.

 

[VAD]

Hello Baldie:

With the ground conditions being sandy and sturated the removal of sheetpiles should ot pose aproblem as the material should flow as the piles are retracted, hence I would not be overly worried about the seepage issue. In any case sheetpiles also provide a seepage path and leaving them inplace would be no advantage. In respect to increased seepage and possibly decrease in overall stability,I do not see this problem materializing.

That's all for now. Interesting practical problem.

 

[Baldie]

Thank you VAD.

I had not considered the increase in seepage that could be caused by the presence of the piles themselves.

 

[layfieldman]

This is within the reservoir? With standing water? Or is it under ground level within the embankment itself? Would you be interested in a geosynthetic dam? They have been built up to 16' and are very easy to install in standing water then dewatering the contained area. Basically a big geotextile tube filled with water placed in a U configuration.

http://www.aquadam.com

 

[kmerl]

Layfieldman has a point with the water filled tubes. I have specified and used theses before and they work well. I've also used a product called Porta-Dam, these are metal frames that support a water retaining membrane, I'm not sure if there is any product local that is similar. (

http://www.portadam.com)

I would be very interested in this case to see if the embankment is currently stable with seepage already occuring.

I would be very careful with disturbing the embankment as a saturated unit of clayey sands and sands. Liquifaction may be issue with gap graded sands/clays. Has any testing been completed on the soils? Some materials mixtures in that condition can liquify when shear stresses are increased beyond the internal shear resitance.

I would prefer to lower the water level in this condition and dewater the embankment to make it more stable. The embankment (based on the soil types) is not retaining water very well at this point and any distrubance with the full load of water on the embankment may make it unstable, especially a breach, even if limited. Toe seepage on this type of embankment is a serious issue and should be considered carefully. Full embankment remediation may be an issue.

I will be intesrested in hearing and understanding results from a embankment stablility anaylsis.

Keithe J. Merl

http://www.princetonhydro.com

 

[Focht3]

I agree with [blue]kmerl[/blue]. I don't like the idea of driving sheet piling - temporary or permanent - into the heel of this embankment. At least not without a helluva lot more engineering analyses than has been indicated so far in this thread...

Has a study been done on the stability of the embankment? How about seepage analyses of the entire embankment - before and after sheetpiling is installed?

What are the consequences should the dam fail? How tall is the dam?



Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[jheidt2543]

Baldie,

I have to take issue with your statement:

"Yet another possibility is to make the contractor responsible for restoring the integrity of any loosened embankment soils."

It seems to me, that as the geotech. designer for the project, it should be your responsibility to come up with at least ONE solution. If the contractor wants to offer an alternate solution and take on the risk for the design of that solution, fine - that's his decision. I just think it is wrong to foist that onto the contractor because the design team can't or will not solve the problem. That's my two cents.

I am wondering about VAD's comment regarding the "self sealing" of the dam when pulling the sheets, would using a vibratory head pile driver help recompact the gravel fill or not?

On related topic, there is a thread regarding the failure of two dams in Marquette, MI. The details have been outlined in an article in Engineering News Record in the last couple of weeks. Basically, the Army Corps of Engineers decided to install a "safety plug" in an existing 100 year old dam. After the installation, a rain/flood event, well under the design criteria, took out the plug and the dams downstream.

 

[DRC1]

As you dive and extract the sheets, the loose to med. sand and gravels will densify, removing any void left by the sheeting.(If the sands are rather dense, it is possible driving the sheets may loosen them) However, the densifcation will result in settlement of the sheeting line, which will also resultin settlement of the part of the embakment over the sheeting line, which may disturb any filters or liners present over the sheeting line. Also the energy and vibration can induce sloughing of the embankment. Vibrationless methods are available for sheeting instalation. A final concern is the cofferdam for the pipe repair. Yuo mention a three sided dam in a sandy soil with excavation well below water table. Sheeting can not be used below the pipe, yet even if the sheeting can span strucurally across the pipe, flow must be cut off prior to dewatering or the bottom could blow. This should be investigated carefully before proceeding. Simply making it the responsibility of the contractor may leave you open for a significant claim.

 

[cvg]

jheidt
the dike in michigan was constructed with a "fuse plug". This is a form of emergency spillway which is a granular embankment, designed to erode away in a controlled manner during overtopping in order to provide additional spillway capacity. Unfortunately, it was either poorly designed or constructed, and failed. This issue is essentially un-related to the issue presented in this thread - however - the result was the same - a catastropic dam failure. This is precisely the reason that any modification done to this dam should be thoroughly investigated for "fatal flaws" to determine what the failure modes are and the risk associated with it.

 

[Grouser]

How much of the intake must be replaced? Maybe the replacement can be floated into place and sunk to its final position without using a cofferdam, a la oilrigs.

 

[Baldie]

Thank you all for your posts. They really helped me put together a thorough report.

 

[Focht3]

Has the work been completed yet?



Please see FAQ731-376 by [blue]VPL[/blue] for tips on how to make the best use of Eng-Tips Fora.

 

[Baldie]

This study was done in several design stages, and the final design report is due next week. My original post was submitted very early in the process, at the 30% design level. Keep in mind that this is a low hazard dam with no or very little downstream development. Following is an update of the design recommendations for those who may be interested.

1) Cofferdams:

The design team (at my suggestion) determined that specific cofferdam designs should be performed by the contractor because the effectiveness of a design depends greatly on construction methods, quality and experience of personnel, and equipment available at the site. A contractor will also be able to optimize a design based on their resources (e.g. pumping capability, equipment, past experience with various methods used on other jobs, etc.).

The maximum water height will be about 12 feet above the mud line depending on the design storm event that the cofferdam needs to accommodate. The following is an excerpt from the cofferdam section of our 90% report (Note the above discussion is also in the report):

"Soils at the site are granular so relatively high seepage beneath a cofferdam system should be expected. In addition a cobble zone was encountered in one boring. Bedrock was encountered at depths 30 to 40 feet below the crest. The existing soils are loose to medium dense and are prone to settlement or liquefaction under vibratory loading.

Sheet piling (single, double, or cellular), earthen embankment, or proprietary systems (e.g. Aquadam, Portadam) may all be feasible for the proposed work. Other cofferdam systems, or combinations of systems, may also be feasible. There are several design considerations that must be taken into account by the contractor. If sheet piles are selected, then driving of the piles should be performed in a manner to prevent liquefaction, and associated slope instability, from occurring. Another concern with a sheet pile system is the potential for a loosened zone of material to be created as the piles are removed from the embankment and foundation soils. A relatively loose zone could increase seepage through the embankment. Grout could be injected at the bottom of the sheet piles as they are removed to fill any voids and densify any loosened soils. The degree of any loosened zone will depend on the contractor’s construction procedures, and if only minimal disturbance occurs stability of the structure would likely not be threatened. It should be noted that the existing dam probably already leaks at a relatively high rate. To alleviate concerns associated with driving piles, it may be desirable to use another type of cofferdam in the embankment area (such as Aquadam or Portadam).

Passive resistance of sheet piles will depend on the amount of upward seepage, which in turn depends on the amount of sheet pile embedment. Adequate factor of safety against piping/heave due to seepage must be included in the contractor design. Well point dewatering or a soil berm may be necessary."

2) Seepage at downstream toe:

A sealed piezometer reading measuring the head within the cobble zone and a separate standpipe measuring the phreatic surface in the same location were installed in the d/s embankment face to get a better handle on the seepage issue. The water levels were nearly identical. We recommended that a drainage blanket consisting of ASTM C33 concrete sand be placed in the area of the relatively small seep. A perforated pipe will be installed within a swale to collect and discharge seepage. The intent of the sand blanket is to prevent any future fines migration, which in my opinion would be very low likelihood.

3) Stability analysis:

Triaxial shear testing indicated effective friction angle of 32 degrees. Zero cohesion was used in the analysis. US and DS slopes are both 3:1. The dam (and foundation) is considered stable w/ static FS=1.5 and pseudo-static FS calculated to be 1.2. The drawdown condition had FS=1.0 so we recommended lowering the reservoir no faster than 1 foot per day. This just seemed like an ok rate, I was unable to find a reference to base this on. Does anybody have an opinion on this 1 ft/day lowering rate. We ran a perm test on one of the triaxial samples; results indicated a coefficient of permeability (vertical) on the order of 2.0x10^-4 cm/sec.

I hope you guys have found this to be an interesting topic.