Seepage in Dams 1

[pelelo]

I am assigned to design a small dam: 10 mts high and about 70 mts long.

Other than slope stability analysis, I understand I need to run a seepage analysis.

I have been wondering when seepage values are too small or too high for this small dam?

Are there any boundaries?

 

[hokie66]

The book "Design of Small Dams", but the US Bureau of Reclamation, is a good source. I think it can be downloaded without cost.

 

[aeoliantexan]

The U.S. Army Corps of Engineers and NRCS also have dam design guidelines on line.

A seepage analysis can determine the hydrostatic uplift pressure under the downstream slope for use in your stability analysis and the seepage gradient at the toe to be sure the ground won't be heaved and/or boiling occur. The seepage quantity can be estimated to evaluate water loss.

This is a good learning opportunity for you, but you need an experienced dam engineer to check your work. There are a lot of pitfalls, and a failing dam is dangerous.

 

[geotechnicalenginee]

I downloaded that book “Design of Small Dams” for free and it’s really helpful.

 

[dgillette]

@ Aeoliantexan - Thank you for your last paragraph.

I've been meaning to ask, are you from the town of Eolian TX? I never knew there was such a place until I saw it on the map while looking for Breckenridge. Does the wind blow a lot there?

 

[fattdad]

There are two reasons to calculate seepage; one is to see if you will leak water to a point that the reservoir quantity is affected, the other is to gauge whether the gradient of the exit flow is "critical."

For an embankment dam (i.e., with an unbound phreatic surface), the construction of a flow net is a touch difficult. I have successfully used numerical modeling (i.e., Modflow 3-d) to solve the quantity problem. You can consider the confocal parabala solutions for a back-of-the-envelope approach. The confocal parabola is really an approximation and best applied on a homogeneous dam design.

Solving flow problems to address the critical gradient is usually the bigger engineering concern. I usually would have internal drainage design to address critical gradient issues.

f-d

¡papá gordo ain’t no madre flaca!

 

[dgillette]

Given my "druthers," I'd put in a filtered chimney drain, and not worry much about the seepage calcs.

 

[cvg]

central filter is a common practice, especially for homogenous earthfill dams. I can't count the number of NRCS FRS's that were built without them and now are being retrofitted. may also want finger or blanket drain to the downstream toe. An alternative is a toe drain system which can be a maintenance nightmare.

 

[aeoliantexan]

dgillette, I spent the first half of my career in the midwest, often dealing with aeolian soil (loess). I blew into Texas about 18 years ago.

I didn't know Eolian, TX was there. I looked on Google Earth, and I don't think you could call it a town nowadays.

I would guess that the wind blows a lot there. Take Google Earth to the Sweetwater area and see all the wind turbines just east of town.

I agree that a chimney drain is a good feature, although we seldom see one in a 35-foot high dam. A toe drain in a trench can be OK if you have a creek channel to spill into. If the soil is very sandy, a dam with no formal drain usually results in a marsh at the toe.

 

[pelelo]

I guess the question I had to ask was if for example if have a 350 gallons/min (gpm) seepage. Against which parameter should I compare this value to conclude it is an ok seepage value or low or high?

Maybe against the total aquifer capacity?. So if for example, my aquifer is 35000 gallons as total vol capacity, it means that only from seeping action, the aquifer will be filled in around 100 mins?, which might be too fast, considering all the utilities that need to evacuate (assuming this is the case).

Therefore in this case the seepage value is too high.

I haven't found a straight answer of this and text books don't provide this very clear answer. If anyone can provide any input, it is appreciated.

 

[cvg]

?
you need to calculate the exit gradient of your seepage to determine if the toe of the dam will blow out and fail.
you need to calculate slope stability with your assumed phreatic surface
you need to estimate the hyd conductivity to analyze if the reservoir will hold water
you need to evaluate the abutment and foundation to see if water flows in and through them

 

[dgillette]

I don't know of any simple criterion for permissible seepage quantities, and if I saw one, I would probably tear it up and throw it away. A seepage quantity in gpm doesn't mean a whole lot without consideration of the materials the large flow passes through (silty sand embankment, fractured hard bedrock, weathered claystone, etc.), whether it's concentrated at a single location or evenly distributed along 500 feet of dam length, whether it passes through a sand filter, how high the gradients are in the more erodible materials, and like cvg says, whether it could cause slope-stability problems. Especially important is whether the seepage changes with time. You have to start by understanding what governs (or will govern) the location and quantity of seepage. Only then can you start trying to put numbers on it.

Internal erosion causes about half of all embankment dam failures. (Spillways cause about half, by inadequate capacity or poor design causing erosion in bad places, which leaves a very small fraction from all other causes like slope instability or earthquake.)

 

[BigH]

and then there is always the question "does the seepage quantity surpass the ability of the reservoir to pond and maintain itself?"

 

[rilxouss]

Tend to agree with dgillete though, better to control the seepage by providing a chimney or even drainage blanket if you are concern with the under seepage.

Unless it is justified to do numerical modelling as suggested by fattdad, which still going to depend on the reliability the parameters used on your model.