small pond outlet culvert - anti seep collar?

[murdock]

Hello,

I am working on the design of a culvert that serves as the primary outlet of a stream-fed pond. I have sized the structure as an 83" x 57" pipe arch. 100-yr storm flow is 163 cfs, and the pond rises 3.25' during the storm. The embankment of the pond is about 10' tall at its highest point, and is about 60' wide.

I have questions regarding anti-seep collar design:

1. The literature I've read so far has rejected antiseep collars for large structures in favor of filter diaphragms. Is my smaller structure OK to use ASC's, or should I be thinking about a filter diaphragm design instead? If ASC's are still fine for my structure, then:

2. The soils onsite are clayey, and consideration for a clay collar was asked for. Is there a spec for the type/thickness of clay used in an antiseep collar?

Thanks in advance,

 

[cvg]

you have a 10 feet high dam which falls within the possible limits of a state regulated structure. How many acre-feet does it impound? Hope it is less than 50. Either way, I would use the filter diaphram. If you use anti-seep collars, they should probably be concrete. You may want to look at a storm larger than 100-year also.

 

[fattdad]

Don't use an anti-seep collar. I learned in graduate school over 20 years ago that they are prone to failure and have witnessed it myself (and acted as an expert witness in claims). I'd need more detail to give you a design that would be more appropriate, but for my money (and the integrity of my practice), I'd do something else.

Here's an example of what I'd do: If you have a 10-ft high embankment and there's an 8 ft crest width with 3H:1V sideslopes the run of pipe would be about 70 ft. I'd place the upper 2/3rds of the pipe in a concrete cradle to the springline and backfill above the springline with compacted structural fill (i.e., just like you'd use for the embankment). For the lower 1/3rd, I'd surround the pipe with at least 12 in of open-graded aggregate. This may warrant the use of a separation geotextile. Then I'd backfill over the pipe drain with normal embankment fill. At some point where the pipe daylights by the toe of slope, I'd include a perforated pipe in the open-graded aggregate to allow the free flow of water from the aggregate.

This will likely be all that you need.

f-d

¡papá gordo ain’t no madre flaca!

 

[cvg]

before you do this, confirm with the state that it is or is not a regulated dam. They will require a permit and will require the diaphram, plus a lot more...

 

[murdock]

Thanks for the responses. Here's a bit more detail.

The structure is pretty small. The dam impounds about 6 ac-ft, which puts us well under our state's jurisdictional threshold (12 ac-ft).

The pond is an existing farm pond that failed in the past (right now the location of the embankment is a chewed-out part of the stream); our proposal is to repair the breach, provide new drainage, and build a driveway/rural road on top of it. The 10' depth is at the breach; the length (perpendicular to flow) of the embankment is about 30', ending where it meets existing grade.

The permanent pool of water is set at the invert of the pipe, the only soil saturation around the culvert will occur during (big) storm events.

Does any of this change your thinking?

 

[cvg]

if you are never impounding water and your outlet pipe easily handles the 100-year outflow, not sure you have much of a seepage problem. But if you do, still don't recommend the collars.

 

[fattdad]

Murdock,

During periods of high flow (i.e., when the water is above the normal pool elevation and the pipe is running full), you will have transient conditions where flow will want to travel along the outside length of the pipe. This flow will result in seepage pressures. You or your contractor will have diffuculties compacting the backfill of the pipe at the haunches (i.e., below the springline) and these seepage pressures will worry away at that poorly compacted backfill. The best way to correct this is by using a concrete cradle (upper 2/3rds) and a pipe drain (lower 1/3rd). This is an appropriate design for a regulated or unregulated dam.

f-d

¡papá gordo ain’t no madre flaca!

 

[dgillette]

Hi fattdad. When you say cradle, are you referring to something cast before the pipe is put in place, or anchoring the pipe in place in a trench and placing concrete under it to form bedding? I've done the latter. Little helical anchors made for tying down house trailers work for the anchors to keep the pipe from floating.

Last I knew, the state of Florida required seepage collars on some dams. When I worked there (>20 years ago), we just left a space so they could have the concrete bedding envelop the lower part of the collar. Then, we required small equipment (Wackers) around the upper part. If constructed well with proper compaction around them, I'm not convinced that collars are a major detriment (as opposed to being approximately worthless), but proper compaction around them is often neglected. It's less trouble to bed the pipe in lean concrete and wrap a filter diaphragm around it.

Regards,
DRG

 

[fattdad]

DRG, I'm referring to the latter. Step the pipe up on blocks or such and then pour concrete to envelop the lower part of the pipe.

In Virginia our Stormwater Management Handbook also references the use of anti-seep collars; however there is an out in that you can actually get an engineer to design something else. In many states, it remains possible for a barber to design a small embankment but not for an engineer to cut hair. Go figure. . .

f-d

¡papá gordo ain’t no madre flaca!

 

[cvg]

encasement of the pipe generally eliminates the compaction or lack of under the pipe haunches. It also provides better protection against leakage through the pipe joints. However, there is still a preferential path for seepage through the embankment which runs right along the edges of the pipe. Either you have to lengthen the seepage path to reduce the chance of internal erosion, or provide a safe way to convey the seepage through the embankment by using a diaphram. The choice of collars vs diaphram vs nothing should be based on your risk assessment of the embankment and the potential failure. In this case, I would say the risk is relatively low with very little impoundment, very large outlet and low dam, you can probably get away without collars or a diaphram. However, as stated before and I agree, encasement of the pipe is always a good practice with any pipe penetration through a dam embankment.

 

[fattdad]

cvg: I agree the weighted creep ratio along any interface will exist. I also agree that the weighted creep ratio, in principal, is greater when you construct a right angle bend in the interface flow path. The use of concrete bedding is to safeguard that there is no preferential flow path through poorly compacted or otherwise disturbed soil (I think we agree on this). I'd guess that the preferential flow path along the concrete cradle would be more likely at the soil to concrete interface as well as the soil to pipe interface above the cradle. The use of a "filter diaphram" is not necessarily the only solution, however. Surrounding the lower 1/3rd of the pipe in gravel also provides a point of interception along this preferential flow path.

On the matter of tolerable risk, bear in mind that the OP is a chemical engineer and may not have any perception of risk with respect to using anti-seep collars or poorly compacted pipe backfill.

I've seen the results of failure and it's not pretty, even when it's not a matter of human risk. Sometimes it's just big amounts of sediment that end up in the adjacent wetlands. Something about environmental regulators and sediment in wetlands that creates a bad day, eh?

f-d

¡papá gordo ain’t no madre flaca!

 

[cvg]

I agree, unless you have designed dams or levees, most engineers have not done any sort of risk assessment. And I agree, the seepage is nearly always at the soil to rigid structure interface. Surrounding the lower third of the pipe with gravel is in effect part of what I generally consider to be a filter diaphram. The diaphram being a large area of granular material near the center of the dam and then granular pipe bedding on the rest of the pipe to convey the seepage to the downstream toe. All considered to be part of the diaphram.

I've seen the effects of a dambreak with just sediment inundating downstream farm fields. Three lawsuits later and everybody has been sued and the lawyers are now dreaming about the fourth one - the cleanup would have been cheaper than the attorneys fees.

 

[fattdad]

I'm just questioning the need for the "large area of granular material near the center of the dam. . ." I've heard these called "chimney drains" as well as diaphram drains. Both of these can have a benefit in controlling the flow net through the embankment if critical gradients are likely at the discharge at the toe of the embankment.

Penetrations through dams are a different matter (in my mind). If you have a dam that does not need internal drainage for the typical embankment condition, then the only precaution is to control drainage along the penetrations. For this case, that's when I'd just consider the use of a pipe drain for the lower 1/3rd of the penetration.

Hopefully, we have not confused the OP at this point. . .

f-d

¡papá gordo ain’t no madre flaca!

 

[cvg]

I agree the "diaphram part" is overkill on this particular project. However, not all dam regulators will agree with your philosopy that penetrations through dams (even dry flood control structures)don't require diaphrams. I know several, that don't and they will point to failures that they believe could have been prevented with better seepage control.

 

[murdock]

I was enjoying the back-and-forth between you two. Thanks for the open discussion, it's an educational experience. Maybe I should change my handle, since I do more stormwater/site layout than chemical engineering now. However, I have not done dam/embankment design, and I know that it is a tricky science, hence my posting here.

My thoughts at this point are that since I am only dealing with limited periods of innundation (greater than 1-year storms), I will be OK with an anti-seep collar.

All that being said, I am now looking into a precast concrete box culvert as an alternative. Will anything change if we decide to go this route as opposed to a metal pipe arch? My preliminary thoughts are that with gravel bedding, I will be providing a preferental pathway for any seepage that would occur and due to the greater mass, would provide a more stable structure to build my road on top of.

 

[fattdad]

Murdock,

Irrespective of whether you use a pipe or a culvert, I'd do the same thing. I'd envelop the lower third of the culvert in open-graded aggregate and I'd make sure that the upper two-thirds of the box culvert was in intimate contact with the subgrade soils and backfilled with embankment fill. I would not "bed" the culvert in open-graded aggregate for the full run as that would certainly create a preferential flow path.

It never hurts to ask local engineers what they do either. . .

f-d

¡papá gordo ain’t no madre flaca!

 

[cvg]

pre-cast box culverts are a bit tricky but may be a cheaper alternative to construct. The joints are highly prone to leakage, if, you don't specify, detail and construct them properly. There has been some studies done and I have attached one which may be useful for your design.