How do I anchor an underground tank?

[TonyES]

I have a large underground concrete tank that is designed to be water tight. The water table could be above the lid resulting in extremely high bouyancy forces. To help get an idea the tank has about 3,000 ft^3 of volume - I would need a 7' thick x 20' x 12' concrete deadman to anchor it too to resist the bouyancy. There is no bedrock nearby to anchor to either. One thought was to pour a skirt - concrete wall that extends around the tank such that the soil above this wall helps keep the tank from floating - this skirt would be 8" tall and as wide as needed to resist the force - but will this work? If I have to assume the water table is above the tank than will there be any soil to help resist the uplift?
Any thoughts?
thanks

 

[ishvaaag]

The question is if the tank has been conceived with strong points where to tie with some advantage of some engineering anchors. You could then tie to anchor piles deep enough. You also can modify the tank design to effectively hold and enclose some upper burden of soil even on floods etc. See starting with your tank what kind of devices can ensure proper anchor, and reasonable alternatives.

 

[cancmm]

Helical anchors may be an option. Relatively cheap and easy to install while providing good tensile capacity in soft soils. Work with the manufacturer (these are generally proprietary) to see the capacities are sufficient based on penetration and spacing restrictions. Be forewarned that any sort of tension pile or anchor will require mobilization (upward movement of a couple inches) before the capacities are realized. If movement is absolutely unacceptable, then you may need to look at other options.

 

[JedClampett]

I would thicken the bottom slab and/or walls until the total concrete weight was enough to counteract the buoyancy (with a factor of safety). You're talking about under 50 cu yds of concrete to counteract the buoyant force. This is nothing, about 6 or 7 batchplant truckloads.
Concrete is cheap, anchors and fancy design details are expensive. Plus it always works.

 

[JStephen]

Can you put checkvalves so that the tank fills if the water table gets that high?

On the deadman idea, keep in mind that you're just basically pouring concrete in a hole, so it's not as expensive as you might imagine. Possibly use flowable fill to backfill and consider the weight of it?

 

[jgailla]

The "skirt" is commonly used in sewage lift station wetwells. You should be able to reduce the dead weight of the concrete needed by using the soil above the skirt. There will be a cone of soil that will act through friction on the sides of the tank even without the lip, although I prefer to use a lower factor of safety against buoyancy and only use the soil directly over the lip.
The lip should help. 3000 sq. ft is ~190,000 pounds of uplift. If you have a 20'x20'x7' deep tank and add a 2' wide skirt, you have 1120 cubic feet of soil, which even at a buoyant weight of 50 pounds per cubic foot is 56,000 pounds. If the tank is buried several feet in the ground, you can also use the soil weight on top of the tank.
You should talk to the project geotech about this.

 

[dcarr82775]

I would just make the thing heavy enough to stay put. Mass concrete is relatively cheap compared to other options.


I have always considered the weak tension link in helical piers to be the single bolt connecting shaft segments. The tension capacities never seem to work for the bolted connection I have seen.

 

[cvg]

there is some risk during construction or during repairs when using a "skirt" with soil load or friction. If it rains hard during construction before backfill is placed, your tank will float. Again during any subsequent repairs, waterproofing etc, where the tank has been partially excavated, you may not be able to count on the full soil load. Better to just thicken the slab to create sufficient dead weight. You can pour a 2 or 3 foot wide concrete skirt around the perimeter, just don't rely on any soil loads or friction, only the dead weight.

 

[hokie66]

Agree with Jed and dcarr. Just ballast it with concrete.

 

[msquared48]

I agree too. Just use CDF.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[paddingtongreen]

I had a case where a the flood stage would come above an underground basement, the tunnels to the basement had entrances above flood stage. I used helical anchors.

An interesting side light on this. The lead company didn't want to do this anchoring, and they questioned our use of a 1.5 safety factor. We agreed that it seemed a little high but was normal. The lead then called a conference of chief engineers from local peer companies. They all agreed that 1.5 was high and that something less would work. Things stuck there until my boss went around the table asking what factor each individual would use. They were unanimous at 1.5 because there was no published tests or recommendations for less.

When I was researching for hold down methods, I found that one engineer had poured a base slab on grade, laid out steel ingots on it in a pattern, drilled dowels in between and then completed the concrete placement to above the ingots.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[VAStrEngr]

We design several of these in many locations where the water table is high. We have explored many of the solutions presented here and almost always discover a mass of concrete to be the cheapest solution with the least question of reliability.

On a few where it has been warranted, we have gone the route of drilled piers but as I mentioned before we typically go with the mass concrete anchor.

 

[HeavyCivil]

What does the site have for grade? A beefy perimeter drain with a nereby daylight in addition to a modest 'skirt' should do the trick.

Assuming your tank has a mat foundation, I think just throwing extra yardage at it can be problematic.

You can't have any cold joints so the second you start pouring you've committed. Not sure where you are located but where we do many of our tanks, in the rural adirondacks, batch plants are often an hour away with only a few trucks. The second there is a hang up - say a broken pumper - your trucks stack up, go over the time limit, and can't roundtrip in time to avoid the previous placement setting up before the pour continues.

Only reason I mention it is because you're doing a small 3000cf tank so I figured you may be in the stix like me. I've been present for the pours this week and last week on a 4000cf tank and we've continuously dealt with hang ups that have made getting these modest pours accomplished without cold joints pretty challenging. IMO the smaller the pour the less can go wrong. Plus - correct me if I'm worng, but with a giant mat arn't you still responsible for meeting ACI 350 crack criteria and would have to fill it to the brim with bar? 50cy of $180/cy mix is $9k extra but that could be exceeded by steel and labor.

 

[hokie66]

Yes, you are wrong. Crack control reinforcement need not be provided for ballast concrete.

 

[msquared48]

VTEIT:

If you have ever seen CDF poured, it is very fast for each truck to empty. Common time per truck is under 5 minutes for a 9 to 10 yard load, then move on to the next. And there are no concerns with construction joints as it is done so quickly. No consolidation, no finishing. No worries.



Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[hokie66]

I suppose you could anchor the tank in controlled density fill (I had to look up CDF), but I would use concrete.

 

[cvg]

CDF is concrete

 

[msquared48]

Sorry Hokied, I guess I do it too. CDF stands for "Controlled Density Fill", and is a low strength concrete, because strength is not a concern, only the density.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[paddingtongreen]

I don't know about now, but CDF was called "flowable fill", mostly we used K-Krete, it was low strength but it could also be excavated with normal equipment so it was suitable fill for where the local dirt was rubbish.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[hokie66]

I think the properties of CDF vary so much from place to place that specifying a concrete with the desired strength and density is the way I would go.