Tanks on slab on grade

[JStructsteel]

I have 2 12' diameter x 30' tanks on a slab on grade. My first concern is do I take a foundation down to frost depth, or do I add a granular fill with drainage pipes under the slab? The contractor wants to do a perimeter foundation wall around the slab (1 side is against a building) but I dont see the benefit of a wall down to frost outside.

Next concern is overturning. To me a empty tank situation would be the worst case with wind, then perhaps a full situation and seismic loading.

What thoughts do you guys have?

Thanks
Jim

 

[phamENG]

Everything sounds about right to me but the seismic. If they're holding liquid, I believe there's a worst case between full and empty where the sloshing of the liquid can exacerbate the seismic ground motions and jack up your base shear/overturning load.

Since it's exterior and unheated, the foundation wall isn't going to do anything for frost heave that I can think of. It could help with erosion and prevent undermining of the slab.

 

[r13]

A ring foundation is recommend to minimize differential settlement, and provide ample space for anchorage. For overturning, the worst case can either be the tank is empty, or full, due to seismic acceleration and the additional hydrodynamic effect.

 

[JStructsteel]

They want to use a slab on grade/mat foundation for both tanks. No specific foundation for each tank. At 12' diameter I tend to think a mat foundation is best, wouldn't you get the extra capacity/bearing of the mat compared to a single foundation around the tank?

I would have the thickness of the slab be such that the anchorage is accounted for.

Would the tank provider provide loads to the foundation, or is that asking too much?

 

[phamENG]

They might, but it's a bit hit or miss. I just did a foundation where the equipment supplier did give me reactions, but they were just single vertical, lateral, and moment numbers. No indication of what is wind, what is dead load, what is live load, how the wind was calculated (return period), etc. So I had no way to go back and run my own loads and tweak until I figured out what they're numbers meant so they could be appropriately applied to our building code.

Other times I've gotten nice IBC based load charts with more information than I felt like dealing with to develop load combinations.

 

[r13]

Yes. You can locally thicken the mat foundation below the tank. Once you have satisfied all requirements, depends on location of the application, the mat could have a depth deeper than the frost depth anyway. But I would still isolate the mat foundation from rest of the SOG due to settlement concern and associated complications. The gradation of the backfill and compaction are very important, you are better of to consult a geotechnical engineer.

In the purchase order, the owner can request the supplier to provide support reactions, stamped calculation, and even require the supplier to furnish the anchorage. All depends on the contract setup and cost.

 

[bones206]

I went through this battle recently. I had designed a mat foundation for a tank with a thickened base below the tank and thickened edges (~2.5 - 3 ft deep) for scour protection and containment wall overturning resistance. I specified a generous layer of granular fill below the mat and sloped grades all around to direct surface water away. I considered the mat/containment slab as part of the OT resistance, but kept it thinner overall for economy and locally thickened it under the tank.

During shop drawing production, the contractor told the client my design didn't meet code because the thickened edges didn't extend down to frost depth (5 ft) and they marked up the shop drawings with 5 ft long skinny frost legs all around the mat. The contractor is also our competitor who is EOR for another concurrent project at this facility, so I think this was partly political. But in the end I lost the argument that the frost legs were unnecessary/ineffective, because everyone just kept pointing to the code. In my opinion, drainage and non-frost-susceptible fill is the more effective treatment. Otherwise all the highways in New England would have 5 ft deep frost legs going down the sides... I think the code does have verbiage that allows for other engineered means of preventing frost movement besides just having some part of the foundation extend down to the frost depth and saying that magically fixes the problem.

 

[r13]

bones206,

I understand where your idea came from, the problem is if you don't protect the backfill by deepen the foundation and relying on soil improvement only, then you need to replace the entire soil within the influence of the frost reach (means down to at least 5' depth , and extend a distance from the footing/tank edge for you case).

 

[JStephen]

Generally, a 12' by 30' high tank will require anchorage for wind. That said, there's quite a few tanks around the country that should be anchored and aren't. The issue is that weight of the product helps anchor them, so they don't blow over to the extent that you would expect.
Quite often, seismic loading will not require anchorage by itself, but if the tank is anchored, then seismic loading will control anchor bolt design and foundation design.
If the tanks are built to API-650 or AWWA D100, the tank standard specifies wind and seismic loading, anchor bolt requirements, and some foundation requirements. Otherwise, that may need to come from ASCE 7. On some small tanks, you run into issues as the standards give minimum bolt sizes but also require bolt anchorage adequate for full yield strength of the bolt, so you can wind up with 500# of anchorage required while you design for 20,000#.
We would normally run foundations to the frost line, but in our area, that's seldom more than a foot or so.

 

[bones206]

retired13 - does a perimeter wall really eliminate the presence of moisture in the soil below the slab? I don't think it hurts, but I don't think it helps that much either. The temperature of the soil under the slab is going to be the same whether you have a frost wall or not. The only thing you can try to control is the moisture content and I think drainage is the best way to control that.

I hear your point on replacing the soil over the full frost depth. I've done that several times in more temperate climates where the frost depth is around 2.5 - 3 ft. I generally coordinate with a local geotech to determine what replacement depth and engineered fill can provide sufficient frost protection, based on experience and local soil types. I've heard some people using rigid insulation below the slab, but I've never tried that nor had a geotech suggest it.

 

[r13]

We all remember the ingredients for frost action - freezing temperature, source/presence of water, and susceptible soil, without one of these, frost will not occur. So, I agree with you that cutting down the water is a solution, but we need to indicate the ways and means on the drawing, or spec, to quell the questioners.

We use rigid insulation on the interior surface of perimeter elements of SOG, and extend it a distance into SOG per recommendation.

 

[phamENG]

If the seasonal water table is below the frost line, then yes - the slab with frost walls would keep the soil dry as long as it is sufficiently waterproof and develops no significant cracking. If the water table is above the frost line, then no - the wall won't help as water will enter the frost zone beneath the slab from underneath.

For an exterior slab, there's not much of anything you can do to control the temperature. Rigid insulation under an exterior slab may slow down the heat transfer and knock the tops off the peaks for short term temperature swings, but if you have a prolonged period below freezing the temperatures will stabilize and the ground will freeze. The rigid insulation under a slab really only helps keep heat in a building, and then in the soil for the stuff at the foundation walls.

So the only thing you can really do to be sure is to use a soil with sufficient strength and enough voids to absorb the volume expansion of the water as it freezees.

 

[bones206]

There's also a capillary zone above the water table that can wick moisture up into the frost depth.

 

[phamENG]

Good point.

 

[r13]

For local with high ground water level, a well planed drainage system needs to be deployed, it is frequently seen pumping well in conjunction with subdrain grids in industry applications. The freezing is a top down and lateral spreading action, for properly installed thermal isolation material, a barrier is formed near the perimeter of the foundation to block out the freezing temperature, and due to the installation above, the inner soil is less susceptible to the variation in ambient temperature. It is an added insurance to well designed/prepared soil.

 

[bones206]

One other thing to consider if you do put in frost walls. The walls create a potential adfreeze surface area that could lead to uplift around the perimeter. To prevent this from occuring, the backfill around the frost wall should be well drained material like gravel so adfreeze cannot develop along the wall.

 

[HTURKAK]

jrisebo (Structural):

You did not define the content of the tanks and frost depth. If the content is water , refer to AWWA D 100, in case of oil , refer to API 650. The tenk manufacturer is expected to provide info . if anchorage is necessary. Depth /dia ratio is 2.5 ..implies anchorage will be necessary at least for wind loading . The foundation could be ring or mat..In cae of mat, the thickness could be in the range of 500 mm. Regarding the frost depth, if the upper must soil is frost susceptible, i will suggest the soil exchange with frost free material.

 

[JStructsteel]

These are tanks that will store milk. Right now I have 2 tanks on a mat foundation approx 16'x32'x12" thick structural slab. I havent refined my wind calcs, but first look seems to have them ok for overturning, bearing based on that slab thickness. Im going to see if the tank folks will provide anchorage requirement, loads to confirm.

Frost depth will be 32" below grade.

 

[HTURKAK]

jrisebo (Structural):

The common foundation slab for both tanks is not a good idea. Separate foundation is better for the case; one of them is empty .
I will suggest separate mats circular or hexagonal, with thickness 400-500 mm, and having nib wall all around with depth 32" .(inverted bowl foundation). If the uppermost soil frost susceptible, soil exchange with sand and gravel ..

 

[Geoff13]

Like JStephen, I'd expect 12'Ø × 30' tanks to need wind anchorage to the slab. Not sure how often a milk tank is full vs empty, but the contents may be all that will hold it in place without anchor bolts.

I ran some quick numbers using API 650 clause 5.11.2.1 requirements, and either 2 or all 3 checks say anchorage would be required. My assumed numbers may be off a bit, but two of the checks requred anchorage by a wide margin. The wind shear could even make the empty tanks slide on the foundation.