Differential Settlement of Circular Concrete Water Tank

[ghgeo]

For help in preparing the scope of work for a geotechnical investigation, I’m looking for some thoughts on how to evaluate the differential settlement across the footprint of a circular concrete water tank.

From the request for proposal: The proposed tank is 400-feet in diameter and contains interior columns spaced about 20-feet on center. Perimeter footing and interior tank footing loads are estimated to be on the order of 5,000 pounds per foot. The maximum differential settlement between column locations is limited to 0.25-inches. Inquiries into the tank height / water height have yet to be answered.

Based on previous work experience in the area, it is anticipated that the Site will be underlain by residual soils (micaceous silty-sand / sandy-silt) that will grade to fair rock (RMR rating) at a depth of approximately 10- to 15-feet below the existing ground surface. The weathering profile of the underlying soil/rock tends to vary both vertically and horizontally

My ‘gut’ tells me that the bearing capacity of underlying soils will not be an issue, but I’m not sure how to evaluate the differential settlement, especially with the potential for the modulus of elasticity to vary horizontally.

Any thoughts, methods, references, etc. would be greatly appreciated.

 

[civilperson]

Too tight of a differential settlement specification unless an enormously stiff slab is under all the structure.( 1/4" between elements). Try either steel design which is more flexible for soil deformations or limit the individual settlements to less than 1" which averages out to 1/2" differential settlement.

 

[STVU]

Very restrictive tolerance for sure-L/960 of differential settlement. Only thing that comes to mind is either have the footings be founded on deeper soils and take advantage of buoyancy effects or place it on piles.

For differential settlement claculations, NAVFAC DM 7.2, I think chapter 5, has quick formula for settlement of continuous & column footings in granular soils based on K value.

 

[BigH]

Is this a water storage tank or does it have to do with clarifiers in a sewage plant?? Just wondering - normally the later have tight tolerances but for a water supply tank I am not sure why it would have - consideration of connections? They could be flexible. . .

 

[fattdad]

I'm confused by the original post. It seems that the interior columns are supporting just the roof load of the concrete tank. It seems that the desire is to use column footings (for the roof support) that are sized for 5,000 psf, which may be reasonable for residual soils. I would think that (depending on the near-surface soils, you could confirm that 5,000 psf would provide the desired maximum settlement of 1/2 inch and the empirical differential settlement of 1/4 inch (i.e., half of the total).

Here's where the body is burried: The water load over a 125,700 sf area will mobilize a seat of settlement that will trump any structual loading. The settlement "bulb" will mobilize greater settlements in one portion of the tank then others (it's not necessarily from middle to edge - refer to "Differential Settlements in Steel Tanks", D'Orazio and Duncan for some insight on settlements beneath large areal tank loads.

Back to the original post: If you are desiging a geotechnical field program, I'd consider using a dilatometer to get a modulus profile with depth. Sure, you may also want some SPT data (for sample collection and correlation to regional data), but I'd also supplement with a few DMT soundings.

I think the RFP and client desires may not be fulfilled!

f-d

¡papá gordo ain’t no madre flaca!

 

[aeoliantexan]

I don't have much experience with residual sandy silts, but quite a bit with tanks. Some issues you need to consider are:

400 feet is really large. I presume the walls will be prestressed. How much experience is there with the behavior of the walls of prestressed tanks that large?

The edge of the tank will probably settle only about one fourth to one third as much as the center. If the rock at 15 feet is relatively incompressible, the full differential due to geometry will occur within about 15 feet of the wall. The floor settlement profile will be shaped like a dog dish, not a saucer. I haven't seen a large tank that settled less than an inch, so it is very unlikely that the 1/4-inch differential settlement criterion can be met between the wall and the outermost row of columns.

Big tanks tend to have big pipes coming in and out, often through the floor. There is likely to be a 90-degree elbow under the floor that is supported on or near the rock surface, so the pipe entrance will settle very little, but the floor a few feet away will experience near the maximum settlement, creating severe distortions in the floor near the pipes. If the pipe backfill is more compressible than the undisturbed soil, the problem becomes worse. Thick backfill in a narrow zone often gets placed quickly without the necessry compaction and testing. Tank builders like to use very thin "membrane" floor slabs, which will not span the backfill zone. The results can be disastrous.

I agree with the others that the differential settlement tolerance between columns is very tight. It probably doesn't need to be. It can probably be met only with a very expensive pile foundation and structural slab or extensive site improvement.

We really need to know how deep the water will be.