local high soil pressure under tank base

[J1D]

I posted a similar thread in Structural Other Topics (thread507-202072) about a tank foundation review. The posting here the geotechnical question related.

A 85ft dia tank to contain sour water about 45ft deep. There is a central column supporting roof, the 20" dia column in turn is supported on a 90" dia base plate on the 0.375" thick bottom plate. With computer modelling, I calculated a 0.35" deflection at the column and a 0.18" at the edge of the base plate. The modulus of subgrade reaction is 150pci (10ft granular engineering fill on sand and till fill).

The question is whether the local 0.35" deflection and the 7ksf local high pressure is acceptable. The stress in base plate and differential settlement are within the allowable.

Thanks for advice.

 

[fattdad]

The areal loading of the fluid contents will cause MUCH more settlement at the column loacation than the axial load of the column. To properly evaluate this condition, you'd need the geotechnical engineer to provide soil stratigraphy and modulus values to look at the compression likely to occur from a 45 ft high fluid over an 85 ft diameter. (Alternatly, you may also need to look at the one-dimensional consolidation characterisics if the site is underlain by saturated clay.)

Please recognize that the center of large tanks often settle inches - not fractions of inches!

f-d

¡papá gordo ain’t no madre flaca!

 

[J1D]

The 0.35" deflection is mainly from column load (about 2/3). The tank content pressure is less than 3000psf, the allowable soil pressure is 4000psf. This may give a clue what a deflection should be.

The 150pci modulus is given by the geotechnical consultants. And I'll double check with them for the calculated deflections. Just want to get some geotechnical idea if the high stress/deflection is very local, what happens.

 

[msucog]

if i understand the conversation thus far, i believe that the "background conversation" fattdad is making is that the center will settle several inches or more due to the large area load (tank liquid). i haven't had to specifically evaluate tanks before but i do know that settlement is not necessarily dependent only on the contact pressures. because the loading area is so large, the stresses will be felt more or less the full depth of the soil profile and not just the upper few feet as with normal column loads. it's sort of like fill induced settlement.
for instance, i'll talk in very round figures relating to something i'm pretty familiar with (area fills). here where i'm at, let's say there's 40' of 20-30bpf material and there's going to be 20-25' of new fill over a large area. without running any tests, i would expect fill induced settlement to be 2% of the soil profile thickness. that's 40x.02/.12=6.5" settlement (some of which will work itself out near the end of fill placement but most of which will occur in the weeks following the end of fill placement). now imagine it's a big tank instead 45' deep (center) x say 60pcf = 2700psf contact pressure. well, 25'x115pcf=2875psf. i say you've got yourself a comparison there...and that's on a pretty good soil profile. my scenario is not exactly how it works but it's close enough for a rough idea. a geotech would need to use the appropriate chart, tables, data, etc to perform a real settlement analysis for your case.

consult a geotech. and remember, moderate loads over big areas equals bigtime settlement even in the best of situations. the other folks on this board can definitely provide a more thorough school lesson than i can.

 

[fattdad]

Bearing stress at 4,000 psf may be perfectly adequate for the structural design for a 2 or 4 ft wide footing. Subgrade modulus values have nothing to do with the deflection realized by the areal loading of the tank contents. Please do not ignore (or allow your geotechnical consultant to ignore) the potential for large settlements beneath your tank.

Refer to this thread for more information and a reference to a paper on this topic by Prof. Duncan (Virginia Tech).

http://www.eng-tips.com/viewthread.cfm?qid=194438&page=4

f-d

¡papá gordo ain’t no madre flaca!

 

[J1D]

Many good points in the post your referred, thanks.

I can see the total settlement can reach several inches. What I am most curious here is whether the high soil pressure (about 7000psf, it is larger than the allowable) adjecent to the column is acceptable because the higher stress is very local. The subbase is 10ft deep engineered granular fill, the soill underneath is sand and clay till.

The vendor of the tank has a strong opinion not to use a footing under the central column. From what I can see, using a footing will make the soil pressure less than the allowable even under the column, but when the stress spread several feet down there shouldn't be a big difference. Also the total settlement of the base may not be noticably affected by a footing. Can you comment on this? Thanks again.

 

[BigH]

I personally don't see the need for a special footing under the centre column. If you want, why not thicken the tank bottom by adding a couple of steel plates directly under the footing - in any even the colunn won't overstress the steel. I have never seen anyone designing a flexible bottom steel storage tank using k values before. I'm assuming you aren't using a rigid raft foundation. For large dia steel tanks, the stress imposed at the centre is higher at any depth than that at the edges - normally, one would estimate that it would be about 2x higher. You can use stress distributions (Bousinesq) to determine the stresses at the centre and at the edges. As you have engineered fill (hopefully well graded sand and gravel and not a clayey silt), you won't find that settlement will be a problem - especially since the base on which the fill is placed is glacial till and you are not using a floating roof tank. Make sure that the engineered fill extends beyond the edge of the tank by some 10 to 15 feet (and for the 10 ft high fill, use a side slope of 2H:1V. In essence, you have produced a tank pad foundation. In the past I have been involved with larger tanks on firm clayey soils. You have to worry sometimes about edge bearing failures in such cases (see Bjerrum, London 1957). There are several books that discuss tank foundations (Lambe and Whitman, Craig) and many papers in the journals - the 1960s seem to be a great decade for such papers.

 

[fattdad]

Here's some further comment: You REALLY need a settlement evaluation of this tank when full. When the settlement is realized at the center of the tank (i.e., whatever is calculated), the support for your center support will be removed. This will lead to loss of support to your tank "roof". Now the question is whether the "roof" can yield with the center. Bear in mind that the settlement of the tank rim will be much less than the settlement at the center of the tank.

Here's where you are kind of lucky: If this site is underlain by glacially consolidated soils, there may be negligable settlement under the load of the tank contents. Not for me to decide, but something for you to evaluate (or your consultant).

¡papá gordo ain’t no madre flaca!

 

[BigH]

fattdad - good point about the roof perhaps not settling with the base!! For a situation that is described - 10 ft of good engineered fill over glacial till, I doubt too many will really carry out anything other than ballpark (viz., back of envelope) settlements.

 

[J1D]

Differential settlement (between the edge and central column of the tank) has been a main concern for Structural.

It is not a rigid raft foundation. The tank is to be supported on a concrete ring beam, subbase and a central column for part of the roof. There is a socket at the column base, to make the column length adjustable at least during the construction. The geotechnical consultant originally proposed a 12'x12' footing under the column, but the vendor is strongly against it, saying because the soil at the immediate outside of the footing will settle more than the footing, this will create a high stress in the bottom plate there. Without a footing, the vendor uses a 90" dia, 2.25" thick plate plate at the column to reinforce.

I'm sending a request for additional evaluation of the tank without a footing. From my view the only concern is the very local high bearing pressure. I personally tend to consider the footing is not neccessary.

The allowabe differential settlement for tank is about 2".

 

[fattdad]

The allowabe differential settlement for tank is about 2".

That's crazy talk. Two inches for a radius of 42 1/2 ft is nothing. You will likely see much more than that when all is said and done.

¡papá gordo ain’t no madre flaca!

 

[J1D]

2" is the allowable differential settlement from structural point of view. Different structures, with different integraty requirements, have differential limits. The value is not the most strict one.

I've sent the evaluation request to the geotechnical consultants. I should be able to tell you what they figure out the total and differenct settlements for the tank described above later.