Tank Differential Settlement 1

[pgyr]

I'm providing geotechnical recommendations for a 72 foot diameter water tank 20 foot height. Site at ridge top. About 3/4 of the tank footprint will bear on moderately to slightly fractured sandstone bedrock, expected settlement negligible per borings to 50 plus feet-the rock just gets more competent with depth. About 1/4 of the tank daylights out of the slope and will require engineered fill for support. The max fill height is about 11 feet this to include excavation and removal of unconsolidated colluvium overlying the sandstone. I am concerned about differential settlement of the portion of the tank on fill in relation to that on the rock. Calculated loaded tank settlement on fill by using Janbu and elastic modulus values from the literature commensurate well compacted gravels. At the edge of the tank, settlement of results are on the order of 0.4-0.75 inches. This is not taking into account OCR associated with the compaction effort-therefore arguably an overconservative assessment. Conversely, common rule of thumb for settlement of granular material under self weight is 1/2 percent of height, i.e in the same ballpark.
I will specify fill gradation, lift thickness, and compaction requirements i.e. Proctor plus field test strips. With little tank experience, I don't want to miss something obvious due to ignorance. Structural needs to know what they must design for or if site plan/tank size must be adjusted.

 

[BigH]

I would take it that the tank bottom is "flexible" - i.e., steel shell and base plate? For that settlement (i.e., <3/4 inch)of the fill, I wouldn't expect any problems. The way around it is, of course, to add about 2-3% cement to a well graded crushed rock upfill. This will stiffen up your fill to a degree that you can rest assured to have no problems - presuming that you have looked at the stability of the tank near a slope. In my view you should allow for a bit of crushed soil fill below the tank even on the rock and then top it off with a thin layer of fine crushings for base protection.

 

[JedClampett]

For a tank, there's going to be storm drainage off the roof for the entire perimeter. If (when) there's a real "gully washer" the fill around the tank will get saturated, greatly increasing its weight, 11 feet of saturated soil, plus any change in its behavior due to getting wet, is likely to increase the settlement beyond calculated. I'd expect 2% of the fill height as eventual settlement.

 

[pgyr]

BigH, slope stability analysis shows high FOS.
Cast in place reinforced concrete is proposed. I've informed structural designers of 0.5 inch potential differential settlement, awaiting reply.
Soil cement is an intriguing idea, I've seen mass production of soil cement, silty sand soil, using a pugmill. Can well graded gravel be mixed successfully with a "tilling" methodology? Wouldn't this technique prove less expensive than the mill?

JedClampett, A concrete slab 10'width will surround the tank, properly graded to drain this should prevent fill under the tank load from becoming wetted.

 

[BigH]

I don't think a well graded crushed stone (with or without cement) that is compacted say to 98% Modified MDD will have that much settlement - 2% of 11 ft = 2.6"; can't see that. I've used pugmills for cemented road base course. It will give the best mix - better than tilling and a crushed stone won't be that easy to till. Basically you are using RCC (which is typically 4 to 5% cement) - Surely, you could use transit mixer, add the cement and mix at the site. Would you still need a concrete slab? That would sort of defeat the purpose.

 

[pgyr]

Big H
Elastic settlement using conservative modulus values with a FOS of 2 results in about 0.5 inches. I have never spec'd a fill this thick to support a structure, my experience with fills of 6-7' our local gravels and good construction control leads me to believe that the settlement will be 0.5 inches or less. A belief not strong enough to give it the title of engineering judgement. Yet.

 

[hokie66]

With 3/4 of a concrete tank bearing on rock, I would want the rest bearing on rock as well, probably by boring some piers through the fill to bear in the rock.

 

[BigH]

@pgyr - I agree and that is what I stated - 2% as another poster opined is, in my opinion, far too high - you are, given your experience in your area, the best to estimate and I agree that you are more on target. @hokie66 - understand your concern which is one of the reasons that I suggested using a cement added to a well graded crushed stone fill compacted as a road base. This will give sufficient rigidity.

 

[dik]

Can the loose granular material be removed and CLSM be placed to make the base more or less the same? It's expensive, but 'foolproof'. Can some of the bedrock be removed 2' or 3' if badly fractured, and replaced by compacted fill. The problem is the difference in the hardness of the bearing. This would 'soften' the base. Tanks, in these environs are often 'bedded' in sand with the perimeter of the tank supported on a concrete 'ring beam'; is it possible to put a foundation ring beam on rock, with a sand bedding?

Dik

 

[pgyr]

Questions posed to forum (and to project structural designers) geotechnical questions for structural.
Given a maximum of 0.05 feet of differential settlement over 20 linear feet can a concrete tank be designed to prevent cracking and consequent leakage?
If not then what is acceptable? How are the limiting values of deflection ratio and angular distortion determined?
EM-1110-1-1094 and other references familiar to me give ranges of values for acceptable limits of angular deflection for structures. I assume there is a reference for tanks-I just don't seem to be coming up with one via internet search and my references.

Seems we could limit the quantity of cement treated fill and or deep foundations to only areas required given a limiting value of angular deflection.

 

[EBSEngineering]

I agree with hokie66. Put in some helical piles and end bear on the rock. Then the whole foundation is on the same medium.

 

[Mercycity]

I concur with mr. Hookie66 9(structure)suggestion above.
Although it may look expensive but the whole tank foundation anchored on the rock you may eliminate differential setlemnt and other unknown which may occur in the future.

 

[BigH]

I just think the talk of helical piles etc is overkill. Even without using cement in the fill, his expected settlements are small

 

[pgyr]

Rock under the majority of the tank will be overexacavated to minimum of 3 feet of fill under that portion of the tank.
I am also exploring the idea of preloading with placed fill to simulate tank loading for two to three weeks. Fill would then be removed and concrete poured.

 

[pgyr]

And another option for fill is locally available "washed" rounded alluvial rock, 0.5 inch to 3 inch diameter. Rock is hard quartzite-placed in confined conditions and compacted to form the fill, I would expect less settlement than for a compacted gravel and sand mixture. Anyone care to comment on that assumption?