waste water aeration tank allowable settlement 2

[babak1977]

we have a 80mx40m aeration tank in a waste water treatment plant, subgrade modulus of the soil is very low, considering structural piles only under tank walls and using FLAC code we obtain the maximum settlement about 700mm at the center of the rectangular tank, safety factor against soil instability is more than 5, and the concrete floor (which is structurally separated from the reservoir walls) thickness is 500mm which acts as a watertightening medium, its stresses are controlled due to its high length to height ratio (high flexibility) and its maximum reinforcement ratio is .02, the only critical issue is whether this big settlement should control the design or not?
note: all connections and pipings are flexible so no limitation is to place on the design due to mechanical and piping issue.

 

[hokie66]

Let me get this right. Are you saying that the walls do not move, but the 500 slab moves down by as much as 700 in the middle? How much does it move down adjacent to the walls? 2% Ag is a lot of reinforcement in a floating slab. Why not spend some of that money on piling under the tank floor?

 

[JedClampett]

I wouldn't do it. You'll end up with a cracked, leaking floor and a very expensive fix. If the piles are too expensive (unlikely), consider improving the bad soil or removing and replacing it.

 

[jayrod12]

All the aeration tanks I've seen on poor soils were fully piled. It's the cheaper route. Thinner slabs and less rebar with negligible settlement.

 

[babak1977]

thank you all,
hokie66: yes got it right, we will use dowels as shear keys between the well-based walls and loose-based floor to constrain the vertical displacement of the floor to the wall at the edge of the wall foundation, and about reinforcement I agree it's high but with dividing the floor to some smaller slabs with structural joints with using dowels between them as shear keys we can reduce the amount of reinforcement substantially.
JedClampett: why will we end up with cracks, due to which stresses or differential settlements these cracks may occur? we may have a very slender beam which sits on two simple supports, its sag may be big, but it will stand loads to the point it would fail (instability occurs) due to yielding and forming a plastic hinge in the middle of the span, I mean controlling the settlement or deflection should have some external structural reason. For concrete slab delta/length due to big dimensions of the reservoir is small, although absolute delta (maximum deflection in the middle) is high. Some cracks occur in concrete and the opening of the cracks is limited by means of reinforcements, and in the most conservative way we can cover the top surface of the slab by some thin layer of epoxy primer or bituminous isolator sheets.
The Floors of the aeration lagoons are only consisted of geo-membranes and not any restriction due to settlement is placed on their design, what make a tank floor this much different in controlling the settlement, when opening of the cracks are simply controllable and the walls for which settlement should remain in low ranges are underpinned by structural piles? All connection are to the wall not to the floor except for the aeration nozzles which in the worst condition can sit on a stainless steel framing connecting to the walls.
constructing piles under the floor is very expensive because the floor elevation code is 0.6 m.a.s.l., NGL is near the sea level and it's fully saturated, and the ground is nearly muddy, using piles will not help in such situation, only remained method is soil improvement (like jet grouting) which is very expensive.

 

[hokie66]

I doubt that anyone here will concur with your assessment of the suitability of this system.

If you can install piling under the walls, you can also do it under the floor. Driven precast piles or grout injected piles would be two options for consideration.

 

[SlideRuleEra]

Babak1977 - For starters, I agree completely with Hokie66, JedClampett, and Jayrod12. But let's not go there, take a look at another aspect of the project:

Constructing piles under the floor is very expensive because the floor elevation code is 0.6 m.a.s.l., NGL is near the sea level and it's fully saturated, and the ground is nearly muddy, using piles will not help in such situation, only remained method is soil improvement...

I have to disagree with you, an adequate number of piling are probably the only cost effective solution. I have spent a career working on various aspects of the design, construction, and (most importantly) the multi-year follow-up of the structural performance of heavy industrial projects (electric utility power stations, 500 megawatt class) with exactly the soil conditions you describe.

From your statement, I assume the tank / foundation is almost completely above the water table - all load, for the entire tank & its contents will be carried by the piling under the wall - no buoyancy from foundations submerged in ground water.

The 80 meter by 40 meter tank has a perimeter of 240 meters. Let's say that one row of nominal 400 mm concrete piling are used under the wall. Minimum pile spacing is normally considered to be 3 pile diameters (or three times the length of one side, for square piling). Therefore, typical pile spacing is no less than 1.2 meters, for a maximum of 200 each, 400 mm piling to support everything (walls, floor, tank contents, piping, etc.) You could put a footing under the wall and get more than one row of perimeter piling, but that makes things more complicated. If my 400 mm assumption of pile size is wrong, it does not really matter. There could be more smaller piling, but less allowable load per pile. Larger piles, more load but fewer piling.

You have not mentioned the depth of the tank so I can't make a meaningful guess at the load per pile, but the 500 mm thick floor alone is over 18,000 kilograms / pile. That by itself, is a lot for a 400 mm pile in poor soils. Have you calculated an average pile loading value that includes the walls and tank contents, also?

How long are the piling? Because of the soil conditions, the piling are certainly not going to rely on friction for bearing anywhere close to the ground surface. Are the piling point bearing instead? How far down? The poor soils are worthless for providing lateral support to the piling. Piling with heavy loads and long, laterally unsupported length = failure by buckling under load.

Just some things to think about, that could be unintended consequences of not having an adequate number of piling.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[Ron]

You are dealing with a 2+ million gallon per day treatment plant. You cannot afford for a structural issue in the future to shut the plant down or to dump 2 million gallons of untreated wastewater into the environment. If the project is viable, it has to be viable at a reasonable cost to counter the existing conditions.

You have not given the soil conditions at the site other than "muddy" and "saturated". Depending on the soil conditions (classification, strength, etc.), there might be some ground improvement options that will work. Jet grouting is by no means the only option for ground improvement....there are numerous ways. A pile supported slab and walls might very well be the more cost-effective solution, but consider other options of ground improvement as well.

For the size of your tank, if the soils are sands, clayey sands, or silty sands, you could consider using a 900mm diameter pipe vibrated with a vibratory pile hammer and inserted on a 2m center-to-center grid, held in place to achieve the desired consolidation of the soil and then extracted, moving on the the next grid point. I used a similar design for ground improvement of the Nassau Harbor cruise ship berths in the Bahamas about 25 years ago. It worked well.

Vibroflotation is another, similar method that might be effective. Both of these will depend on the soil conditions, which we don't know at this point.

 

[babak1977]

SlideRuleEra: tank height is 7 m with 0.5 m free board, soil cohesion is 50 kPa (doubtable because suction effect may affect the test results) , soil friction angle is 10 degree, and the subgrade modulus is 3e5 N/m3, Poisson ratio is .49, so we are dealing with a loose soil which its bulk modulus is very higher than its shear modulus, and it's nearly "incompressible", in such a soil using peripheral piles provides "confinement" and due to soil incompressibility when you prevent the lateral movement of soil its vertical settlement will be limited to any range you are designing your under-wall piling for moments, my main point is: in such soils which continues to depth of more than 100 m without any change in mechanical parameters, piles are neither rely on end bearing nor on friction, and on the other hand, initial elastic vertical settlement due to low subgrade modulus is substantial and due to low surcharge (about 6.5 meter of water) with respect to the soil burden history (which in first 10 m of depth is overconsolidated) the consolidation settlement is not considerable, I should mention that the hydraulic conductivity of the soil is about 1e-10 m/s, which means the soil is nearly impermeable, so not any kind of internal or external impaction or vibrating is efficient, due to extension of the soil parameters to the depths, piling doesn't work well vertically, replacing the weak soil with any better material should be done to depths, but enjoying the incompressibility of the soil by confining it by means of lateral stiffness of piles we can control the settlement, as I said before, modeling the situation with FLAC code and using "moment-resistant-piles" (they can form a diaphragm wall under the tank walls instead of individual piles) total uniform settlement of 700 mm is reduced to 150 mm which is acceptable for such dimensions of the tank.
one big concern in modeling of such materials with implicit finite element methods, is volumetric locking, but FLAC code is based on explicit full dynamic method, and so not any compensating method to keep the total volume fixed such as penalty method or Lagrange Multipliers are needed in this code, I encourage you to utilize this code and model the complicated situation.
In such situation, Piles can not bear vertical loads since shear modulus is very low and settlement of the pile is controlling the load bearing capacity, underpinning the tank floor is not reasonable since no stress transfer is occurred, and the whole soil body under the tank floor will settle with a profile which extends up to 2-3 times the dimension of the tank beyond it, but peripheral confinement reduce the settlement under the tank between the walls, and beyond the walls soil up-heaves, the amount of displaced soil inside and outside the tank walls boundary compensate each other, I know this seems odd, but it works, and the only concern remained is how to deal with differential settlement between the tank wall and connecting waste water channels since this peripheral piles under the wall act like a seesaw,

 

[SlideRuleEra]

Babak1977 - Thank you for the explanation. Your proposed design makes a lot more sense to me now. Sounds like it should work. I'm familiar with the use of piling for soil confinement, but have not encountered a situation where they would intentionally settle with the entire structure while providing confinement.

You have mentioned that the floor is to be divided into relatively small slabs with structural joints between them. During construction would you expect differential settlement of the small floor slabs? I would be concerned that the structure is not at steady state and soil confinement is not complete until all floor slabs are in place.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[LannyBudd]

Sounds like good old Chicago clay...you better go with the piles!

 

[hokie66]

I have great reservations about your plan, but that is possibly because I am unaware of any successful precedents. Note that none of our usual geotechnical experts have commented. Is this a scheme recommended by your geotechnical consultant?

 

[babak1977]

SlideRuleEra: thank you for following, and commenting on this issue.
confining comes to effect when the whole structure is constructed, and it's ready for filling with waste water which exerts the vertical loads on the floor slab to be confined in the soil beneath by peripheral piles, and in this stage all floor slab parts are completed and ready to interact vertically with each other and on the borders with tank walls using shear dowels between them. Note that all parts of the floor slab are undergoing the same load (hydrostatic head) in all operational and emergency scenarios.
Since filling the tank with water occurs in some few hours and considering the fact that the greater part of the settlement is its initial amount then almost all settlement occurs immediate after filling. In the events such as dewatering for maintenance or other purposes, the soil under the floor bulges back to its initial level quickly, structural joints through which water stops are built should stand the rotational and translational displacement due to settlement profile which includes and defines all differential settlements.
hokie66: I'm also a geotechnical expert, I discussed and reviewed my design proposal with my structural and geotechnical colleagues as a part of a Value Engineering, we are seeking for newer procedures rather than traditional and conventional ones, it's not a single project it includes a vast variety of waste water plants which are to be built along the shorelines.

 

[SAIL3]

ok, the hydrostatic load is uniform.....not sure if the soil settlement or it's expansion when the hydrostatic load removed would be uniform....

 

[SlideRuleEra]

Babak1977 - Thanks, I should have done the math... with over 80% of the load being the weight of the wastewater, of course load on the soil is low until the tank is filled.

...peripheral confinement reduces the settlement under the tank between the walls, and beyond the walls soil up-heaves, the amount of displaced soil inside and outside the tank walls boundary compensate each other, I know this seems odd, but it works...

To me, it does not seem odd at all. In coastal South Carolina (and elsewhere along the southeast USA coast) there is a very poor "soil" commonly referred to as "pluff mud". I've had occasions to deal with structures both in and on top of it. Funny thing about a soil this bad, and probably about the soil at the location for your project, it can act pretty much like a fluid. As with a fluid, it can provide a support similar to buoyancy... but without the structure actually being immersed in it. Your quoted description is very similar to loaded pluff mud behavior.

When a large amount of settlement is acceptable, as on your project, properly designed structures on soil underlain by pluff mud usually work. Problems arise when a portion of a structure has meaningful support (such as bearing piles) and the remainder does not. Surprisingly, the mistake happens far too often.

You have convinced me that your design using containment piling, but no fixed support, will accomplish the stated objectives. Perhaps you can keep us updated on the project's status, I would like to hear how it goes. Good Luck

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net