Soil Friction to Help Resist A Buoyant Tank

[KennyRogers]

Morning all,

I have an underground tank we are looking at for rehabilitation. I am concerned about uplift due to buoyancy from a relatively high water table. With the all of the gravity/self weights, I am coming up shy on balancing the uplift forces. I was considering taking into account the effect of soil friction on the vertical walls of this tank to help with resisting uplift. Does anyone have input if this has ever been used? It seems reasonable to assume that the soil friction will provide some benefit.

In terms of the analytical frictional force, I was considering taking the at rest soil pressure acting horizontally on the walls and computing a friction force based on the soil coefficient of friction (from tanδ, about 0.30). Since the water table is relatively high, I was going to use the effective stress to compute the lateral earth pressure at locations below the water table. These are clay soils and I was thinking about adding a cohesion component, however since the uplift would be a long term effect, I don't think this is an undrained analysis case. I will however check with cohesion only since there may be some short term effects, which would then transition to long term sustained loading.

To me this seems like a reasonable approach, but wanted to bounce this off someone else. Unfortunately, there aren't too many geotech savvy folks here.

When considering gravity loads only, we would only need to drop the water table about 1.5 feet. I figure we could pump water to draw down the water table, but this presents other issues with where to discharge. The water table was measured this winter, so I am thinking it could have been a relatively high (wet, warm winter) water table, but I am thinking spring could cause it to rise further. If it was late summer, I think we'd have a chance the water table would drop enough to not cause worry.

Thanks!

Kenny

 

[dik]

I would not normally include soil friction because you could have a good backfill drainage layer, and if the soil is 'filled' with water, there may not be much adhesion.

Dik

 

[oldestguy]

What is "rehabilitation"? If a significant job is planned you can do it right, such as anchor it to a concrete slab and get some projection of that also.

 

[GeoPaveTraffic]

I also, would not recommend counting on any soil friction to hold the tank down. There will be some, but if the soil is saturated it will be small. Best to dewater around the tank during the work.

Mike Lambert

 

[dik]

if the soil is saturated

thanks... that's the word I couldn't think of... old age mind fart...

Dik

 

[KennyRogers]

Thanks guys. I thought I was sort of grasping at straws taking into account for soil friction/skin friction, but figured it is a reality. I figured it would act like a drilled shaft and have some resistance. I ran through some computations and I am still pretty far away from being balanced. I was conservative for obvious reasons in computing soil friction, but there is a heck of lot variability I don't have a lot of good soils data. I am going to tell the owner to see if the underdrain system can be redeveloped or install a high capacity discharge well and pump it down. Luckily there is a monitoring well, so we can can keep an eye on what is happening with the water table.

Kenny

 

[BridgeSmith]

It's not really comparable to a normal drilled shaft, which generally has a rough surface in granular soils. In clay, horizontal grooving is required to provide friction. For a drilled shaft with permanent casing, which would be the comparable situation, generally skin friction is neglected as being inconsequential and inconsistent, especially with clay soils.

 

[dik]

Hotrod10:

Another thing for friction piles, caissons and spread bored piles, I always spec that the top 10' is mechanically vibrated... on 40' piles I've had the concrete level drop nearly 1' with vibration.

Dik

 

[BridgeSmith]

"I always spec that the top 10' is mechanically vibrated..."

We our standard spec calls for vibration to a depth of 2 diameters for our drilled shafts. We use clear spacing between bars not less than 5" or 5 times the max aggregate size. We also specify a minimum slump of 7", so that it will flow into the annular space.

"...on 40' piles I've had the concrete level drop nearly 1' with vibration."

Sounds like you're getting some serious voids. I'd recommend following the guidance in the AASHTO LRFD Bridge Design spec listed above. The earlier AASHTO standard spec had a spacing limit of 3x the max aggregate, and it apparently wasn't enough.

 

[dik]

HotRod10:

Maybe a 4" max slump... and almost no vertical reinforcing... often for 16" friction pile, a single 15M x 19'8"... (Bars typically, in this area, come in 18m lengths... and chop them into 3.

Dik

 

[oldestguy]

Hey guys. The subject is a tank, not piles.

 

[dik]

OG: We knew that...

Dik

 

[JoelTXCive]

Our firm designs submerged wet wells and we DO include skin friction on our wells BUT the geotechnical engineer provides the skin friction value to us.

Some caveats:

We use a safety factor of 3 on the skin friction component of our buoyant calculations.

We have expansive clays in Texas so we have to worry about the clay shrinking and creating an annulus around the submerged tank/well. We have a general note requiring that the contractor grout around the well. On new construction they can use small grout ducts, but you would have to drill holes on a remodel.

Attached is the language that the geotech typically provides us. The actual values vary from job to job, but the language is pretty much the same from all the geotechs we use.

 

[fattdad]

I'd use the free-body diagram and safety factors. The FBD includes side shear and horizontal/vertical forces - including both at-rest and hydrostatic pressure.

f-d

ípapß gordo ainÆt no madre flaca!

 

[dik]

Joel... I would not normally backfill with clay... want a free draining material. Do you accommodate the swelling pressures of the clay on the tank?

Dik

 

[JoelTXCive]

In reality, we never backfill at all.

The geotechs always have language in their reports about excavating and backfilling with select fill or sand, but it's not really practical. The stormwater lift stations we design range in diameter from 10ft to 32ft and the depths range from 20ft to 70ft. I can't imagine how big an area you would need to excavate if you actually wanted to drop a lift station down into the ground in an OSHA safe way.

To my knowledge, all of the wet wells are constructed using the caisson method where the contractor just casts ~10ft sections of well and then sinks the well into existing ground. Thus the wells are surrounded by Houston clay. The amount of concrete required to sink the and overcome the cohesion can get crazy. I'm just the EIT, but in my 3 years; I have seen 36+ inch walls several times.

We will often provide lift station plans with 18" walls. The contractor will then submit signed and sealed shop drawings with 24" walls and not request a change order for the additional concrete (must be a high profit margin business or they bid the work knowing they are going to thicken the walls).

 

[dik]

Joel: Thanks... Dik