Concrete tank storing hot liquid

[patswfc]

I'm looking at the design of a water tank that will store water at temperatures of up to 80 degrees C.
The tank will be below ground with roof slab at ground level. Does anyone know of a good reference for designing such a structure for thermal loads?

I've designed several tanks in the past but have never had to consider such a hot liquid being stored. My initial approach was to consider a monolithic construction (how i've designed tanks in the past), however the fixity at slab and wall edges prevents free rotation of the elements and causes large bending moments in the elements due to the thermal loading. Initially I tried to thicken the elements to deal with the moments however this makes the design moments worse as M = alpha * delta T * E * I / h, where alpha is coeff. of thermal expansion, delta T is the temperature difference between the 2 faces of the section, thus if the section thickness is doubled M increases by a factor of 4 as I contains h cubed! Thus to reduce the thermal stresses I should reduce the wall thickness however I cant make them any thinner due to the lateral pressures from the ground, ground water and stored water, so am stuck!

I am now thinking it may be better to design the elements with pinned edges as this will allow the elements to rotate and deflect without generating moments (thermal strain
without stress).

Other options are to insulate the inside of the tank (assuming monolithic construction) to reduce the temperature gradient through the concrete elements (although I still
think that this will cause major headaches with the RC design) or my preferred option which would be to provide an insulated tank within the RC tank to effectively remove the thermal load case from the design of the RC tank.

Any thoughts on the above or suggested literature on the subject greatly appreciated.

Thanks.

 

[fegenbush]

It depends on how large and what geometry the tank is. Please provide volume and shape so that the solution can be tailored to the specific problem.

 

[patswfc]

fegenbush,

The attached pdf shows the geometry of the tank which is located beside a partially burried waste material fuel bunker (this is an energy from waste power plant).

I have played about with the thermal loading in my FE model reducing the temperature on the inside face of the concrete first to 40 and then 30 degrees C with the outside being at 20 degrees C, to see if we go down the insulate route what temperature we could live with on the inside face. Even with this much smaller thermal gradient through the sections the induced moments are very high. The 30 degress inside and 20 degrees outside would be typical of a heated swimming pool. How would you guys/girls design a swimming pool, would you even consider the thermal load? I wouldnt have though 30 to 20 degrees would have caused dominant load effects, but it would appear that they do!

Any guidance greatly appreciated, I've been banging my head against a brick wall the last couple of days trying to figure this out.