Ring Wall for Oil Storage Tank 2

[jeg]

Consider:
an oil storage tank 30 metres diam. and 14 metres high, full of oil, sitting on a 50mm sand/bitumen cushion, on top of a compacted, crushed stone pad.
The stone pad is retained by a circular ring wall, 30 metres diam and approx 1.5 metres high (0.6 above GL and 0.9 below GL )

I can calculate the hoop tension in the ring wall and provide rebar to suit.
BUT, I am having a BIG problem with thermal expansion.
On a hot day, the wall expands and the stone fill shakes down a little. Cold night, wall tries to contract, but the stone fill prevents return to original size. Next hot day further expansion, and only partial contraction at night. After several cycles of this, the concrete will tear apart in tension and the rebar will enter the plastic zone and continue to stretch.
If my scenario is correct - how does one prevent it happening?
If my reasoning is wrong - can someone point me in the right direction and suggest how to calculate the stresses involved?

 

[vmirat]

I'm surprised that there is enough thermal expansion to make a difference since there is probably only about 25% of the ring wall exposed to sunlight. Anyway, just fill the inside of the ring wall during the day when the inside diameter is at maximum. Then, all you're dealing with is compression of the aggregate bed by the ring wall at night.

 

[DaveAtkins]

I engineered repairs (which were only partially successful) a few years ago to a wall very similar to what you are describing. The storage tank it supported contained hot liquid asphalt, and the tank was completely empty in the winter. The wall had vertical cracks at about 3'-0" oc. Sika CarboDur straps around the perimeter (sort of like hoops around a barrel) were used to stop the cracks from getting any worse.

I came away from that project with the following opinions. First, if at all possible, DON'T PUT A RING WALL UNDER A STORAGE TANK. We have many tanks around here that merely sit on a gravel base on the ground. Second, if you do use a ring wall, PUT THE REINFORCING IN TWO LAYERS (ONE EACH FACE OF THE WALL). The wall I was dealing with had one layer of bars, in the middle of the wall, and I believe if it had two layers, the outer bars would have greatly restricted the cracking.

Other than that, I don't have a good answer for you. I think it is accurate on your part to be concerned about the thermal movement.

 

[jeg]

Thanks DaveAtkins, unfortunately the industry trend, driven by CIRIA Report No. 164, is toward ring walls for large tanks on stone pads.

 

[austim]

jeg,

Clearly Dave Atkins has good advice. However, if you are really stuck with a ring wall partially exposed to thermal effects, then I question some of your logic.

First, I am assuming that your reinforcement will be capable of resisting the calculated ring tension resulting from the lateral pressures in the contained fill without any assistance from tension in the concrete. If that is an incorrect assumption, then it ought not to be.

Now suppose that you are quite correct as far as the first expansion, infill shake-down and first cold night contraction are concerned.

If the passive resistance that would be developed by any attempt to restore the infill to original diameter is greater than the cracking load of the wall, then the wall will crack and create its own random 'expansion joints' ready to take up the next expansion.

So the next cycle of expansion will be taken up firstly by closing up the cracks,and any further increase in diameter should only occur when the daily temperature exceeds any previously achieved temperature.

In effect, you should only have to worry about one full contraction movement from the maximum probable temperature to be expected within the design life of your tank.

I would recommend that you calculate the total likely 'crack width' from such movement (plus the calculated total shrinkage), and put sufficient well-engineered contraction joints in the wall to accomodate that movement (plus an Austim safety margin of at least 30%). I would aim to reduce the concrete section by even as much as 50% at each joint, to prevent cracks occurring between your joints.

Finally seal the joints with the best product that you can secure, and you should have the basis of a successful project.

 

[jeg]

austim
your logic is elegant and reminds me of very long continuously welded rails where the anticipated expansion is taken out in advance by stretching the rail before welding to the next length.
Yes, the hoop tension is a function of the lateral pressures and I shall certainly incorporate the Austim safety margin.
Many Thanks