Voids Beneath Bottom Plates 2

[zikzag]

How do you address voids underneath the bottom plate. Should we cut the plate and refill the void with compacted soil? Is there an acceptance or rejection limit for voids beneath tank floors? Your experience and expert advice is greatly appreciated.

 

[gr2vessels]

I can give some advice from my experience;- others can complete the rest of the requested field.
Depends, always depends. The size of the tank and the size of the voids are important. A large tank with significant static pressure will exert a significant bending stress on the floor designed to take no bending stresses. Remember, the API 650 specifically stipulates the full support of the flat bottom. There is a good chance the fillet weld will give in and the floor plate will break. As always, depending on the size, you might dig under the empty tank and use some pressure filling which hardens to similar strength or consistency like the rest of soil. Lift the tank or cut the plate if your calculations show the risk of excessive deformation and weld overstress. Can you afford a leaking tank? if yes, then ignore the leak.
If you need specific help, then you might want to share with us some secrets of your tank like size, content, design standard and conditions, estimated size of voids and some explanation of the origin of those voids that were not supposed be there.
Cheers,
gr2vessels

 

[Duwe6]

As usual, gr2vessels got it. One simple field test is to have 2-3 large, sturdy men [aka Fat Guys] stand on the unsupported area and see if the floor will flex downwards enough to rest on the dirt. If so, your problem is cosmetic and only exists when the tank is empty or nearly empty. The full tank will have the required 'fully supported floor'.

If the floor fails the 'Fat Guy' test, look into cutting some holes and filling the voids with "Flowable Fill"; it is a slurry of sand, water and a LITTLE cement. Flowable has a compressive strength a little higher than well-compacted soil, but is easy to dig into, and like all cenent mixtures is somewhat alkaline so it is 'friendly' to carbon steel. Any concrete supplier will know what Flowable is, and it is MUCH cheaper than properly compacting dirt.

The fix is pretty quick and fairly cheap; do not take the risk of refilling a tank that has failed the Fat Guy test. You may have the floor rip open at a lap seam and "catastrophically deinventory into the local aquifier". $5-15K repair vs risk of $0.5-20MM environmental damages. Not a risk a prudent engineer would sign off on.

 

[LittleInch]

The "Fat Guy Test". Absolute classic. I'll remember that one long after all the others.

When I walk across a tank roof it flexes as well which is quite alarming....

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[zikzag]

Thank you guys for the advice. More info: It is a crude oil tank with a diameter of 70m. It passed the 'fat guys test'. The bottom plate was punctured due to MIC attack. Through the 10mm dia hole, we can see a void approx 40mm (measured gap distance from underside of bottom to the soil). So far, we have discovered 7 holes altogether scattered all around tank bottom but located about 2 meters from the shell wall (radially). We have not cut the plate yet for repair because the tank floor scanning activity is still in progress. Eventhough there are holes on the tank floor, we did not receive any report of bottom leak from Operation. I presume the hole was sealed by thick crude sludge. The plan now is: we will cut a small section of the plate and assess the soil condition for any contamination. If soil is not contaminated, we will follow the use of slurry cement to fill-up the voids.
a) Is this the right way to address the issue?
b) Referring to API 653 Appendix B, it mentions on local bulges and depressions, should we apply the rules only after the plate has undergone permanent deformation? Appreciate if you can share some lights on this. Thanks