Rotation of a Buried Propane Tank

[chicopee]

A buried propane tank rotated a few degrees about its long axis. This tank was strapped and it appears that the rotation was caused by uneven tension in the straps-- much like a belt around a sheave and the uneven tension causing rotation of the sheave.

Has anyone come across an article or reference material analyzing the forces of earth pressure on such tank. I am basically looking at the frictional analysis between tank and earth that would resist rotation.

I am pretty much set on the strap analysis that includes preloading and its friction on the tank however, I am somewhat uncertain about my analysis of the earth pressure on the tank which includes semispherical heads.

 

[ishvaaag]

Since you are wanting to get the overall effect of the surrounding ground on a buried tank, it may be useful to model the tank surrounded by ground material, i.e., include the soil in your model. Assume for a start full contact between tank and the soil, and let gravity show its effects on both the tank and the soil, i.e., both deforming interactively under the effects of gravity. The straps' shape and tension will have a say, and that's most you will have to get of one model like this; if the forces in the straps reveal enough differential deformation as to explain what observed you are quite done with the model; if not, you may to complicate more the model to include more soil layers or wedges of different compressibility till the rotation is found modeled.

Nor this model withoug further complications at the nodes nor the one without the ground can fully represent well the situation respect what friction must say about; in any case must be small effect when compared with the overall gravity forces acting and compressibility of the modeled soil. In your case because you will be starting from dubious pressures and in the contact surface case because I think friction will be difficult to represent.

 

[dik]

Any chance the soil was unevenly compacted? I wouldn't have thought that torsion on the tank would be an issue given the (generally)large diameter of the cylinder.

Dik

 

[bimr]

In addtion to the straps, you should also install concrete around the base legs of the tank.

http://www.fema.gov/pdf/fima/pbuffd_chapter_3-2.pdf

http://www.propane101.com/undergroundpropanetanks.htm

 

[bimr]

Rather than using the weight of soil on the tank, a simpler analysis would be to use the weight of soil on the concrete base. See the manhole floatation analysis in the link.

One would think that the resistance to rotation exerted by the soil would be so large that rotation of the propane tank would be impossible. Since the propane tank is also symmetrical, why would the propane tank rotate on its own? You can calculate that resistance to movement using the procedures shown in the manhole floatation analysis.

Is it possible that the propane tank was rotated when the tank initially installed?

http://www.concrete-pipe.org/pdfs1/2008-05-DD_41-manhole-flotation.pdf

 

[chicopee]

I am providing some more details. Tank was strapped with polyesther straps to a concrete pad to offset buoyancy. The straps were too long, therefore the contractor decided to knot one end on shackle. Turnbuckles were used to tightened straps(initial tension). There was water in the excavation when the tank was excavated, about 1/3rd full.

What probably happened, after the tank was buried and rain fell during the weekend, the knot tightened creating uneven tension and rotated the tank a few degrees.

What I am looking for is how to analyze the soil pressure components on the tank spherical heads and cylindrical shell to calculate the soil friction against rotation.

 

[bimr]

There is a method in concrete pipe.org for sliding resistance. You can probably modify that method for your needs.

http://www.concrete-pipe.org/pdfs1/2008-05-DD_41-manhole-flotation.pdf

Having said that, if the tank was sitting in a water bath, the resistance to rotating from the soil was probably nil.