Failure of Fuel oil tank Roof with Structure

[kvbalu]

I have a case of a 32M dia x 20 M height Cone roof fuel oil tank, the roof of which collapsed. The roof alongwith the truss structure caved in. At the time of the collapse, the tank had about 2 Metres of fuel oil in it. The tank had steam coils but there wasn't any transfer from / to the tank. Apart from the roof, two shell courses have also caved into the tank pulling the connected foam piping, appertunances away from their places.

Can the forum members throw light no analysing the failure. The roof had one 8" Gooseneck vent on top with a 36" line feeding/taking out liquid to/from the tank. Storage tank flash point is >66 deg.C with a storage temperature of 90 deg.C. Tank is designed to API650. I am worried about the safety of the other three tanks which are similar.

 

[unclesyd]

Sounds like a collapse, suck-in, from having a vacuum, negative pressure, pulled on the tank.

Can you comeback with more details like was material being drawn off at time of collapse?

Had there been any problem with heating system?

Is the vent working properly?

Any signs of corrosion on the steel?

 

[kvbalu]

Looks like a suck-in due to vacuum. No material was being drawn in or out during the time of collapse. The heating system was working normal (unless there were leaks in the steam coils inside). I would be able to look at the steam coils only after the tank is ready for man-entry. No signs of corrosion on the steel.

I am proceeding on the basic assumption that the vent sizing could have been deficient or Vent could have been blocked or that it was a pure structural failure due to corrosion of the Structural Truss supporting the roof plates.

Can the members suggest on methods to calculate the size required for the vents. API 2000 only gives the capacities that need to be considered !

 

[JStephen]

API 2000 does in fact require vents to be rated by testing- even "free" vents. You might check Varec, Shand & Jurs, and similar vents.

Some of the ASHRAE books have duct sizing info that can be adapted to vents, but it won't technically meet API-2000.

 

[unclesyd]

Thoroughly check the steam heating system in the tank and any steam stations on any lines connecting to the tank. I have seen two tanks that failed due to condensing steam and a vent problem.
Try to be close to the tank as much as possible to see what's being done. Things get changed, especially if there is a highly visible failure.
I use to say drink coffee with the operators and keep a tight lip

 

[kvbalu]

Are there cases of condensing vapours (not steam, but petroleum vapours) creating enough vacuum to suck the roof in? Members may like to advice.

 

[gr2vessels]

Yes, the hydrocarbon gases condensation is the main design point for a condensate tank, where the uninsulated roof will cool the vapour space considerably after a quick storm.The temperature differential could be as high as 40-45 deg C. The tank I mentioned is similar in size (a bit smaller) and has now two off 24" manway size vents, sized for fire API 2000 and vacuum.
Cheers,
gr2vessels