Storage Tank Explosion Overpressure

Are you sure the failure mode was pressure? The deflection "inwards" on the tank side walls indicates a negative pressure event. Just an idea...

The incident happened when the tank was empty and N2 gas was supplied in through the bottom 4' bulking line. Therefore, it could not have cause negative interbal pressure. The buckling damage on the circumferential shell on the top section was as such due to the "pulling in" effect by the bent metal trusses supporting the cone roof from within when the roof was pushed upwards due to the event of overpressure.

How is N2 provided, and how is it admitted into the storage tank?

At what temperature and pressure is N2 stored and admitted into the tank?

How is this process controlled?

How is N2 provided, and how is it admitted into the storage tank?

N2 gas (from pump house) is supplied through piping that was connected to the tank via the bulking line.

At what temperature and pressure is N2 stored and admitted into the tank?

Temperature is not taken into consideration as the tank was empty and due to the fact this was a purging process. The pressure supplied was at approximately 6-6.5 bar but the flow rate was unknown as this was a manual process (not controlled by automatic devices) in which the valves involved were opened into a certain degree (say quarter, half, three quarter or full). The N2 2" line was opened half while the 4" bulking line to the tank was opened quarter. On top of the tank, the 1" bleeder was fully opened while the PV relief valve tested to be functional when installed on a nearby tank.

How is this process controlled?

As mentioned earlier the process was controlled manually. The concept is to purge N2 into the tank to displace air until the desirable oxygen (%) level is reached. Therefore, the operator will go up to the tank roof to measure the O2 level using a special device. When the desired O2 is reached the purging process will be stopped.

think that inside pressure in much less than atmosphere... It supposed that the PV relief valve should be closed (or more close-open ratio), rite?

The operator went up to the tank to check the PV relief valve was functioning properly right after the purging process started. According to him, it was behaving properly with the intermittent release of excessive pressure ( when internal pressure exceed ~25mmbar).

Have you considered temperture differentials? in particular if the N2 was still very cold or liquid, once in the tank it would expand rapidly, causing the pressure to rise very rapidly, far in excess of the actual flow in. The flow through the reliefs would be woefully inadequate and the tank would overpressure. Everything would look fine when you shut down as the relief valves and new cracks would relieve the pressure of the expanding gas. Leaving normal pressure in tank and valves that work perfectly.

Ft Lauderdale water department had a leak in a similar water tank. the water ruptured the wall of the tank at the bottom. The weight of the water above the leak was so great that the flow of water was so great that the top corner of the tank was sucked inward from the negative pressure inside the tank. It looked like yours, only worse.

If you are sure that it was over pressure, then compare the flow rate of the bleeder valve and relief valve versus the volume going in.

But I would bet on suction damage.

Were there any flame arrestors installed with the PV vents? If so, were they not partially blocked? Also check the sizing of both the PV vents and the emergency vents for the worst case. This could be maximum nitrogen flow i.e. all supply valves fully open or fire case.

No flame arrester on the PV relief valve model and none were installed on it. The PV relief valve on the damaged storage tank was removed and installed on another tank for testing; afterwhich showed no indications of malfunction or complete/partial blockage. We suspected the N2 flow as well but we also concern if the PV relief valve had complete/partial blockage which could have been cleared during the mechanical explosion overpressure. Is it possible for the PV relief valve to choke at the first place?

Thank you Mr Abdel for correcting the terminology. Based on the operation conducted, our company is sure this was not a case of implosion; but rather it was an event of overpressure. Our concern is what could have gone wrong during the purging process that contributed to such occurrence. And what can we do to prevent such incident from happening in future? Kindly advice. Thank you all.

Dear Mr. Abdel,

Sir you're right that the term "explosion" is misused here. May I asked if the situation should be termed "mechanical explosion" resulted from increased internal pressure within the storage tank exceeding the maximum pressure limit thus overpressure?

And what are the common improper condition on the valves which could have led to such incident?

Thank you again for your input and response.

Dear keverne,

The condition is just a Deformation due to Overpressure, and can be called an Over Pressurizing until Bursting. We used to use term Explosion only if an actual explosion occurred due to fire.

All valves and instruments assembled to the tank must be tested, inspected, and calibrated regularly in accordance to the related codes and standards and the recommendations of the manufacturers.

"...what can we do to prevent such incident from happening in future?..."

Suggest a pressure regulator on the N₂ line valve into the tank. It's got to be cheaper to protect the tank than to repair/replace it.

Due to the large amount of N2 used and the rate as which it was being charged may have provided a cold nitrogen stream to the tank. With the ambient temperature being elevated, the cold nitrogen could have expanded and exceeded the discharge rate.

The air had a certain amount of humidity in it. The cold N2 caused it to condense out creating a negative pressure and the tank caved in. There should have been a functional vaccum breaker as well as a pressure relief valve. It would only require a small vacuum to damage a tank or pressure vessel in this way. I have seen similar things from my oil refinery days usually after "steaming out" a vessel.

Ed "FuzzyE" Kedz - for some unknown reason I am unable to log on.

just an additional insight . . .

i've read some articles, regarding some major accidents, and most of them are cause by human error. . . since your purging operation was not totally automated and it is manually operated, better check your staff, specially the operator, if they were as alert as they should be during operation . . . example: the probability of reading a magazine while in operation, or talking to someone during process. or the possibility of not being attentive/horseplaying . . . i really believe that part of every engineers duty is safety and alertness . . . . .

but this is me . . . just an opinion and no offense intended

I know nothing about this particular subject. But I was wondering why the tank was being filled with N2 in the first place. If I understand correctly it was to remove moisture from the tank.

You stated that the tank was not cleaned between fillings. Did the moisture come from condensation after the tank was emptied?

Was there really that much water in the tank that you had to use N2 to purge it? Seems to me that there shouldn't have been that much water in it to effect the fuel. I know that underground fuel storage tanks eventually contain some water after they've been in the ground a few years. But even then considering the size of the storage tanks there usually isn't enough water to effect the fuel. There is probably more water ratio wise in your trucks fuel tank than there would be in a large storage tank.

Thanks for putting up with my uneducated ramblings.