Tank Double Bottom Accidents 1

[sshep]

I almost think Trevor Kletz could have a subchapter on tank double bottom accidents, as we have just had the 5th one in less than 15 years. One was stupidity- purging the double bottom space of an unfilled tank with line pressure utility nitrogen (and then blocking in the vent). The other incidents basically involved using a nozzle going to the bottom space as if it went into the tank itself- i.e. discharging a vacuum truck, ect. The tank will basically balloon up and snap off its anchor bolts.

Obviously every event has had an investigation which recommends labeling nozzles or locking closed ect, and still it happens.

I don't expect to give out any stars for this sort of inquiry but I am interested in any useful infomation about better design practices, reliable administrative controls, if this has happenned to other sites, or even interesting stories on this topic. Thanks, sshep

 

[jalvarez]

sshep:
5 incidents in 15 years seems too much.
Maybe you simply want to make comments on double bottom tanks, not double wall, but I will write here on these last, thinking that they will include yours.
In your case, I think that the design of the bottom -external- space PSV's contingencies criteria should be revised, and/or the size of the purge connections adjusted to avoid the possibility of a human error. I mean, adjust the size of the PSV, or adjust the size of the purges... if you have any PSV. Don't you?
In the couple of double wall tanks where I was involved, a continuous purging with nitrogen was installed. The nitrogen was provided with a regulator valve, with a good hermeticity to avoid losses of nitrogen. So, manual purging is never necessary. We only purge the air with nitrogen for the first start up, and not even the level changes affect the outer space pressure (I'm referring to criogenic tanks at -104°C).
Accidents: I remember only one with a reformer naphta feedstock tank, it is not a double wall case, but it's interesting, so I'm describing it:
During a start up, the naphta feed pump to the reformer stopped (MCC protection) and the check valve in the discharge failed open. 30 bar reforming gas flowed back and the (fixed) roof of the tank failed (blowed, making a 10 ft opening). Bottom bolts, etc. were not affected.
The check valve (simple swing) was mounted horizontally and this was incorrect (the valve maker confirmed this). A survey in all the refinery was made; we found other cases that we fixed.
The failure of the check maybe have occurred before, but never during a start up, when the attention was concentrated in many issues, and nobody realized the feedstock pump problem.
The original contingency analysis of the tank PSV didn't care about the mentioned possibility. We think we were lucky, the consequences would have been worse...
Have a safe day
J.Alvarz

 

[JoeTank]

The weight of a 1/4 inch thick tank bottom plate is equivalent to only 2 inches of water column. Why would you even allow a nitrogen source with much higher pressures to be connected to this space? I think your 15 years of experience speaks for itself. DO NOT DO THIS!

Steve Braune
Tank Industry Consultants

http://www.tankindustry.com

 

[sshep]

Thanks Steve, but I only mentioned that N2 case for your amusement- I think the N2 hose was connected to dry the bottom. No further comments are neccessary on that as it is probably the only investigation I have ever read to cite gross stupidity (basically) as the root cause of an accident.

I am interested in nozzle mix-up incidents, and designs or controls that minimize the chance of reasonable human error.

 

[corrosionman]

There was a case in Northern U.K. a few years ago
A tower was being purged with Nitrogen from a 3 inch pipe.
The safety vent was only a 2 inch dia pipe.
Tragically there were two men on the roof when it blew off.
These things are all so unbelievable when you hear about them.REgards D W

 

[jalvarez]

Nozzle mixing? The most hazardous case I can imagine is the replacement of Instrument air or any air with nitrogen. Though it sounds quite obvious for us, it's very tempting for unexperienced people to use N[sub]2[/sub] in lieu of instrument air if you are in a hurry, and I know two cases where fixed connections were installed in order to have a coverage for an Instrument air upset.
In both cases, these fixed connections were eliminated, and an "air gap" and announcements identifying the risk were installed. Further, N[sub]2[/sub] quick connectors are different from other utilities, to provide an additional protection against N[sub]2[/sub] abuses.
Anyway, I don't remember an accident involving mixing of utilities. In the other hand, I know or was close to some cases where incorrect or abusive use of N[sub]2[/sub] ended with an accident. Two of them, involving fatalities.
Good practices and continuous training are the most effective manners to reduce these hazards.
Have a safe day.
J.Alvarez