Implosion of Pressure Vessels. 7

[march1971]

During the process of start-up of one of our flowstations,while pumping down the high liquid level from a vertical crude oil surge vessel with two pumps, the vessel suddenly imploded. All pumps were immediately stopped, facility closed in and wells secured. Pleases what colud be the possible/general cause of implosion of pressure vessels and the remedies.

 

[athomas236]

march1971,

Heard of a similar case in August 03 that was caused by blocked air vents. When the tank was drained the pressure in the tank was reduced sufficiently below atmospheric pressure to cause the tank to buckle inwards.

Have photo if you are interested.

athomas236

 

[arto]

Run a check with ASME Code SEc. VIII Div 1 par's. UG-28 to UG-34 to check external pressure design

 

[gpape]

I saw one in Houston where the outside of the tank was being painted and the painters had put plastic to protect the diaphram on the PSV and forgot to take it off. They turned on the pumps and the 100 ft diameter X 30 ft high tank was sucked to the ground!

 

[march1971]

Thank you all. Grateful for your information.

Thanks,
March1971

 

[unclesyd]

Pulling a vacuum in a tank with a pump is a quite common occurrence and as stated most tanks as normally designed want handle more than a few inches of water, vacuum.

Most API tanks if not vented properly can be sucked in simply by gravity draining.

Another good way is to use an extra pump to speed up the process and again if the vent is sized for the additional pump the tank is sucked in.

Going down the list any high melting point material that has a vapor that can condense and solidify in vent needs to have a special conservation vent, heated, to keep this from occurring.

Water vapor in some areas condensing in vents can freeze and plug the vent.

Please don't try to pressurize the tank to get the wrinkles out as the consequences can be quite dramatic.

 

[chili]

It's good practice to always check your vessel for full vacuum. It is a fairly simple task via Megyesy. In our process plants today you just never know what may happen. Even birds can plug a critical vent! Painters are infamous for taping over pipes and openings.

If you have checked and you know that a vessel can or cannot implode, you can then go into your start-up or test with either confidence or acute awareness of what might happen if every vent is not checked.

 

[jay165]

Something similar happened at a refinery where I worked at years ago. Two things contributed to the tank collapse:

1. Undersized vacuum relief device (two pumps running instead of one) could not handle the air in breathing requirement.

2. Internal corrosion weakened the roof structure. The roof was a dome type self supporting design with no rafters.

 

[qualityManager]

Probably not associated with your failure, but I have seen the following failure states. FYI for other readers:

1) Empty tank was cleaned by a CIP (clean-in-place) process with 170-190 F caustic solution. Tank drained leaving misty vapor in tank. Next, hit tank with 50 degree F cold water rinse. Vapor condenses to water faster than vent can keep up - watch tank collapse.

2) Maintenace crew places strip of duct tape over vent tube for some reason and leaves for next tank drainout. Watch tank collapse.

3) Tank clad with closed insulation cover and heat transfer jacket in between. Jacket leaks over time and forms just enough water in the space so that when the correct condition arrives (low tank volume + steam in jacket = steam created from the leaked water and raises pressure in insulation space) client can watch tank collapse.

4) Plant material handler clips low point manhole with forklift. Since the vent is not sized for an 18" drain wide open - watch the tank collapse.

This is anecdotal type information, but in my experiance 90% of vacuum failures have to do with a blocked or closed vents combined with operator errors.

Cheers,
WRW

 

[nur12]

To athomas236: Please, send me this picture also, I am working on presentation about equipment failures. Thanks a lot!

 

[waskillywabbit]

I saw a PV just recently that was pulled together in the center so dramatically that is looked like an hour glass. The failure of the PV was caused by a faulty pump where no backup had been installed. Needless to say, they had it built without full vacuum as a design criteria. All the new PVs we have built for them since come designed for full vacuum, which in its simplest case, caused the PVs to be the next size, 1/8" thicker in shell and head thickness. Not a whole lot of material cost, but a bunch of headaches saved in the long run.

Design PVs for full vaccuum if it is even a remote possibility.

Brian

 

[jte]

Folks-

We've often discussed books in this forum, but there's some good reading out there which I've not seen discussed. Check out "What Went Wrong? Case Histories of Process Plant Disasters" by Trevor Kletz (

http://www.amazon.com/exec/obidos/tg/detail/-/0884159205/103-0581154-6870261?v=glance).

Its handed out to all of the engineers at my worksite and actually makes for some good reading.

A quick check of the index in this book (4th ed.) shows three sections under "Tanks, sucking in" and two entries under "Vessels, sucking in."

jt

 

[nur12]

I think I forgot to mention my e-mail: nura11@hotmail.com. Jte, thanks for the reference, it's a great sourse, but unfortunately it does not have a lot of pictures (as least I've got this impression), and my presentation is Powerpoint based on pictures. Is there any other source?

 

[MJCronin]

march,

e-mail these people....they have records/photos etc of implosions and other accidents

http://www.accidents-inspections.com/Investigations/Accidentrecord.shtml

Let us know what info you get....

MJC

 

[LuizSouza]

Hi, Thomas,
Could you send that photos too.
Thanks in Advance.

luizhenrique_99@yahoo.com