High Temperature Tank Bottom Deformation

[MrVessels]

I am looking for any past experiences that members may have on this subject.
I have a 64 meter diameter high temperature (220C) petroleum storage tank built in 1997 that has suffered extensive buckling of the single fillet lap welded bottom plates to such a degree that one of these lap joints failed in the HAZ adjacent to the toe of the weld, causing subsequent leakage.
The tank is a cone up flat bottom API 650 App M design and the foundation is of conventional packed sand and concrete ringwall construction.
The tank has 4 drawoff sumps that are more or less equi-spaced around the periphery and there was no provision made to allow for free expansion of these sumps which are thought to be a possible contibutory factor in the buckling.
The buckling is asymmetric and approximates the form of a Y with the maximum deformation and failure located at the center of the tank and top of the stem of the Y pattern. It should also be noted that the buckling is so severe that the surface of the ASTM A-36 plate adjacent to the cracked joint also exhibits signs of tearing running parallel to the joint.
This tank suffered a significant seismic event some 6 years ago but no external damage was observed and the bottom only started leaking earlier this year.
After repairing the damage but prior to putting this tank back into service, it is required to introduce modifications to hopefully avoid repetitive buckiling occuring.
Apart from providing free expansion in the foundations at the sumps, any other suggestions based on past experiences of the membership would be apprcited

 

[IFRs]

Is there an annular plate, if so how thick and how wide?

 

[MrVessels]

Yes there is. The tank bottom plates are 8mm thick and the annular plate is 14mm thick and 1150mm wide.

 

[JStephen]

One of the requirements of API-650, when the fatigue design cannot be met, is to maintain the tank with a minimum head of heated product. Is this something that can be done in this case, or has this been done?

 

[MrVessels]

Hi JS.
No this has not been done and cannot be done. The tank sees dual product service and when shifting from the lower temperature service (80C) to the higher it is necessary to empty and physically enter the tank for cleaning. The tank is then nitrogen purged and the high temperature product is introduced. Although the procedures state that the high temperature product should be reintroduced "at a very slow rate" it is not better defined. Also the tank inlet nozzle is flush with the indside of the shell (i.e. no distributor is fitted) so this will potentially lead to an uneven temperature distribution throught the tank bottom plates. It has been recommended to the operator that a suitable distributor be fitted to avoid this.

 

[MJCronin]

Mr.....

Was your tank originally designed for 220C ? Did the design incorporate the temperature limits in API-620

Have you reviewed this thread ?

http://www.eng-tips.com/viewthread.cfm?qid=200237&page=1

What do your original design drawings say ?

-MJC

 

[JoeTank]

Mr Vessels,
There have been a significant number of such occurences to large diameter high temp tanks. But, they are not well documented nor publicized. API-653 makes vague refernce to them as "bulges" in the tank bottom.

Joe Tank