I am performing a stress analysis on a replacement tank line for hot crude oil service. All considered piping was found to be within the code allowable. My concern is at the tank line to tank bottom connection. I don't have acceptance criteria as far as allowables go for this connection.

It is a vertical connection that is welded directly to the tank bottom and resembles more of a drain line than the typical API 650 horizontal shell nozzle. Inside a concrete vault, the 36" tank line is connected to the tank bottom by a short radius 90 deg ell and a 48" x 36" reducer. The reducer is welded directly to a 48" opening in one of the bottom plates of the annular bottom ring. There is also a 0.75" thick reinforcing plate at the opening.

Can Ceasar II validate the tank bottom connection loads as acceptable or not? Are there any suggestions on allowables other than rule of thumb? The connection doesn't fit the API 650 nozzle loading critera. WRC 107 is geared toward pressure vessels (not storage tanks?). Any suggestions would be appreciated. Thanks.

I think that only the tank manufacturer can give you the acceptance criteria. Since you perform a replacement, first ttry to find something that can help you from the existing documentation (ask the owner). If not, th enext thing to do is to recreate the old configuration and to come up with something new that is below that loads.

Regards,

Well, I'm not able to understand what does it means!

The main reason for an annular bottom ring is to assure a proper design and service of shell to bottom junction.
The rotation of the shell to-bottom joint under hydrostatic head induces rotation and stresses in the bottom plating and API 650 Code gives rules, which dictate the thickness and width requirements for the bottom plates which are immediately under the shell. Also, annular plates play a role in case of seismic uplift, settlement, etc.
For this reason any "opening" in annular plates should be avoided.

Anyway, I consider that in case you have piping connected to both shell and annular plates, it is very difficult to count realistic "imposed displacements and rotations" in these two points other than by following a FEA analysis.

Regards.

mariog,


Thanks for your response. Just calling it like I see it. One of the plates in the annular bottom ring is different than all the others. It has a much larger surface area than the others to account for the 48" field cut tank inlet. This plate is butt welded per API 650 just like the remainder of the annular bottom ring plates.


I don't consider myself an API 650 expert. Just curious, is the above statement in API 650? If so, could you please cite the section within API 650? FYI, this tank was designed and built in the 1970s.

Without FEA tools at my disposal, it seems that the most logical thing to do would be to model as danb stated (thanks danb); recreate the old and compare the old connection loads to the new config connection loads. If the client accepts this approach, I'll pursue this approach.


Utilizing CII, if I approach the problem as danb states (comparing the old tank line config to the new), it would seem to me that the difficultly to account for realistic "imposed displacements and rotations" would basically be simplified into whatever is assumed for the old config should be assumed for the new config. For example, if the old tank line connection to tank bottom is modeled as rigid, the new tank line connection would be modeled as a rigid. If one is flexible, the other has the same flexibility. If one has imposed displacements and rotations, the other would have the same imposed displacements and rotations, etc., etc......in any case the new loads at this connection should be less than the old. Thoughts anyone?

Your application is not cover by API 650 (or maybe I was not able to understand what it is!). For Sumps, API 650 imposed some distance as is indicated in Table 5-16—Dimensions for Draw-off Sumps. Looking for a 6" nozzle, distance from shell to sump edge is comfortable. This is not really an argument for your application other than API tries to place the "opening in bottom" to a "comfortable" distance from the shell.

What I tried to say is that both shell and annular bottom respond together to the hydrostatic load, thermal, seismic etc loads. Depending on the shell to bottom size connection, this means that the connection would be subject to imposed displacement and rotations and/or it would restrict the annular bottom/shell response to an seismic event.
Especially for seismic calculation, it is difficult to say something without FEA, but even in this case such analysis applied for seismic case should consider all tank and the connection shell- annular(a very expensive and time consuming one). By the other hand, your Client would ask you to extend the analysis for your connection under the provisions of paragraph E.7.3 PIPING FLEXIBILITY from Appendix E of API 650, case in which I consider it would be very difficult to qualify that connection or to provide an alternative calculation without FEA.