Stress analysis for large diamater storage tanks

[jguillory]

I am trying to do a stress analysis on a 42" nozzle in a 200' diameter storage tank. Talking with another engineer, I have found that because of the large diameter, WRC-297 does not apply, and API-650 APP-P is limited to the bottom two shell rings. This nozzle is near the top of the shell. I have been provided nozzle loadings (forces and moments in all axis). My method of apporach was first to use several formulas from Roarks flat plate analysis to develop stress for input into 2D mohr's circle formula to find principle stresses, but was left with an axial load that was not accounted for. I have tried to work through the 3D analysis but have had trouble understanding the procedure. Any help would be appreciated. I can provide my calcs through email for review.

 

[TGS4]

If you're near the top (and near the roof flange), Roark, Biljard, etc are of no help. You're kinda stuck with an FEA.

 

[jguillory]

I made the assumption of flat plate because over an inspection area of about 80 in, there is only 3/16" arc

 

[TGS4]

It's a reasonable assumption. However, if you are near the top flange, then the boundary conditions for a Roark-type evaluation are probably not respected.

Will this nozzle see any hydrostatic pressure?

 

[jguillory]

Yes, there will be about 11 ft of water column pressure on the nozzle. when you say near the top flange, are you talking about the top of the tank or the top of the nozzle flange. The tank is 34' tall and the nozzle is installed at 22'-9". OD of the reinforcement pad is about 88"

 

[jtseng123]

Recommned you contact Chicago Bridge & Iron. They are expert in storage tank. I believe your problem might have been done many times. No need to create new approach.
Your pad is 23" wide. We have a common practice that intermediate weld is required if pad width > 16 x pad thickness to ensure no separation of pad to shell, and to ensure loads will be well distributed. This is for typical vessels, not knowing if it is suitable for tank or not. Check it out.

 

[JStephen]

Some considerations:
1) If no one has performed an analysis of the nozzle, then presumably no one knows what the stiffness of the nozzle is (nor the deflections of the nozzle due to hydrostatic load in the tank), in which case, the proposed loads are likely to very wrong. Check back for the sources of those loads before spending too much time trying to confirm the stresses. Garbage in = garbage out.
2) Consider detailing the nozzle and/or the piping to minimize the loads transferred into the nozzle.
3) Generally, with the WRC-297 or related analysis, you'll run into issues with the t/R ratios, not the absolute diameter of the nozzle or the tank, check specifically into the limits.
4) If the nozzle is near a fixed roof, the stress situation, so far as applied loads go, should not be drastically different from a nozzle near the bottom- you still have a nozzle in a cylinder near a fixed end.
5) Typically, cylindrical shells will be very strong axially, and very flexible radially. If the primary loading is vertical, it likely doesn't merit a stress analysis.

 

[jguillory]

Thank you all for the direction. I have decided to go the FEA route. Intermediate welding is not a common practice is AST's, but is a good idea considering the size of the reinforcement pad (about 88").


JStephen -
1. the loads provided are the loads from the piping that will be connected to the flange. How would you go about giving an educated guess to the deflections? Assumptions have to be made, but I don't have any experience dealing with nozzle loading on tanks this large.
3. I was unclear with my diameter comment, but was trying to say the same thing...at a diameter of 200', the shell thickness would need to be on the order of feet, not fractions of an inch. In this case, the shell thickness is 3/8" thk
4. This is an open top tank so that would not apply, but assuming it would be a fixed roof, would I be able to use the head pressure of the fluid above and still apply API 650 app P Calcs(run the calcs as if the tank was up side down)? That is a very creative idea and if it would be applicable, that would be of great use to me in the future!

 

[TGS4]

Now that you've decided to go the FEA route, request the analyst to calculate the nozzle flexibility for you. Feed that back to the piping group to recalculate the nozzle loads.

 

[MJCronin]

Consider use of the fine products from the Paulin Research Group.

Purpose developed FEA sofware is available just for your analysis

http://www.paulin.com/Products.html

Regards