Nozzle rotation and flexibility for nozzle with internal piping

[PaperEngr]

Hello,

I have an API-650 tank, with two 24" nozzles. One nozzle has internal piping going back to the center of the tank, the other has none. The internal piping would be supported about every 20' with goalpost type supports (no hold-downs). I am trying to determine what effect the internal piping has on the nozzle flexibility and rotation.

Tank Data: D=222', Height = 30', Shell at bottom row = 0.64" CSG=0.88
Nozzle Data: 24", 2' above floor of tank

My App P spreadsheet gives me about 0.79" of radial growth due to product head, 0.47" radial growth due to thermal expansion, and -1 degree of rotation.

My AutoPipe software calculates nozzle flexibility based on API-650 for axial deflection, and circumferential and longitudinal bending.

I figure the nozzle with the internal piping would have close to zero rotation, and zero bending flexibility. It would still have the radial growth. Does that sound reasonable? Eliminating the 1 degree of rotation does wonders for the moments on my nozzle loads - like a 80% reduction.

 

[Geoff13]

The horizontal product pressure on the tank shell would normally cause it to expand radially. However the tank bottom is "clamped" to the foundation by the vertical product pressure and the shell-bottom corner junction cannot expand. This creates a "bulge" over the lower few feet of the shell as the radial growth goes from zero at the junction to the "free" radial growth somewhere above. The nozzle movement calculated by Annex P is how much the nozzle face has moved due to this shell bulge.

The Annex P stiffness coefficient is just a measure of how much the shell is predicted to move under nozzle loading. It is only a function of the tank geometry. It can't become zero by adding internal piping (or maybe I've misinterpreted your words). Hopefully AutoPipe is using the Annex P coefficients.

I suppose you could use the moment from your internal pipe weight to see how much the shell rotates under this nozzle load, and deduct this from the product rotation. Of course as the nozzle rotates back towards zero the internal pipe will deflect down to one of the supports sooner, limiting how much pipe weight/moment you can count on for this reverse rotation. The nozzle rotation will never be zero as there would then only be the pipe weight halfway to the first pipe support causing the reverse rotation.

If you choose to take advantage of this then you must also be 110% convinced the owner will never remove this internal piping, and that the internal pipe will never corrode and become lighter.

 

[HTURKAK]

...I figure the nozzle with the internal piping would have close to zero rotation, and zero bending flexibility. It would still have the radial growth. Does that sound reasonable? Eliminating the 1 degree of rotation does wonders for the moments on my nozzle loads - like a 80% reduction.

The assumption for internal piping with zero bending flexibility and close zero rotation is not reasonable.

What about the external piping? If you post a sketch showing the external piping, internal piping , support conditions ,you may get more valuable comments.

 

[PaperEngr]

Gentlemen,
Appreciate your responses. I will re-incorporate the rotation as per your recommendations.

 

[saplanti]

PaperEngr,

I would consider modelling both internal and external piping to check against Appx P allowable since the same nozzle location will be under the total load generated by the thermal load.