API 650 tank - Floating roof annular space tolerance

[FrCivEng]

I am looking for information related to design and new built or repaired API 650 tanks floating roof annular space tolerance criteria.

The point is :
- We have a new built deformed tank shell with identified out-of-roundness local areas,
- We have to design and built a new floating roof with an appropriate sealing system,
- We have performed a complete and detailed topographical survey by scan laser
- We have defined a new annular space about 250 mm (10 ft)but the remaining space between the total floating roof diameter (including bumpers on each side)and the as of today deformed tank shell is very narrow,
- The question is "What remaining space value in order to guarantee tank floating roof working conditions in normal operating service at long term ",
- This remaining space is supposed to cover thermal expansion of floating roof (double deck) / shell, construction tolerance defects, ect

Is there any code giving such remaining space criteria or tolerance ?
Thak you very much for your help.

 

[IFRs]

API says the seals must be capable of +/- 4" of movement. I have designed seals for +/- 12" of rim space variation, but it is not very common. Each seal has a different 'thickness' where it is metal-to-metal and can not compress any more or where it will be damaged. Bumpers will maintain a minimum rim space but if push comes to shove the floater will lose to the tank shell and could be damaged and possibly sink. I would do three things: (1) determine if the tank is leaning, if so in what direction, how much, and if the lean is expected to change over time, (2) determine the max and min rim space for each location around the tank with and without the lean factored in and (3) determine the total rim space both sides along diameters coinciding with each location. Start with locations on approximately 3 meter / 10 foot centers and get more data at the bad spots. Plot the data to visualize the situation and then determine what seal design will work. Note that filling the tank can change the shape / extent / size of the buckles and tank settlement can change things also. You may have to measure the tank as it is being filled to get more information. I worked on a tank that "popped" in an out about 8" as it was filled and emptied. Modifications to the floater and a special seal worked just fine. Good luck and please report back what you do!

 

[FrCivEng]

Thank you for your answer.

Some additional infos :
- The tank is 83 m diam, 20 m height, 18 m operating height
- The tank is 1 year new, with 6 months operations only, when at 11 m high the floating roof has sunken during loading
- Shell plumbness checking shows about - 78 mm inside horizontal buckle on radius at 18 m high which is out of (French) code tolerance and a - 58 mm inside buckle on radius at 11 m high
- Diameter measured above bottom corner is within the roundness tolerance,
- Foundation settlement is within API 653 App B tolerance
- So the tank does not seem to lean
- In the deformed shell, from 0 to 18 m high, the max possible inscribed circle diam is 82.844 mm
- We have already studied the rim space according to your (2)and (3 above, this gives a 250 mm annular space for a trapezoidal PU foam primary seal with 430 mm original thickness
- New EFR outside diam including bumpers is 82.750 mm
- The remaining space for EFR movement is 94 mm on diam ie 47 mm on radius when the EFR is centred

Do you mean that 47 mm (1.9") is enough to guarantee tank floating roof working conditions in normal operating service at long term ?
Ot should I consider 102 mm (4") instead of 47 mm ?
What API 650 chapter is talking about 4" ?

Thank you again

 

[IFRs]

H.4.4.3 talks about the +/- 4" and while this technically does not apply to external floating roofs, it is a good guideline. I doubt if the PU seal was designed to do that. The PU seal can't center the roof at all. Without centering devices, the roof will always be grinding up and down at the bumpers. Consider a series of spring loaded rollers to center the roof. What anti-rotation devices do you have? At 47mm I'd expect the seal to fail quickly. I agree that 100mm is a good practical minimum.

 

[FrCivEng]

Thank you for your answer.
Attached are the anti-rotation device dwgs
I guess that with this device it will act as you describe
The PU foam seal will act as centring device on 125 mm gap on each side , but it is not designed for that and so it will be quickly damaged.

 

[FrCivEng]

Attached are the anti-rotation device dwgs

 

[FrCivEng]

Attached are the anti-rotation device dwgs (2)

 

[FrCivEng]

IFR and all, happy new year !
Do the 4"apply on the radius or on the diameter ; shall we consider 8" on the total diameter, corresponding to 4" on each side when the EFR is centred ?
Thank you very much for your help

 

[IFRs]

Happy New Year to you!
The 4" is on the radius. It refers to the gap between the floating roof rim and the tank. A common gap is 8" plus or minus 4" (200mm +/- 100mm). Most wiper type seals can't really do this unless they are about 15" wide at which point you have to make sure the tip will not be in the liquid when the rim space is at a minimum and at the maximum there is still some extra wiper to make good contact with the shell. A shoe seal is often a better choice as it's geometry can be adjusted for extraordinary conditions.