API 650 Tank - Stiffeners Shell for Fluid Pressure

[mmikola]

I was brought into the design of a tank after a client ordered and began the erection of an ethanol processing tank. After running the numbers we realized the tank was not only build, but the shell thickness was to thin.

My question: Is it acceptable by API 650 to provide stiffeners to a shell for the purpose of resisting stresses due to fluid pressures (Circumferential Stresses). I need hard evidence, please no opinions. I need to have something in writing from API that states that this is an option. I have attempted to contact people at the API, however, I need an answer soon.

We developed a finite element model in order to design the "fluid bands".

Thanks

 

[jte]

I need hard evidence, please no opinions. I need to have something in writing from API that states that this is an option.

MMikola-

How do you expect someone to provide "hard evidence" that something does not exist?

jt

 

[mmikola]

I am looking statements either in the code or otherwise, from API, stating that it is acceptable to use stiffening bands and that the method I described above is acceptable by the API. API 650, section 3.6.5 states that an elastic analysis can be performed in the selection of the shell thickness if the L/H is greater then 2 (since the one-foot method and variable design point method would not be acceptable). Can the same be done (elastic analysis) if the L/H is less then 2 even though the one-foot method or variable point method could be used?

I'm not asking to provide evidence that something does not exist...I'm looking to an example where it (stiffening bands and FE analysis) has been done and the approval from API that states the method could be used.

mmikola

 

[JoeTank]

API-650 does not even mention reinforcing bands as a means of tank design. API requires adequate shell thickness. Is that definitive enough?

Joe Tank

 

[JoeTank]

I forgot to mention that API will not "approve" anything you do whether it meets the requirements of the standard or not.

Joe Tank

 

[mmikola]

Joe Tank,

I know that API does not approve the tank, I just want to know if they approve of the method. Also, the shell thicknesses are adequate if stiffening bands are utilized. My client basically wants that warm and fuzzy feeling then when he signs his name to certify the tank is built according to the API 650 that it actually is designed acceptable by the API.

mmikola

 

[JStephen]

The elastic analysis method is allowed on any tank- see the options in the instructions in App. L, under "Shell Design", which allows the owner to specify it or one of the other methods if desired.

However, I doubt this type of retrofit would be considered acceptable by API. (If so, it might be very useful for stainless tanks). An official interpretation would be the only way to confirm for sure. But certainly API doesn't address the issues of designing hoops for tension, welding and radiography of shell hoops, etc, that would be required to make this approach comparable to standard shell design.

 

[zerosum]

When we made nitric acid, we had about five atmospheric tanks of light gage stainless with outer carbon steel hoops (flatbar rolled the hard way with transverse (vertical) stabilizer bars maybe every 60 degrees. (Use angle and you probably wouldn't need the stabilizers).

They didn't carry an API data plate but were quite functional, built before FEA was invented.

I have also seen quite a few municipal water tanks built with reinforcing rings. You might try searching the AWWA for a standard with reinforcing ring details that your client, his underwriters, and the jurisdiction could find acceptable.

I have a sometimes foggy recollection that water and ethanol don't difer too much in specific gravity. (Lighter by 20%?). So a water tank-based beef-up should be conservative?

 

[JStephen]

This isn't addressed in the water tank standards, either. AWWA D100 uses lower stresses than API as well.

I've never seen a municipal tank built with rings for that reason. I've seen decorative rings on a few tanks, and seen one or two standpipes reinforced in that method when the height was extended- and that was probably a bad idea as well.

 

[JoeTank]

Again, API does not recognize shell stiffeners to resist hoop tension. Those tanks do not meet the minimum requirements of API-650.

Joe Tank

 

[mmikola]

First of all I want to thank you all for you opinions offered here. I have spoken with an API 653 inspector and he told me that he has seen tanks in the south using the stiffening bands as I have discribed. I realize that the API 650 does not have a design nor speak of stiffening bands used to resist outward fluid pressures. However, it does again state that an elastic analysis can be performed. Basically, it doesn't say you can and doesn't say you can't use stiffeners to account for excess internal pressures in the shell. The way I see the FE analysis that was performed is that it was an allowable stress method of design using the elastic analysis and the stresses were greatly reduced by adding the bands to shell thicknesses that were as little as 1/16" to 1/8" underdesigned per the one foot method. The stresses that are developed are actually under less stress then the one-foot method thicknesses would have offered. I would really like to here from someone who has been involved with a tank repair that used such a method and if they had any correspondance with the API on such as issue.

 

[JoeTank]

Since you said you did not want opinions, but wanted hard facts about what API permits, I did not go into the area you have chosen to pursue. The elastic analysis method was not intended to be used for stiffeneing rings. It was only inteneded to allow one to look at the stresses at the shell-to-bottom joint in detail. All the courses further up do not benefit much from a more detailed analysis beyond the one-foot method or the variable design method of design.
I and others have used stiffening rings to reinforce tanks due to poor design, deterioration or changes in service. This method is not recogniozed by API-653 either. In fact it was rejected as a repair method in the early drafts of the document. So, in the end you are on your own. You cannot point to a recognized Code or Standard to rely upon to validate your actions. It does not mean that what you are doing is not adequate, but it is clearly none Code. there is of course a business risk to the owner to pursue this approach. If something went wrong (even unrelated to this issue) they will be be hung out to dry for not complying with recognized Codes and Standards. For example, how would you resolve non-compliance with NFPA-30 (which require tanks per UL, API or ASME) Ethanol is a flammable product and I am assuming that your state building Code references this document. As with many things the devil is in the details, so please put lots of control on the ring installation, tightness if fit, NDE, weld spacing. If they can't even design a tank correctly, are you sure the rest of the tank is right. Is the welding up to Code? Do you have the NDE records to verify that? Are the tanks within dimensional tolerances now? They will distort further when the rings are wellded on the shell. Well, I guess that's enough for now. Best wishes on this one.

Joe Tank