Storage tank finite element analysis - fitness for service 3

[ultrazero]

Can any point me in the right direction?

I need to do a finite element analysis on an existing chemical storage tank to validate its fitness for service. Is there a software package available? What procedure should I use? The tank is not for petroleum product. The closest that I can come up with is in the API.

Thank you much.

 

[corus]

You could consider this as a storage tank and use pressure vessel codes to assess it. There may be design codes explicitly for storage tanks but you'd have to check. In general these design codes should be enough to assess it and you can buy software to help in that regard. You'd only need to use finite element methods if the vessel design fell outside of the code limits.

corus

 

[JoeTank]

API-650 is also for use in the design of chemical storage tanks. API-653 is for evaluation of chemical tanks too. They're used extensively for that purpose. What portion of the tank must you analyze by FEM?

Steve Braune
Tank Industry Consultants

http://www.tankindustry.com

 

[ultrazero]

Corus & Steve, thank you much for your valuable inputs.

My main interests are out of plumb and foundation settlement (approximate 1" difference for a 50 ft dia. tank). The tank was built 30 years ago. I do not know if it remained that way after construction, or a gradually worsening problem.

Since posting the thread, I came across API-579 storage tank fitness for service. Based on the two replies, I assume it is good for chemical tanks also. I realize if it exceeds the code limit, then FE is needed.

I checked into several FEA software for storage tanks. Coade & Codeware (FEPipe) seem to be the closest match. Can anyone comment on their experience with FE for storage tank analysis?

Ultrazero

 

[JoeTank]

If the one-inch difference is due to simple tilting (rigid body rotation) there is no reason to do FEM. API-653 has no criteia for out-of-plumb for an existing tank. Someone may be giving you erroneous info about API requirements.

Steve Braune
Tank Industry Consultants

http://www.tankindustry.com

 

[corus]

Looking at Coade, I don't think they use FE as such, other than Caesar which is for pipes. PVElite may come from the program PVE5 which was a pressure vessel design software package. This program merely followed the rules and calculations of the design standards and as such saved a lot of the manual work needed to perform the calculations. I'm guessing that the tank software will do the same.

If you still have concerns over the current design that you need to perform a FE analysis then any FE program will do. Look at other threads to see which will be suitable, ie. cheapest. For pressure vessel designs, Appendix A of BS 5500 (EN something or other) is used to assess the stresses from FE analysis. These general rules are applicable for most structures regardless of the application. It's up to you to consider which load case you need to assess though it's best to consider the worst case possible to satisfy yourself, say for example if support was provided by 4 legs then assess it for only 3 supports. As it stands, a 1 inch settlement on a 50 foot diameter vessel isn't going to change the distribution of loads in any significant way, if support is still provided at all points.

corus

 

[MToft]

I've used Caesar for pipework but seen Finglow used specifically for pressure vessels (you can hire it) - we used to treat pipework terminations as constraints in Caesar then impose the loads and moments on the vessel model in Finglow.

Done some work assessing old pipework and associated plant - for large-scale projects like dams it can be worth seeing how progressive the settlement is eg by laser surveying of the tank and datum points monthly, weekly or even daily if really worried.

Supporting soil and concrete properties may be key for settlement; may be worth doing some quick sensitivity studies on these unless you can get some data.

Good luck!

 

[ultrazero]

Thank you much all.

Accurate measurement of the settlement is vital. I am going to investigate laser survey instrument compared with an optical unit. The former should have much higher accuracy.

My most recent discovery: API 650 Appendix B and API 653 Appendix B discuss tank foundation design and settlement. I gather that if the severity of the problem goes beyond the API documents, then a finite element analysis is justified.

I also found Coade sells a "Tank" software program that implements API 650 and 653. It has a module that analyze tank settlement (presumably not using FEA).

Has anyone tried any of the API tank software programs?

Ultrazero

 

[JoeTank]

All the accuracy in world will not resolve the problem that you probably do not know the original elevations. Therefore, the ultra accurate measurements are somewhat useless in that it is the net settlement that you are really after. That is the big mistake that most of the commericial programs make in that they often assume the non-planar measurements are the result of in-service settlement. This is absolutely a great way to make oneself look really stupid to your client. Be careful. It really comes down to this... tank shell are never built in a flat plane. The tank takes the shape of the original foundation grade and that is not a structural problem at all.

Steve Braune
Tank Industry Consultants

http://www.tankindustry.com