Seismic Design/Analysis of Horizontal Tank

[David A]

Hello all,

This is my first time posting. I have been to the forums many times over the years, but never have I been this stumped. Hopefully someone has some experience with this design.

We are evaluating an existing tank for a client located in the bay area of California. The tank is a 2,500 gallon above ground tank containing vehicle fuel. It's horizontally oriented, cylindrical tank and is constructed within a secondary containment/double wall enclosure. The tank base is a completely flat bottom. We have been tasked with evaluating the anchors and restraints for overturning and shear. We analyzed it at a rigid, non-building structure. The regulator has denied our calculations as deficient and states that we need to use ASCE 7, Chapter 15.7 for the basis of our analysis. I disagree with their code reference, as 15.7 is for design/analysis of tanks and vessels supported at the base, like a water storage reservoir or grain silo. They also want an analysis of the available freeboard and resultant force of the sloshing of the tank contents. The freeboard equations provided in Ch. 15.7 assume a vertically oriented structure with constant cross-sectional area with regards to the height of the contents. A horizontal cylinder has a varying cross-sectional area.

Here's my question: Is there a better reference or design guidance for horizontally oriented tanks? Am I off base here, or is the regulator incorrectly siting code?

Any help would be greatly appreciated.

 

[MotorCity]

I disagree with each of you.

The tank is certainly not a component. Typically in ASCE 7, "component" refers to an architectural, mechanical, or electrical part WITHIN a structure. In your case, your tank IS the structure.

I also agree that the part of ASCE 7 referenced by the reviewer is geared more toward vertical, flat bottom tanks. Personally, I would not try to adopt those same provisions for a horizontally mounted tank

such as yours. I suggest taking a look at API 650 for guidance.

 

[David A]

Thank you MotorCity. I corrected my first post to state that we analyzed as a Rigid, Non-building Structure. I was incorrect in saying component. API 650 Appendix E has design criteria for flat bottom tanks, but here again, it reads as they are describing flat-bottom vertically oriented tanks. I have found a few white papers describing analysis of cylindrical, horizontal tanks, but no code references specifically called out, yet. I will continue my search!

 

[JGard1985]

I've got a seismic screening and analysis book for nuclear power plants. Looking at screening for horizontal tanks, there's no discussion on sloshing. In this book weight of the fluid drives seismic forces and longitudinal frequency of the tank (rigid/flexible). Longitudinal frequency is based on stiffness of support saddles of tank

Jeff
Pipe Stress Analysis
Finite Element Analysis

http://www.xceed-eng.com

 

[MotorCity]

That makes sense. Seems like there is a higher potential for sloshing in a vertical tank than a horizontal tank. Not much room for a whole lot of free board in a horizontal tank.

The biggest challenge will probably be trying to convince the reviewer of this logic.

 

[David A]

MotorCity, that's exactly right. I'm just pounding my head on the wall right now with this one.

 

[JStephen]

I'm not sure of your exact tank arrangement. But note that 15.7.14 in ASCE 7-10 covers saddle-supported horizontal vessels, which will be the closest configuration to your horizontal tank.
I don't think there's a requirement to consider sloshing in that section, though.
For vertical cylindrical flat-bottom tanks, the shape of the liquid part is fixed well enough that there are canned approaches for considering sloshing. For horizontal tanks, this is not the case.
If a horizontal tank is completely full (maximum seismic mass), you have no sloshing. So that only comes up in a partially-full tank. Many of the tanks/vessels in this configuration have a lot more strength relative to a sloshing wave than a vertical tank, so it is less of an issue for the tank itself.

With large vertical tanks, consideration of sloshing increases the required freeboard, but actually lowers the seismic forces on the tank (due to the much lower frequency involved with sloshing).

Neither API-650 nor AWWA D100 address horizontal tanks in any way, and the seismic approaches used there are not applicable to horizontal tanks.

You might find some seismic info in some of the vessel handbooks- I don't think Megyesy or Bednar have anything useful on the topic, Moss's book might, but I don't have it. (Scanning Bednar's book just now, he does mention that the "rigid structure" approach is probably appropriate for horizontal drums on two saddles.)

If you can actually allow your tank to slide around, that might lend itself to alternative approaches.

 

[SnTMan]

David A, as I recall, for tanks / PV's designed to a recognized Code or standard there are specific provisions in ASCE for the same. I don't have the ASCE to hand, sorry I can't be more specific.

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[David A]

JStephen has the right provision in ASCE, but there is no mention to sloshing. Thank you guys for your time!

 

[LittleInch]

Just so you know you really shouldn't have double posted in two forums

https://www.eng-tips.com/viewthread.cfm?qid=449124

If you do do it because you didn't get responses, always include the link to the previous post.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[dhengr]

David A:
I’m not an experienced tank designer by any stretch, and many of you know the code specifics much better than I do. But, it seems to me that you might have a problem which needs to be addressed in several separate steps. Your picture and some of the others suggest that the primary tank is a horiz. cylinder on a couple saddles, and would be analyzed as such. And, I agree with JStephen that sloshing should be much less an issue in the horiz. tank than in a large vert., flat bottomed storage tank, which would be much more susceptible to their detrimental and unbalanced actions. Thus, it has typically been (probably been) ignored in normal horiz. cylindrical tanks. You need the exact design and construction details of this system so you know how to treat the two saddles as they attach to the foundation in your analysis, maybe through the bottom structure of the outer rectangular containment container/tank. Then, the second problem pertains to the outer rectangular containment tank, which might be partially filled if there is a failure in the primary cylindrical tank. In this configuration sloshing could/would be an issue. If you look through EQ design and analysis textbooks, ASCE Jounals and papers written by Prof. Anil K. Chopra from Berkley you might find something to hang your hat on. He and some others out at Berkley did a bunch of work on liquid sloshing due to EQs.

 

[bones206]

Is your tank like the one in this photo?

I would think it would qualify for rigid mass analysis per 15.7.6.1(a). Unless it is Risk Category IV, which would trigger the sloshing calc per 15.7.6.1(b). This rigid mass approach (V = 0.3S_DSWI_E from Section 15.4.2) is what I usually use with horizontal flat bottom tanks like this. This part of the code has always given me some consternation because it directs you to use rigid mass analysis in part (a), then goes on to say hydrodynamic analysis is required within the same clause as if part (a) wasn't there.

I sort of agree that 15.7.14 is also applicable to your tank, which directs you to assume that the tank and contents are a rigid mass when designing anchorage and foundations. It's fuzzy how to meet that requirement, but I would argue that the above equation from 15.4.2 meets the code intent in 15.7.14.

Putting confusing code requirements aside, IMHO it just doesn't make much sense to consider sloshing on a tank like this for the reasons stated in the OP. But you can try to point to 15.7.6.1(a) or 15.7.14 to back up your stance with the regulator. Good luck