One of the rules of thumb I've heard is: in lieu of allowable nozzle loads, if the pipe stress at the nozzle is less than 5,000 psi (or 10% allowable), then the nozzle loads are probably acceptable.

I would assume this stress is found in the operating case reports where the forces are "real". Whereas the occasional/final load cases used for evaluating the allowable stresses contain "fake" forces due to the matrix math involved.

Is this correct, and are there any other guidelines/references to similar rules of thumbs?

"... then the nozzle loads are probably acceptable."
"Is this correct"

Probably.

Now, we fully expect you to be willing to stand with your face next to that particular nozzle for the next decade or so while that vessel is in service.

Except it's not that simple. You're asking us if it's ok to keep your face up to any and all nozzles based on this assumption.

I'm not willing to say yes to this question. Any number of things could have gone wrong from the beginning to the end, and that you're being asked to qualify something you don't even have real numbers for only proves that something HAS gone wrong.

Unfortunately it's all too common for equipment manufacturers to not provide allowable nozzle loads when it's included in the specification and/or the client to no longer have the applicable documentation.

I understand that FEA would be best practice.

When I took the ASME B31.3 course, the instructor was the one who provided the piping stresses < 10% allowable stress rule of thumb.

I can relate to having to wade into pipe stress waters, without the appropriate armaments to perform your work completely, as all too many of us have had to in the past.

Going back and rereading, I see that you were non-specific on what your nozzles were attached to.

Based on this logic, all you'd have to do is make pipe extra thick, i.e. stiff and higher nozzle loads, in order to force things to pass.

For that reason, I reject the notion that pipe stress less than a specified value means all nozzles pass, or even all nozzles of a specific kind of equipment pass.

A pipe is different from a vessel, which is different from a pump, which is different from a plate frame heat exchanger, which is different from a box header, which is different from an FRP tank, which is different from an API tank.

CAESAR has some options for some pieces of equipment built into the calculation modules in the "Analysis" tab. You need to open the User Manual, Technical Guides, etc and read about those if you haven't already.

For all other items, look into the equipment data sheet and the equipment design standard. They may have guidance. Your company should be providing you access to these standards.

If the equipment manufacturer is contractually obligated to provide permissible loads, and isn't, you need to send that up the flag pole. Someone's not doing their job.

Be prepared for additional shenanigans, which may include but is not limited to "0 permissible loads." This was unacceptable 50 years ago; it's unacceptable now.

Failing all else, be prepared to send your calculated nozzle loads to the manufacturer via certified transmittal for the manufacturer to approve. The manufacturer might not be able to approve the loads you've provided. The manufacturer might have to alter their design to approve the loads you've provided. But if the manufacturer can't even determine if the loads you provide are acceptable, then you and your client probably need a new manufacturer.

Hello,

If I introduce stiffners on 36" tank nozzle, will the the allowable of tank nozzle will increase.

In that case is it correct to use api 650 or FE stiffness in caesar as the nozzle is stiff.

Ashish P Joshi

Be careful trying to apply API-650 stiffnesses. The rules and guidance in API-650 Appendix P came from limited testing of full-scale tanks, and the assumption is that your tank has a diameter greater than or equal to 120-ft.

"Will the allowable of tank nozzle increase?"

Generally yes, but not always. If you add too much stiffness, your stiffener stops absorbing strain and just passes the loads on to the unstiffened portion of the vessel.

Here's an analogy.

A vehicle bumper protects the driver from the other drivers on the road.

Does this mean stronger and thicker and more bumper means more protection?

No. The purpose of the bumper is to absorb impact. It stops absorbing impact and transmitting it to the rest of the vehicle once it becomes significantly stiffer than the rest of the vehicle.

The same applies to nozzles - I don't care what kind of nozzle, API or otherwise.

"Is it correct to use API 650 or FE stiffness."

It isn't wrong to be inaccurate, so long as your inaccuracies aren't so inaccurate that there's a real world failure when there isn't a theoretical failure.

It's even acceptable to be wildly inaccurate, so long as your results are conservative, but only so long as you do not result in undue waste of resources.

Edit to add: Generally, FEA will be seen as more accurate, especially the further you deviate from what API-650 was designed to emulate.

Michael,

I agree with your assessment of vendors saying "0 permissible loads" is unacceptable. The issue my company seems to frequently have we're never in a position to force the issue.

The client we work for, who ordered the equipment, tells them to provide loads. The vendor will say "thats what it is" or something to that effect, and include the nugget of "well we could run *FEA or some detailed analysis* but that is a cost adder". Then we, the engineers, just have to deal with it. It's quite frustrating.

I imagine if the client/plant engineers were doing the design and were in the position to say "we won't accept it unless you provide the allowable loads per the spec" something may get done.

Is there a particular kind of equipment you had in mind?

For what it's worth, I did find a written reference in Pipe Stress Engineering by Peng.

The old rule of thumb was limiting the pipe stress to 6,000 psi at the vessel nozzle, which is assumed to be rigid. This 6,000 psi limiting stress is generally considered sufficient conservative for thick & thin wall pressure vessels.