Nozzle Loads 3

[jcfoley]

I know that I may be searching for a non-existent answer, but I have to ask. In lieu of a customer’s specification for a nozzle load, what should a vessel supplier use for this calculation?

It has been a practice for some suppliers to use nothing. I think that this is a just ignoring the problem so that it goes away. Thermal reactions, PRV shock and other loads must be considered by UG-22.

If the supplier is mounting a PRV on top a vessel, I expect that supplier to properly support the PRV to handle the side loads generated or derate the flange for this load. If I am mounting this PRV, them it is my responsibility to specify the side loads to the vessel supplier.

For other piping connections, it is not so straight forward. You can strategically place pipe supports to minimize the loads, but you can never eliminate them. That being said, what loads should be used in the calcs?

Chris Foley
Midland, TX

 

[waskillywabbit]

I think it depends on the scope of your work. If you are an engineering firm then you are being contracted by a customer to determine these loading(s) on the nozzle. If you are simply a vessel fabricator is it the responsibility of the customer or the customer's engineer to supply you w/ these loadings. Specifying inaccurate loads is just as bad as specifying none at all if there are really some present that need attention.

Brian
Pressure Vessels and Autoclave Systems

http://www.mcabeeconstruction.com

The above comments/opinions are solely my own and not those of McAbee Construction.

 

[arto]

how about a nominal F = xxx# per inch of pipe size & M = xxxx ft# per [inch of pipe size]^2, Unless otherwise noted?

or something like NEMA SM-23 or API-610 or Kelloggs

That way you'll get them thinking & they'll realize it's neither an anchor nor "zero"

 

[johnnymist2003]

It is possible for the vessel supplier to "back calculate" the maximum allowable loads and moments and supply these to the user.The problem is that the manufacturer's guess on maximum forces and moments is normally not correct.It must also be borne in mind that there are typically 3 forces and 3 moments in a nozzle loading calculation.When the end user sees the loads they normally revise them to some degree in any case.A number of end user specifications use tabulated loads based on the nominal pipe size, sometimes with factoring for temperature, and/ or flange ratings.When it comes to rections due to PRV's, yes, the vessel manufacturer could probably calculate the correct support system, again assuming that he has the correct information.From my experience it is always better for the end user to specify the loads, this will eliminate any confusion and unnecessary recalculation.
John

 

[warrenw]

If this is of any use to you:
API 560 section5 lists maximum allowable nozzle loads.

 

[DSB123]

warrenw,
What has API 560 got to do with allowable nozzle loads??? Do you mean API 650???

 

[warrenw]

DSB123,
I checked to make sure my Dyslexia haddn't kicked in, it IS API 560 Section 5 that lists allowable nozzle loads. Once in a while one of our clints will requests that our nozzles be designed to these load.
I just thought this may be of some use for jcfoley.

 

[jcfoley]

Thanks for all your valuable input. This topic is is somewhat controversial when you try to make something "standard" (see waskillywabbit's comments). There just are not enough resources to perform this analysis on every single unit that we build. That is why I am looking for a 50% solution so that I can concentrate on the the jobs that really need the detailed analysis.

I have never used API 560 but we buy a lot of API 661 heat exchangers. API 661 has a listing of nozzle loads but I find these loads to be quite high.

Thanks,
Chris

 

[NozzleTwister]

The specification API 560 is for Fired Heaters and the nozzle loads addressed in that standard does not apply to any other kind of equipment.

API 661 applies only to Air Coolers (Fin-Fans).

NozzleTwister
Houston, Texas

 

[DSB123]

warrenw,
Thanks for the clarity. Sorry to have queried.

 

[MJCronin]

jcfoley (nice irish name...)


There have been several discussions of nozzles and nozzle loadings within this forum...For atmospheric tanks, vessels and rotating equipment.

http://www.eng-tips.com/viewthread.cfm?qid=75687

A long time ago, when dinosaurs roamed the earth, I worked for a company in Boston Mass called "Stone & Webster Engineering". This very question came up then on the specification of design loads for new vessels and tanks. After much discussion and the review of various piping stress analyses for similar components, it was decided that an approximation based on the connected pipe diameter was about the best anyone could do....

If you use loads of about ~250D (forces in each of three directions) and ~300D (moment in each of three directions) for nozzles of 4inch to 16 inch NPS ( where D is the nozzle size in inches and forces are in lbs.; moments in ft-lbs)....you will have a reasonable design loading that will cover most cases

Example: 10 inch NPS pipe nozzle
Fx=Fy=Fz = (250)(10)= 2500 lbs
Mx=My=Mz = (300)(10)= 3000 ft-lbs

Again, this is just a general rule of thumb....if nothing else is available

Now, I know there will be objections to this rule of thumb.......so lets hear them !!!!..........

My opinion only......


MJC

 

[NozzleTwister]

I stand corrected on my previous post.

I found out today, that API 537, 2003, Flare Details for General Refinery and Petrochemical Service, refers to API 560 for nozzle allowables as a default if the purcahser doesn't provide any loads to the vender.

Personally I think this is a mis-application. The nozzle table in API 560, Fired Heaters for General Refinery Service, tops out at 12" and I've had very few Flare Stack Inlets that small.

NozzleTwister
Houston, Texas

 

[jcfoley]

Mr. Cronin,

A star for your comments. This is exactly the sort of thing that I was looking for. I knew in the back of mind that it would come from someone with "years & years" of experience. But I didn't think that it would go back to the age of dinosaurs!

Just goes to show that the latest is not necessarily the greatest.


PS: Yes, my family had some deep Irish roots a very long time ago. My son & my wife have red hair so we even look Irish!

 

[StressGuy]

jcfoley - It's funny that you mention API-661 loads being high. We, as a standard, require vendors to provide for 2X API-661 loads on finfan nozzles.

As I understand, some time back, companies specd 3X API. Then, API raised the loads by 50%, so now we call for 2X API to get them same numbers.

Edward L. Klein
Pipe Stress Engineer
Houston, Texas

"All the world is a Spring"

All opinions expressed here are my own and not my company's.

 

[jcfoley]

Yes, I have had to modify my view on 661 loads being high. Since the original posting, I have noticed quite a few specs requiring more.

I am still very curious as to how manufacturer's say that they will comply and then still use the maximum temperature & pressure rating of a B16.5 flange. What happened to calculating the equivelent pressure and derating the flange based on these loads?

 

[davidribeiro]

We are experiencing the same problem here in SA. Lately, the end user is not supplying nozzle loads in time (as the piping has not been designed yet), and the onus is left on the PV designer to specify the loads that the nozzles can take (WRC107). I usually make a note that the piping designer must ensure that these loads are not exceeded.

 

[JohnGP]

A client is asking for heartache by not specifying nozzle loads, or at least a basis for nozzle design, and then expect the PV designer to later handle piping loads. Having been metricated a long time ago now, I commonly use 40D for forces (N) in 3 directions, and 80D (Nm) for the 3 moments, where D is the nominal nozzle diameter (mm). In my experience it is not often that the piping loads are known when the vessel design needs to be finalized.

These loads are applied at the nozzle shell junction, but I have heard of instances where loading is specified at the flange face, which introduces a further moment at the shell interface due to the flange forces multiplied by the nozzle outstand.

I checked against MLC's numbers above and in comparison my forces are about 10% lighter, but moments nearly 5 times greater (if I got the conversions right).

I have found that only for vessels with lower pressure, that I may have to slacken off on the above loads so as not to overdesign. Generally though the resulting design has proven satisfactory when checked against piping loads provided when the piping stress guys catch up.

Regards,
JohnGP

 

[MJCronin]

JohnGP,

Thank you for your post....

I completely agree with the statement that clients are asking for heartache by not specifying any nozzle loads.

Please understand that my earlier posting was based on a "shakey" memory from an employer long ago...I only mentioned them if the specifier had literally nothing else to use.

My intention was to apply these "design loads" at the nozzle/shell junction.

I believe that your methodology may be more useful and the larger nozzle moments a much better idea, but only for pressure vessels (MAWP>50 psi)....not for large atmospheric tanks.

I also recall some corporate "design nozzle load guidelines" where the torsional moment on the nozzle is about three times the other two.

and....I completely agree that the piping stress guys are always the last ones to the table. (sometimes they are not even invited at all !!)

I have always particularly liked the situation where you are performing a flexibility analysis on the piping as the system is being fabricated.

My opinion only

-MJC

 

[JohnGP]

Ah yes, too true. And then quite often the question becomes ".....how much load can that nozzle really take?". That's why I have stuck with the factors noted above - I know they are generally conservative, but save some of the hassle in trying to change vessel design details after an order has been placed, when piping flexibility is being addressed.

I have not had to specify loadings for tanks, but as suggested, the above factors would be excessive. API650 now has a method to readily determine allowable loadings for tank nozzles, and I suspect it is generally easier to position pipe supports for a tank, to reduce nozzle loadings, then it is for some pressure vessels I have worked with.

I only mentioned the fact that I apply loadings at the nozzle shell interface because it does make a difference, and definition of assumed loading point of reference should be made clear to the vessel designer, and piping stress analyst, so that everyone is playing in the same ball game.

Regards,
JohnGP

PS I know all about "shakey" memories......or at least I think I used to.