Nozzle loadings exceeding exchanger inlet nozzle 6

[MQM90]

Dear

My nozzle loadings are exceeding for exchanger inlet nozzle (drawings attached)

My piping parameters are as follows:

1. Piping material is SA 335 Gr 11 Cl 2 & dia 18 inches
2. Operating Temperature = 350 deg.C
3. Design Pressure = 30 kg/cm2g4.
4. Flange rating is 300 lb (Weld Neck)

Overall flexibility/stress ratio is 0.90 as reported in the software output file.

Can any piping expert or an experienced person tell me how to reduce the stresses on a piping system whose stress ratio are still within the limits but nozzle loading are exceeding.

My idea if i go for higher strength material or high rating flange.

Thanks


Osama Nusrat Ali

 

[ColonelSanders83]

You have two options,

1) additional piping flexibility through loops aditional lenght and the judicious use of supports.

2) Have the vessel nozzle handle more load, this is a simple request to the vessle manufacturer and will cost more money.

It's hard for me to tell from you attached Iso so I need to ask how are you modeling the vessle nozzle the piping is attached to? Are you taking into acount the vessles growths? Are you modeling the nozzle to vessle connection as a rigid anchor at the nozzle flange (with thermal ancor displacements acounted for)? Are you modeling it like a pipe tee (this will gove some flexibility and acount for vessle thermal growths)? Are you using the vessles flexibility to your advantage, using the blijaard, ASME flexibilty method, or vessle vendor provided flexibilities at the vessle shell to nozzle interface?

Just my two cents worth

A question properly stated is a problem half solved.

Always remember, free advice is worth exactly what you pay for it!

http://www.ap-dynamics.ab.ca/

 

[BigInch]

Its no wonder that its exceeding allowables. How do you expect to handle any expansion at all with what is a configuration composed of nothing but fitting to fitting make up?

**********************
"The problem isn't finding the solution, its trying to get to the real question." BigInch

http://virtualpipeline.spaces.live.com/

 

[DSB123]

MQM90,
18in Dia pipe is relatively stiff and you would need to generate significant forces to overstress the pipework. With a stress ratio of 0.9 the the internal loadings within the piping must be significant. Generally with larger sized piping it's not the pipe stress that is the limiting factor for system acceptability it is the loads generated onto equipment as you have discovered. Not having looked at your configuration I think you need to consider the recommendations given above. The best way forward is to increase the inherent flexibility of your "stiff" system.

 

[SNORGY]

Build your model through the connected equipment, including modeling the exchanger shell down to its supports and incorporating whether they are fixed or sliding. The exchanger nozzle (E-311) will want to move upwards and that should help offset the nozzle loads. The nozzle oriented horizontally (R-311) will want to move towards E-311, which will create some issues, but if the support nearest E-311 is sliding that might help quite a bit. Specify both nozzles as such in CAESAR II (or whatever software) and incorporate the nozzle stiffnesses generated.

I high-rating flange may not be the answer, as much as maybe a repad or gusseted nozzle might be. It might also help if you can move the reducer to the E-311 nozzle to take advantage of more smaller diameter piping in the bends.

I agree that the piping looks pretty stiff, being fitting to fitting, but I have seen tight piping in exchanger banks before. On first inspection, this didn't look too bad to me.

Regards,

SNORGY.

 

[BigInch]

Snorgy, You have inadvertantly redefined the term, "It didn't look too bad".

**********************
"The problem isn't finding the solution, its trying to get to the real question." BigInch

http://virtualpipeline.spaces.live.com/

 

[SNORGY]

It's like anything else, I guess. You see if it works; then if it doesn't, you see how close it is to working; then you fiddle with what you can fiddle with until it works. What I saw in the sketch - if the sketch shows the extent of what has been modeled - is an incomplete model that may be painting a worse picture than what might be the case in reality if one expands the analysis a bit.

If after all that there is still a problem, then one can go to the next level towards fixing it.

One would have a better understanding of space and clearance constraints if one was sitting at the CAD station looking at the 3-D model. Presumably, the piping was laid out as shown with a view towards available constraints.


Regards,

SNORGY.

 

[LSThill]

MQM90 (Mechanical)


You may want to run an additonal check using
PCL Gold

Level 3 FEA: PCL-Gold The First Pipe Sttess Tool to Automaticall Run FEA

 

[BigInch]

Then again, maybe he should just address the lack of flexibility.

**********************
"The problem isn't finding the solution, its trying to get to the real question." BigInch

http://virtualpipeline.spaces.live.com/

 

[DSB123]

I'm with BigInch - just address the lack of inherent flexibility of the system - no need to go to PCL Gold.

 

[SNORGY]

Fair enough.

But, the way to address the lack of flexibility is to completely and accurately assess what you have, quantify how far away that is from what you need, and take the required steps to fix it.

Regards,

SNORGY.

 

[BigInch]

Agreed, but its for sure you can solve the problem faster, if you don't start with 350C and F/F.

**********************
"The problem isn't finding the solution, its trying to get to the real question." BigInch

http://virtualpipeline.spaces.live.com/

 

[MQM90]

ColonelSanders83.

"Additional piping flexibility through loops or additional length" is not possible as per field constraints (ohh). However, i have taken equipment nozzle as an anchor with its thermal displacements.

Model is attached.

BigInch,

You are right with a configuration composed of nothing but fitting to fitting, the nozzle load may exceeds its allowables.

DSB123,

Inherent flexibility cannot be increased due to the field constraints Sir.

SNORGY,

If i assume complete rigid anchor at R-311 and move the reducer to the E-311 nozzle to take advantage of more smaller diameter piping in the bends, the piping system stress ratio reduces to 0.89

WONDERFUL APPROACH BUT PLEASE CONFIRM TAKING R-311 anchor as complete rigid anchor i.e. without incorporating thermal growths and shifting reducer to E-311 side. SNORGY is this a correct approach or not?



Osama Nusrat Ali

 

[SNORGY]

Personally, I would avoid modeling the endpoints as rigid anchors for a variety of reasons that could be justified even though it might not constitute good practice within the boundaries of comfort for some (whose opinions and judgement should be paid due respect).

(1) There will be some flexibility at the nozzles at both ends. I would model the system through the equipment as I mentioned earlier to account for these flexibilities, assuming that the software has the capability to generate the nozzle flexibilities (or stiffnesses) for you.
(2) There will be small lateral and axial displacements between flange faces, as well as some gasket crush or gasket compression. Again, not much, but enough to avoid having to model perfectly rigid anchors.
(3) You could try to see if you could find some data for axial, torsional and lateral nozzle stiffnesses - or calculate them (WRC-107, WRC-297)? Then you could model the endpoints using suitable restraint stiffnesses.

Be aware that if you try any of these things, you are really taking away anywhere between "some" and "all" conservatism in your design, and that your results will only be as accurate as your assumptions. I would only take these approaches if I was "close to passing", i.e., within about 25% or so of allowables. Even at that, you are obliged to clearly specify the assumptions that you are forced to make - given your layout constraints - in order to arrive at a model that "passes". You wouldn't normally want to go about things in this way.

Is there a process justification for not moving the reducer to take advantage of smaller pipe? If that is a deal-breaker from the point of view of hydraulics (increased pressure drop), for example, then my opinion is - given the sizes involved - you have other design issues besides stress and nozzle loads.

There are some very smart people here who have expressed their concern that the flexibility appears inadequate to them, based on their first inspection and drawing upon their experience. Even if you do proceed with cutting things a bit finer in your modeling, pay due regard to the advice rendered by those folks. The *best* answer might be just as they suggest: increase flexibility by re-routing the pipe. That might give some bosses and Project Managers / Engineers some heartburn, but detailed design people and stress engineers often have to be the bearers of bad news and, while they can make an honest effort to achieve compromises in their analyses, they ultimately cannot cave to project-driven pressure if something just isn't right.

Regards,

SNORGY.

 

[rneill]

Myself, I usually require that we start by modeling nozzles as fixed and then if we can't reasonably meet nozzle load requirements with this conservative assumption, I relent and allow nozzle flexibility to be considered.

I would caution that you have a solid sound technical basis for determining the degree of flexibility assigned to the nozzle. I had a very bad situation on one project where an EPC firm arbitrarily assigned a very high degree of nozzle flexibility to a injection compressor discharge nozzle in order to get themselves out of a bind in the stress study. It was not until the system was built that I became aware of this - at that point, I noted that there was in fact no discharge nozzle - the mating flange facing was machined into the compressor body (designed and built for 7200 psi) - and in fact the assumption of nozzle flexibility was completely invalid ...

 

[MQM90]

Hi All,

Actually inherent piping flexibility or additional length is not possible as per our field constraints.

I am finally going with the reducer shifted to E-311 side and converting the remaining pipe to 16 inches. And incorporating thermal growths (as per above recommendations) at both ends can atleast brings stress low i.e. 0.85 than the original 0.9.

This way i understand i am not taking care of nozzle flexibilities but in actual i think the system will be more flexibile and will not have any adverse effect on flange or nozzles.

Special thanks to SNORGY , BigInch and ColonelSanders83.

If i am still not very correct, please advice.



Osama Nusrat Ali

 

[SNORGY]

I would have to see how "over-loaded" your nozzles were to make a judgement on the appropriateness of your path forward. But, if you were "over - but close..." then your approach is one I would probably try.

Make sure your calculation assumptions are clearly documented. You might want an internal independent check or peer review for QC purposes. I often request one of my colleagues to cast an independent set of eyes on my results or use a different methodology to see if they come up with something close to what I came up with.

Regards,

SNORGY.

 

[4Pipes]

It reminds me of a modification associated with similar size nozzles on an HE. Why do you want to change the pipe layout - we have 10 identical systems which have all worked for 30 years? On a site visit, I noticed 10-15 fire extinguishers near each system. I asked an operator why. Oh, we have a flange fire every 3 months. No problem - we just replace the gasket. We have done so for the last 30 years!
A realistic full system calc is a must but you also need to consider how realistic the theory is for short legs even if you did manage to get an acceptable answer. In this case, I don't believe very much. I am with BigInch on this one.

 

[SNORGY]

4Pipes:

I don't disagree with either of you.

However...if a properly and accurately modeled piping / nozzle / equipment system gave rise to computed nozzle loads that met "pass / fail" acceptance criteria, would you ordinarily seek to further increase the piping flexibility in order to additionally reduce the loads on the flanges and gaskets and, if so, by what percentage of the loads imposed on the nozzle?


Regards,

SNORGY.

 

[TGS4]

SNORGY, I think that performing "flange" calculations should be an intergral part of any decent pipe stress analysis. (I favour the equivalent pressure method myself). In many equipment, the flange loads may govern rather than the stresses in the equipment.

And I'm willing to bet a malt beverage that the leaking flange in the previous example was an 8NPS Class 150 B16.5 flange.

To the OP, if the stress doesn't work, it doesn't matter how much the project folks of the designer, etc don't like it. It either gets fixed or it doesn't go in. I've gone toe-to-toe with some of the most "important" project managers around. As much as they yell and scream, they can't argue with the "math". Only weak and impotent engineers back down in the face of baseless impositions by (often non-engineering) project managers.