We did a stress analysis for the pump line, in which the vertical load acting at the anchor point for the sustained case is well within the allowable limit (say only 15% of the allowable). But during the construction / erection of pump piping, the deflection of the pipe at the end of pump flange (before the pump connection was made) is (10mm downward) from the first support of the straight length pipe which includes Y-strainer weight. Owing to the above, from construction they provided an additional permanent support in-between Y-strainer and pump flange to limit the deflection.

Is it, the responsibility of the stress engineer to consider the load case of erection while analyzing it (i.e. a load case “W”) or the responsibility of the construction engineer to provide a temporary support at the end of the pump flange to limit the deflection?

Hi Paldex,

It is the stress engineer who has to check the pipe for safety, I think its a good practice if you include the erection case also while doing the analysis. Some time it helps. I am doing it in all my jobs. In erection case you dont have to considered the density of fluid.

Regadrs,
RK

Frequently the construction outfit is nothing to do with the company that engineered the plant, and consequently the designer cannot predict how the erection of piping will proceed. In such a case the stress engineer cannot be responsible for any construction or erection cases.

Thanks Mr. Rk and Mr. Moverz,

Is there is any construction practice, that the pump piping should not have any deflection during construction time or when the pump is under maintainence. Also, is there is any practice that the temporary support shall not be used for the pump piping during erection or while the pump is under maintainence.

Please clarify it.

There is no way that you can desing for all possible construction cases but you should consider the cases that you know will occur. You know that at some point during the operation of this pump that the nozzle will be disconnected from the piping and your design should account for this.

In this case I think it would be the stress engineers job to check the deflections and stress when the pump is disconnected from the piping. For each Nozzle on a vessel or a pump the pipe should not be overstressed when that nozzle is disconnected from the piece of equipment. If it will be that information must be passed on to the owners maitenance and operations group.

Thanks for your comment. Its correct that designer can not predict the erection of piping. But though we are considering the stand by and operating cases. What I want to say is just we need to check that if there is any sagging when we the pipes will be ereceted on site & there is no fluid inside it.

Regards,
RK

Rotating equipment connections are somewhat special as millwrights are involved in making sure everything is properly aligned.

The truth is that, if your pipe flange at the pump has significant deflection when disconnected from the pipe, you have failed to provide a properly designed piping system.

At our firm, one of the cases we run is exactly a case of, springs locked, pipe empty, flanges unrestrained at the pump (I picked up a trick a few years ago of changing the stiffness of the CNODE anchor at the pump flange to 0. Otherwise, you can delete the rigid elements that make up your pump representation from the model).

I generally look to keep deflections in these cases to less than 1/4". What you want to avoid in a piping layout is a support arrangement that depends on a perfectly balanced see-saw.

Also, you will want to make sure that the support closest to the pump nozzle is either on a spring, or has a threaded adjustable base to help with the field alignment.

At least from my history, if the millwright can't put the flange bolts through with a minimum amount of jacking the pipe, he will refuse to sign off on it. That's one of the reason's we've taken to checking the disconnect case.

These comments apply much more so to horizontal flanges rather than top flanges. One thing to keep in mind - there is no requirement that the pump flanges have zero load - at ANY time. Even in a well supported system, there will be some amount of adjustment necessary to bolt up. This is why I target 1/4' for deflection.

What has been a problem, particularly for top suction/discharge nozzles, is that millwrights will try to "float" the piping flange, such that there is essentially no contact between the two flanges before bolt up. We seen cases where the field reports that the spring over the nozzle has "bottomed out" because they incorrectly pulled the travel stops and tried cranking on the spring to get the pipe flange to lift off the pump. It's pretty obvious when we get a call like this. Thankfully, it's been pretty rare of late.

PALDEX,

Pl. read Guidelines for the design and installation of pump piping system , WRC Bulletin no.:449.

Regards

PALDEX,

In addition to my previous post, this is what I try to follow:

Check the free end dipslacment in WNC case.Try to restraint it to 1/16 inch ( there is no basis for this number, but i have seen this no. followed in the comapanies that i have worked for and in my opinion a reasonable and justifiable number). Check for the allowable loading variation that is possible in the springs ( if any). Typically around 20% is possible but depends on the spring manufacturer. Make sure that for larger lines ( typically more than 12") weights of all pipe support and spring hardware i included in your model. The argument is, with this condition, if the WNC tip deflection is kept within this limit, in reality when the installation will take place, due to diference in actual loads w.r.t the ones used in the analysis, there will be requirement at site, to alter the status of the pipe supports a little bit ( which may include changing the spring loads as mentioned before) to result in proper alignment. With these changed loadings, generally there is no issue in the operating condition also , as you are actually imposing near zero load due to alignment , if you have followed the above procedure.

Hope this helps.

Hi,
Normally during rotary equipment installs the connecting pipework should be in its right possision for alignment checks and so.
The method I use is as it is mentioned above, change the connecting CNODE ANC stiffness to zero and then if the deflection is too much, add a temporary support for installation purposes. Remeber you should add a note for this kind of support to clear its role, as it might remain after installation causing problems.
Hope it helps

Hi Anindya

Thanks for your valuable information.
I like to clarify the following statement

*** Make sure that for larger lines ( typically more than 12") weights of all pipe support and spring hardware i included in your model***

I understand you are talking only for primary pipe support ( Shoe & trunion) weight to include in Caesar model for more accurate load.

For spring hardware I think we should not include it in piping model because that weight is taken care of by external structure ( Spring hanger ) or ground base ( Cane spring).

I like to consider spring as force only in piping system ( Ignoring very high level of accuracy).
I want your advice on my understanding.

Regards
Habib

Regarding the displacement at the free flange end, we'd better follow 335.1 (c) in B31.3 saying 3mm (1/8") max offset and 1mm in 200mm (1/16 in/ft).

Thanks,