This is regarding Reactor - Heater system. The Reactor height is almost 120m from grade & is subjected to a sway displacement value of 1/200. Also, there is a structure around the reactor, which is subjected to a sway displacement value of 1/300. I have a 42" Process line @ 549ºC coming out of reactor at El-66.5 mtrs from anchor, drops 63m vertically down & with only 1 support on horizontal run goes directly into Heater's radiation section.
Piping is supported using the same structure. Also, Reactor is guided using the structure at 40m from anchor.

My question is, what will be the actual displacement values for the reactor & apart for reactor guide, at what other locations displacments needs to entered.
This system fails in occassional stresses at the 1st elbow near reactor nozzle. Note, as per the client, there is no scope for change in layout or adding flexibility near nozzle.

I have considered following load cases for the system, is it correct?

L1 W HGR
L2 W+D1+T1+P1 HGR
L3 WW+HP+H HYD
L4 W+T1+P1+H OPE
L5 W+T2+P1+H OPE
L6 W+T3+P1+H OPE
L7 W+T1+P1+H OPE
L8 W+T2+P1+H OPE
L9 W+T3+P1+H OPE
L10 W+T1+P1+H+D1 OPE
L11 W+T1+P1+H-D1 OPE
L12 W+T1+P1+H+D2 OPE
L13 W+T1+P1+H-D2 OPE
L14 W+T2+P1+H+D1 OPE
L15 W+T2+P1+H-D1 OPE
L16 W+T2+P1+H+D2 OPE
L17 W+T2+P1+H-D2 OPE
L18 W+T3+P1+H+D1 OPE
L19 W+T3+P1+H-D1 OPE
L20 W+T3+P1+H+D2 OPE
L21 W+T3+P1+H-D2 OPE
L22 W+T2+P1+H+D1+WIN1 OPE
L23 W+T2+P1+H-D1+WIN2 OPE
L24 W+T2+P1+H+D2+WIN3 OPE
L25 W+T2+P1+H-D2+WIN4 OPE
L26 W+P1+H SUS
L27 WNC+H SUS
L28 T1 EXP
L29 T2 EXP
L30 T3 EXP
L31 L4-L26 EXP ALG
L32 L5-L26 EXP ALG
L33 L6-L26 EXP ALG
L34 L4-L28 SUS ALG
L35 L5-L29 SUS ALG
L36 L6-L30 SUS ALG
L37 L22-L5 OCC ALG
L38 L23-L5 OCC ALG
L39 L24-L5 OCC ALG
L40 L25-L5 OCC ALG
L41 L37+L34 OCC SCALAR
L42 L38+L34 OCC SCALAR
L43 L39+L34 OCC SCALAR
L44 L40+L34 OCC SCALAR
L45 L26,L27,L34,L35,L36 SUS MAX
L46 L4,L5,L6,L7,L8,L9 OPE MAX
L47 L10,L11,L12,L13,L14,L15,L16,L17,L18,L19,L20,L21 OPE MAX
L48 L22,L23,L24,L25 OPE MAX



here D1 = Sway in X direction, D2 = Sway in Z direction.
Win1 =+X dir, Win2 =-X dir, Win3=+Z dir, Win4 =-Z dir

please help with the displacement values for sway.
Also, provide me email add where i can send my CII file for your ready reference & easy understanding.

Thanks, Good day.

Here is an idea:
Model the reactor with the piping and supports. Whatever (wind, seismic etc...) gives that sway needs to be modeled (added to the model as action) with the stiffness of the reactor in addition to the thermal loads on the reactor if available. I am not going into details of connectivity between pipe and the reactor here, you can figure it out.

There might be other ways making the model to achieve the results, but I can claim that this is a lot more simple and reliable.

Ibrahim,

Thanks for your guidance, however my question regarding the methodology to use sway effect on reactor, piping & its supports and also the displacement values that needs to be considered for analysis at equipment side & at piping side remains unanswered.

Best regards

If fails in occasional stress then it's not properly supported I supposed, check your first support on vertical. Even guide on the reactor may help on the occasional stresses. There is no need to provide flexibility, stiffen the system instead if your problem is wind load.

~clyde

Without a proper drawing it is difficult to visualize the entire system. In general terms let us assume the tower is on a structure, and pipe is supported by the tower at tower levels and below by the structure. I hope you do not use different structure to guide/support the piping.

In this case the structure is going to be under compression and shortened under the dead load. The tower is under the thermal loads, so as the piping. The entire system is under the wind and seismic loads which, I assume, gives the sway at the top of the tower.

Unless you do a proper FEA you can never get proper displacements of the lower levels without modeling the structure, the tower and the piping all together.

So modelling them all together is the best solution since the pipe will be under thermal displacement (perhaps downward) while the tower is under thermal displacement upward. Thus is tricky part of the analysis. If you do not provide adequate model and your calculation will have some discrepancy to simulate the support locations under thermal loads and you may end up with different answer in your calculation.

The model that I described above will solve the sway action at the top of the tower. Since the model include all the members attached together you do not need to calculate the displacements of the support locations on the tower due the the assigned flexibility/stiffness of the system within the model. With proper load cases Caesar II will do the rest for you.

I trust this answer your question if I understood correctly.

Your reactor is guided by the structure then use the sway displacement of structure. Confirm your sway displacement since they have not the same value, your reactor has larger sway displacement (1/200) compared to your structure (1/300) and your reactor is guided using the same structure. If the reactor should have the same displacement with your structure and also your support then that may solve your problem.
You also have to check where are the connection point of your support point on your drawing then interpret it by modeling in CII.

You calculated the occasional stress using the following I suppose,

L41 L37+L34 OCC SCALAR
L42 L38+L34 OCC SCALAR
L43 L39+L34 OCC SCALAR
L44 L40+L34 OCC SCALAR

I noticed you used the hot sustained stress, it may have a higher value compared to cold sustained stress. Better you have to use the cold sustained stress (L26) on this.

Thank you Clyde & Ibrahim for your guidance.

Clyde - Your understanding reg same displacement on reactor, & structure will give no differential displacement between two, hence all 3 elements, viz. reactor, structure & piping resting on structure will move with same displacement. However, in actual structure & reactor are swaying at the rate of L/300 & L/200 resp.
Also, reactor is guided from the structure at the center. Hence, there is a differential sway displacement that needs to be considered.

Secondly, client demands to check the stresses using HOT Sustained, hence have no options other than that.

I have attached the snapshots of CII file & SPR model for your better understanding. Hope this gives the clear idea bout the system arrangement, which may further help to provide a better solution.

Thanks, Regards

The reactor (I called it the reactor) might have different stiffness along the height due to the differential diameters and wall thicknesses. Additionally the guide on the reactor is half way as you said and provided by the structure. Therefore the displacements at lower levels cannot be easily estimated while the top of the reactor has the sway. The best is to model the reactor, and the structure if possible in addition to the piping. The structure stiffness is another contributor at the guide level of the reactor. No easy way out.

However this type model will simplify your findings, and you will not need to take data from one (FEA) model to another (Caesar II model) for the displacements. So you can easily see the entire system thermal, seismic behavior and make adequate decisions in case a problem exist.

Unfortunately, sometimes the best is to model everything in details.

Selection of location for the line stopper on the vertical piping is very important for the nozzle analysis. If you can find the thermally neutral (the reactor at nozzle elevation will rise with the vertical pipe at the same rate - I suggest again to model the reactor and the vertical pipe, and leave the thermal expansion calculation to Caesar II because small deviation in the expansion coefficients may cause problem with hand calculation input) location for the line stopper on the vertical pipe the SUSTAIN load effect on the nozzle will be reduced substantially. So you are able to manage the seismic loads better.

I suggest you to extend the pipe model to heater radiation section with thermal displacement of this section or by modeling the tubes entirely in this section. Sometimes this type large pipes can create big problem at the tube header connections and you cannot see them without modeling.

Have you modeled the equipment? I'm not sure if you modeled it, but as Ibrahim said, the reactor should be modeled together with the piping, don't forget to support the reactor with guide in your model, that is considered a support point.

Regarding the displacement, ofcourse use the structure displacement at the guided point. At reactor point, calculate the displacement of a certain point of the reactor using its own stiffness value, then add to the previously calculated displacement at guide point. Make it sure to check your stiffness reference elevation for both reactor and structure. Provide a reasonable gaps to your guide support, since every elevation deflects at different rate. Use the structure stiffness to calculate the displacement of your support based on your description.

Thanks, Clyde & Ibrahim for your guidance.
Sorry for late response, as I was busy in other systems.

However, I will try with as what is suggested above & shall update you regarding the results.

Thanks, Good day.

I see that despite what B31.3 is indicating, everybody uses displacements due to occasional events as occasional loads. So pleeeeeeeeeeeeeeeeeease update the code.