Hi all,

I would have some clarification about manifold saddle modeling.
Heater manifold (in accordance with API 560) are designed with two saddles (one is fixed WITH BOLTS and the other is sliding); I'm focalizing my attention on the fixed saddles;

I'm asking what is the better way to model this saddle:

1) as anchor e.g. all 6 degree of freedom restrained

2) as x,y,z restraints e.g. with rotation allowed (no moment on the node)

I mean, is correct to consider that saddle as inifinitely rigid or not (the saddle is usually anchored at the convevtion beam with bolts)?
Have someone experience about this matter?

TY.

In our previous project we just simply model it as fixed anchor (6 DOF's restrained) and give the anchor load to Heater Vendor for them to check the structure to which the saddle will be bolted & the saddle/pipe connection stress check. This fixed saddle is the interface connection between EPC piping to Heater Vendor piping. If you want to buy in some flexibility you can model the saddle which I'm not sure is available in Caesar or obtain it's stiffness value. This is the usual dilemma in Heater Piping where the API 650 allowable is relatively low with respect to pipe size. So the best way is to establish your agreement with the vendor as early as possible. And in our case by submitting to them the Piping loads. But we have tried to make it also as low as possible by providing a flexible layout and some springs.

Sorry I mean API 560...

Hi Borzki,

Thanks for reply, do you remember the values of these reactions? (I mean not the exact values but an order of magnitude of the moments acting on the saddle).

Do you think that moments of 300 KNm can be absorbed by the structure?

TY.

We only have 6" pipe connecting to Heater. As far as I remember it's approximately 9 KN-m..300 KN-m seems big. Anyway what's the pipe size & temperature?

The problem is that I've a 24" (120 schedule) connected to the heater manifold and the design temperature is 550°C; the line is connected to a vertical vessel and I've not much space to loop the piping and make it more flexible.

The heater vendor gives me allowable loads and displacement at battery limit (e.g connection at manifold) and it has considered 5 times API 560 values.

My opinion is that I would never reach these allowable loads due to the dimension and the design temperature of the pipe so maybe I should model the manifold with the two saddles (one fixed and the other sliding)and give the reactions to the heater vendor.

Fixing the saddle (6 DOF's restrained) I still have important reactions (specially moments) at the fixed saddle and I'don't know if these values have been considered in the heater engineering (or they "just" apllied the 5xAPI at "nozzle" manifold (e.g. battery limit) in their structural calculation).

Hi, One quick check,

Is it possible to check the stiffness of the heater support? You can reduce your forces and moments to the considerable extend. Also try out by using EH instate of EC value and compaire the loads.

Regards,
R.K.