Nozzle Loads & Gusseting

[Strook]

G'day guys,

Working with some significant nozzle loads (250 kN axial, 580 kN.m for DN900 into 1800OD vessel) and thinking through various options for strengthening the nozzle.

At this stage, reinforced set-in nozzle construction is yielding shell & comp plate thicknesses around 40mm, a little higher than I'm hoping to get away with.

External gussets are subject to thermal differential, leading to cracking of welds etc. Could this be counteracted simply by insulating the vessel? Vessel operating at ambient to 115degC.

The other suggestion I have is nozzle internal projection providing real estate for an internal gusseting arrangement in lieu of the typical external gusseting. I suspect such an arrangement wouldn't be as likely subject to thermal differential, and would provide direct connection to shell rather than to comp plate. Obviously this wouldn't help the nozzle neck much but conceivably would reduce shell & comp plate thicknesses.

Thoughts on the principle of such a design are welcome, please.

EDIT: The most obscure aspect - existing vessel operating at -8 kPa(g) / +37 kPa(g), 1800 OD and 15m tall with skirt. Extremely low pressure equipment but somehow (?!) piping loads onto this plant are astronomical. I've suggested the piping design is revised to reduce these loads, but in the mean time thinking through worst case options.

 

[MJCronin]

Strook,

Astronomical Nozzle loads strongly implies a newbie piping stress analyst who simply uses the default stiffness of CAESAR-II or similar piping program.... In my opinion.... This issue has been around since at least the 1960s....

This ancient problem has always been most acute in larger diameter stainless nozzles (>8")in low design pressure, vertical axis vessels

Newly minted MBA project managers frequently assign the most inexperienced to be the piping stress analyst and, of course, could care less about who this hurts in the long run.

Before you gusset all of your nozzles and install giant anchor bolts, see if this problem could be dealt with via a sharper pencil. And by sharper pencil, I mean a re analysis of the piping with more realistic nozzle stiffnesses

There are computer programs available to calculate this more realistic nozzle stiffnesses and incorporate some of the Knowledge of the past (WRC 297, PD 5500 comes to mind, but I need to take my medication...). CAESAR-II analysis models can easily develop or accept manual calculation of these more realistic nozzle stiffnesses.

Under the designation "Anchors" please see "Flexible Nozzles"

https://docs.hexagonppm.com/r/en-US/CAESAR-II-Applications-Guide/Version-12/329153

Also, search eng-tips ... this topic has been discussed in the past

https://www.ijedr.org/papers/IJEDR1603035.pdf

Oh, and BTW, ....I also believe that the only time a nozzle should be gusseted in a metallic tank or vessel is when it is recommended by the vendor of a mixer/agitator and mounted in the center head of the vessel. Gussets supply stiffness for the significant mechanical loadings and the nozzle should also incorporate a repad in order to better engage the vessel head. Prevention of mixer sway while operating is very important

Keep us in the loop regarding your final solution

Regards



MJCronin
Sr. Process Engineer

 

[Strook]

MJCronin,

I share your opinion re: newbie piping stress analyst. I'm already pushing for a revision of the loads to be developed with more accuracy and expertise.

I'm aware of the stigma around gusseting nozzles and how it should be restricted to vendor-specified agitator nozzle. I have no problem making this generalisation to prevent newbies making mistakes. But I want to understand from other experienced engineers' perspectives what are the reason/assumptions behind this restriction?

Is it that thermal differential stress will inevitably cause weld cracking? If so, how do insulated nozzles or internal gussets challenge this?

Is it because gussets change the stiffness of the nozzle and therefore change the piping loads? If so, why can't piping loads be developed iteratively with nozzle stiffness as gusset changes are made, for a one-off design? Obviously this becomes uneconomical on small vessels but for large, I see potential value.

Is it because of another reason I haven't thought of? Possibly.

Let's discuss. There must be a practical way to strengthen nozzles in addition to "thicker shell, larger comp plate, larger nozzle, different material."

Strook

 

[saplanti]

You mentioned that there is an existing vertical vessel. Did you check how the piping connected onto it, and how the attached piping supported along the way?

I do not know how you got the job. Did you have specifications for the vessel and as well as the piping?
Some companies distribute the design and manufacture jobs with their own specifications. In some cases, such as yours, the nozzle loads become stupid. In this case you both parties and the client come together to discuss for the workable solution.

You did not mention about the fluid in the piping. Some fluids can allow you to use metal expansion joint in case there is no alternative.