Hi sir;
i was examining a existing spool going from reboiler to stripper. design values are as:
nomial dia= 8"
design temp=253F
design pressure=12psig
this spool is failing in operation and expansion case due to thermnal expansion i checked WRC of nozzle and nozzles also failed. Yet this line is working successfully for last 1 year. It made me surprized. Now i am thinking that i am making mistake in modeling.
This spool is easily passing in suspension case. In my opinion high values of stress and forces at restraint are due to exapnsion in element 90 to 100 which is 8" nomial dia and 14 feet long . according to ASME at 253 F it will expand upto .21 in.
can any body guide me through it

Hello,

You are not really showing us enough of the "system" for us to provide meaningful comments. We need also to see the node numbers. You have two "anchors" pushing against each other with resulting high "X" loads. The system is "over-restrained". But there is not enough information to see where your "anchors" (or other restraints) are located.

Regards, John

i am extreemly sorry for being so late. I got ill and couldnot come on board for three days and now when i saw first file attached with my post . It was not the file which i wanted to post. SO please discard first file i am going to attach images again. I Hope these two images will clearify my question.

Hi Sami
You can not model reboiler nozzles as absolute anchor.
You have to attached reboiler ( may be on spring support or attach with equipment) with this piping model to take flexibility of that two nozzles.

Hi habib.
thnx for advice.
so should i model reboiler nozzle in caeser ii. two nozzles on reboiler are present on same horizontal shell. how can i use spring support.
further more my question is that this model is working in field for last 1 and half year. there is no spring neither any special support. nozzles are direclty attached to main pipe of 8in via flange and tee. calculation shows that it is failing both nozzle as equimpment and piping from tee.
but in field it is working. Is this the case due to factor of safety?

Hi sami

Just model the horizontal shell & support of shell as it is at site.
I understand you have almost same temperature at shell & in line.So it will pass the system.
Your present piping model is failing because expansion of piping between two nozzle is blocked by anchor support.

hi habib from your answer i am understanding that i should model shell in two nozzles in caeser ii. shell is made of 3/8" thick plate rolled to 60" innder dia.
how can i model it in caeser.
i think i am ignoring shell linear growth. I am going to calculate it and then give displacement on anchor May be this is the mistake i am making?
as both of shell and pipe are of carbon steel so perhape there will not be much difference.
I ill post again when i have checked this possiblity.

yup that was the case i was ignoring expansion in vessel shell. Now when i calculated expansion in shell and give values as initial displacement in place of anchor at reboiler nozzles, my system is easily passing and forces are in range of 4 to 5 thousands as compared to 186000 lbs in earlier case.

Your on the right track. While, as stress engineers, we are concerned with the design of the piping system, we need to take into account the effect of attached equipment - especially thermal expansion. Towers, tanks, drums, exchangers all made of metal grow just like the pipe does.


One other bit I'll leave you with - you noted several times that the system is working in the field after one year of service. That's probably not long enough to know. Remember that the expansion stress limits are based on how many cycles the system can survive going from install to operating temperature and back again. Typical B31.3 basis is 7000 cycles (which is modified by changing the f value accordingly for different cycle life)

So, your piping system can be well over the allowable expansion stress values, and still survive a few cycles before failing. If the unit is one that is intended to come up and run at a fairly steady condition, you may just not have hit the (low) cycle limit that your "overstressed" system has.

In conclusion, when someone tells you it "works fine in the field" take it with some skepticism as that doesn't mean you have a code compliant design.