Flange leak problem observing after regeneration cycle of ethylene dryer

[PP007]

In our plant there is Ethylene dryer to remove the moisture & after bed exhaust it required to do regeneration..High temp N2 pass through it & vessel & piping under go from atm temp to 330 degree centigrade temp.

After regeneration several times found that one or two flange get leak while taking the service in line.

Hot bolting is method but there is high risk. is there any way which can solve the problem leaking after regeneration.

Note we are using correct bolt , Correct Graphite filled gasket & proper tightening also .. still facing problem

awaiting for swift response.

Regards

 

[EdStainless]

I presume that you have gone back and double checked the loading and alignment due to thermal expansion.
And you have verified the materials in the piping, flanges, and bolting.
I suspect that you have excessive loading on these flanges.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[Tmoose]

"Hot bolting is method but there is high risk."
What is the "high risk?"


Detail drawings and Pictures of the piping arrangement including the first 4 supports in each direction of the leaking flanges would be a good first step.

Was fastener torque checked after the leaks developed ?
( I know t*rque is a dirty word around these parts. Maybe complete sets of bolts for all the leaking flanges, and several of the non-leaking flanges could be machined/center drilled on each end for checking stretched length at some various temperatures.

http://photos.motoiq.com/MotoIQ/Project-Cars/Project-Honda-Civic/i-SVsTkCS/0/L/IMG1980-L.jpg

)

As others mentioned, pipe strain should be investigated.
A starting point is how even the gap between flanges is prior to bolt tightening.
Also flange offset.
Also analysis of the piping system showing loads and how thermal growth is expected, and accommodated. Sliding supports, loops, offsets, maybe even expansion joints.

 

[PP007]

Thanks Tmoose.

Hot Bolting considering with risk as working required with live line during hot bolting

This flange is 10" 300# so we are using manual tightening technique not as per torque.

After leakage flange torque not checked but flange tighten again & leak was arrested.

 

[SEP87]

How are you validating they have 'proper tightening' when you are using 'manual tightening technique not as per torque'?

I would advise you tighten them with appropriate torque rather than manual tightening. Whilst this is sufficient for many applications if you have a problem site then torques are the first place to look. A deviation in torques around a flanged joint is inevitable if doing it manually. Access and angles on the tools will make a massive difference, you might think they're all being tightened equally but they're not. Even just ruling out torque issues would be a start.

 

[XL83NL]

I suspect that you have excessive loading on these flanges.

My thoughts as well. Either get a bolting consultant to determine the root cause, or (jump ahead and) do a pipe stress ansylsis to find out the loadings on the flanged joint, and check if this flange should (theoreticlally) be able to handle these loads.

 

[Tmoose]

"Note we are using correct bolt , Correct Graphite filled gasket & proper tightening also"
Please provide details (pictures and drawings) of the flange design.

Is this equipment designed by others?
If so, what is the OEM's response to this leakage?

Generally speaking, "correctly" tightened fasteners don't loosen in service.
Several "correct" components don't always add up to a "correct" system without some product development.
I'd measure stuff before and after installation and leakage, expecting to find signs of gasket compression.

Stuff = torque, bolt stretch, distance across flange faces

Maybe after a few leak/tighten cycles the gasket will stabilize and leakage will stop. What that tells me is, the "correct" tightening or gasket material are in fact NOT correct.

 

[gerhardl]

All above is about flanges, bolts and correct thightening, there is only one reasonable inquiry about alignment.

A valve (endflanges?) leaking is most often caused by to weak or missing support for the pipeline , and/or for the valve itself. Also better quality and type for sealing material could improve the situaton.

My guess is missing support for the valve or pipeline,causing misalignment and/or skew forces under the heating cycle.

 

[TGS4]

This is not uncommon. I even wrote a paper about solving one such problem.

There are many potential factors that go into an investigation of such a leakage situation. I highly recommend that you secure the services of an expert in the topic.