Q & A > Question Details
Recently we have a failure ( leak ) of RCSP (Regenerated catalyst stand pipe) bellow.
Following are the details of the bellow:
Bellow type : Two ply double elementpurgeless bellow with with pentographic linkage design.
Size: Element Length: 646mm(compressed), 7 convolutions each, MOC: Inconel 625LCF, pipe OD: 1982mm, bellow is providede with equalising & leak detection arrangement, packed with cerawool with wore wire mesh&silica cloth.
Operating temp: 343deg cent, (510 design) with axial expansion 210mm.
Leak detail: leak observed from 2nd crest of the top bellow element (inlet end) at approx 11 'o' clock position looking from outlet end. Opening (crack) was of 3.5" length
We are in the process of analysing the failure, kindly let us know various reasons for failure ?
 
Answers
18/09/2011 A: Steven Kalota, ECI, steve@fluidcatcracker.com
The 2 most common and probable causes of Expansion Joint failure are corrosion and cyclic stress. The packed joints are much more resistant to potential corrosion since there is much less exposure to moisture than with purged joints. Corrosion, if it occurs, is usually initiated during shutdowns when the standpipe is allowed to cool and there is residual sulfur and humidity present. Once initiated, corrosion mechanisms can be accelerated by the increased operating temperatures. This is far more prevalent in Flue Gas service than RCSP. Corrosion is also usually concentrated on bottom of bellows where there is a greater potential for low point moisture accumulation during cold, empty conditions.
Cyclic stress is the other probable failure mode. You have the correct metallurgy of Inconel 625 LCF (Low Cycle Fatigue) so the final life of the bellows would be impacted by the number of thermal cycles the joint has experienced. If your RCSP supports are satisfactory and you have the proper bellows equalizing rings, then failure potential should be minimal unless you have had many cold-to-hot-to-cold operating cycles.
Other than these causes, there is just poor quality fabrication, poor ceramic fiber packing installation, or loss/failure of the braided hose seal. Metallurgical failure is most unlikely since the leak involves failure of both plys.