Cracks which emanated from the interior parts of the plate could be due to Wet H2S cracking.I would like to know the crack pattern whether is it aligned or stepwise cracking. Carrying out H2 Bake out (Heating the weld to 400 Deg.C for 4 hrs)of weld before doing repair will solve cracks appearing again and again. The H2 bake out operation will drive out the entrapped H2 within the voids/inclusions. This problem is prodominantly high in high strength steel. please inform the base metal metallurgy for knowledge sake. regards,

Luis You have not mentioned the plate material used. If the plate material is equivalent to A 537 Cl 1 , then it is prone for this or older version of Japanese steel with Ni addition upto 1% - to C-Mn steel., these two steel materials used in propane / butane / LPG to reduce over-all tonnage in sphere construction / design, are known to develop sub surface cracks and also surface fissures due to HzS contaminant after some years. Delayed cracks - possibility of Sulphide stress cracks - Type 1 or 2 are known to occur in Propane / Butane / LPG spheres - with operational life of 5-35 years. The fact that you had experienced cracks after PWHT shows this could be a high tensile steel with still high residual stresses high enough to develop cracks - at sub surfaces . The fact the first PWHT had shown fresh cracks in a different and new area than what had been tested by MPT/TOFD, it appears to me that the material supplied had inherent hardness high enough to develop cracks after PWHT. Did you do a thorough PWHT hardness profile in areas where you had repair welded?. Did you do a PWHT locally or for the entire sphere- It is not clear from your questions raised? For sphere - which is operated at temp varying from slightly lower than ambient to design temp around max 50 deg cent - there is no need to do prior hydrogen bake out (suggested by Krish). Phenomenon of hydrogen bake out exercise does not arise for this service where temp of operation is less than 50 deg cent. Hydrogen bake out is usually done in hydrogen service vessels which are operated continuously at 300-450 deg cent with partial pressure of hydrogen likely to be quite high (100- 200 KSC at temp of operation around 300-500 deg cent). Here partial pressure is practically negligible with sphere design pressure / temp - around 10-15 KSC at 50 deg cent. In Propane or Butane or LPG service - moist LPG with mercaptans and some traces of wet H2S are known to develop fine surface fissures. Sub-surface cracks shown in new areas - is quite likely due to high hardness still in the weld and HAZ i.e, the original PWHT had not effectively proved to be a success. Mechanism of H2S - Type 1 or Type 2 is basically due to moist nascent hydrogen + assisted by sulphides to cause SSCC (Sulphide stress corrosion cracks) in plain carbon steels in this service. Intergranular crack is the predominant mode for SSCC. Linkages due to inclusions in steel (MnS, Alumino-Manganese, alumino silicon, Phosphides, different type of alloying inclusions) with H2S can assist in promotion but inclusions by themselves (arising due to un-clean steel making) may not contribute largely. Your apprehension some micro-voids could have contributed - to my mind- does not appear to be the basic reason. This is not an exotic grade material to look from micro-void aspect contributing to the failure. To me it appears the basic welding qualification at the fabrication stage and possibility of omission of high hardness both in WPS/PQR during stage wise testing during fabrication is the most likely cause ( which had not responded to PWHT well). I had occasion to study similar problem faced in LPG/Propane spheres - some years back - where sub surface cracks did appear after few years. Hope this helps you C.V.Srinivasan Nishi Engineers Pvt Ltd India March 25, 2008 E-mail: nishi@vsnl.com >After making 100% TOFD to the welds of a propane sphere >there were detected liner discontinuities such as localized >cracks in the welds link of the dome to shell buds at >several deep thickness (plates thickness 40mm). > >After this first inspection the defects were mechanically >removed and the remaining thicknesses were controlled by UT > >The crack repairs have been done through filler welding at a >controlled temperature and 100% UT control and magnetoscopy > >The sphere was then submitted to a PWHT at 620%C2ºC during 2 >hours following by slow cooling. > >After PWHT, the sphere was again submitted to 100% of UT and >magnetoscopy and there has been found new defects in zones >out of the repaired ones. > >These defects have also been removed and after being >repaired they originate new defects in adjacent zones. > >After metallographic replicas we found intergranular micro >cracks > >Now we are going to repair all the indications given by >TOFD, even the ones which can be accepted by the codes. > >We thing that the defects found have been induced by atomic >hydrogen which is now molecular hydrogen arrested inside the >micro voids, so we need to have the steel as much clean as >possible to avoid more defects induction. > >Please Share > > >Luis Marques >