The Pitting Started Where the Weld Map Did Not Show a Weld
Quote from chief_editor on June 4, 2026, 3:00 amCorrosion failures in Chinese stainless steel and alloy equipment frequently initiate at weld locations that do not appear on the as-built weld maps. The discrepancy is the failure cause.
A chemical plant in South Korea installed a Chinese-manufactured stainless steel reactor vessel in 2018. The vessel was fabricated from 316L stainless steel, ASME Section VIII Division 1 compliant, U-stamp certified by a third-party inspection body. After thirty months of operation handling a mildly corrosive process fluid, localized corrosion pitting appeared at three locations on the vessel shell.
The metallurgical investigation of the corrosion sites established that the pitting had initiated at weld heat-affected zones. The unusual finding: the as-built weld map for the vessel showed no welds at two of the three corrosion initiation sites. The heat-affected zone microstructure visible in the metallographic sections indicated that a weld had been made and subsequently ground flush, removing the weld bead but not the heat-affected zone in the base metal.
The fabricator had made repair welds at these locations during fabrication -- probably to address surface defects in the plate that were discovered during assembly -- and had ground them flush without documenting them in the weld record. The heat-affected zone sensitization that resulted from the weld thermal cycle created a metallurgical condition that accelerated corrosion in the process environment.
Why Undisclosed Welds Are More Common Than Buyers Assume
Stainless steel fabrication involves a production environment where plate surface defects, dimensional adjustments, and assembly fit-up issues frequently require localized repair welding during the fabrication process. These repair welds are a normal part of fabrication. What distinguishes professional fabrication practice from problematic fabrication practice is whether the repair welds are documented, inspected, and heat-treated in accordance with the same procedures that govern the primary welds.
In Chinese pressure vessel fabrication -- including fabrication under ASME U-stamp certification -- the documentation of repair welds is sometimes incomplete. The repair weld itself may be performed by a qualified welder using an approved procedure, satisfying the mechanical requirement. The failure to document the repair weld in the weld record creates a gap in the as-built condition record that has two consequences: the NDT coverage plan may not include the repair weld location (because it was not known to exist), and the heat-affected zone condition at the repair weld is not evaluated in the quality record.
For austenitic stainless steels, the heat-affected zone sensitization created by welding -- where chromium carbides precipitate at grain boundaries, reducing the local chromium content available for corrosion resistance -- is the mechanism through which weld locations become corrosion initiation sites. Solution annealing after welding can restore the corrosion resistance. In a complete weld record, the post-weld heat treatment of all welds is documented. In a record with undisclosed repair welds, the solution annealing of those welds is unverified.
The Inspection Protocol That Surfaces Undisclosed Welds
The most effective detection mechanism for undisclosed repair welds in stainless steel fabrication is ferrite measurement at the completion of fabrication, before the vessel is internally pickled and passivated. The delta ferrite content of austenitic stainless steel welds and heat-affected zones differs measurably from the base metal. A ferrite measurement survey of the vessel surface, conducted with a portable ferritescope, will reveal weld-affected zones as areas of elevated ferrite reading even where the surface has been ground flush.
This inspection takes approximately four hours for a standard pressure vessel and requires a ferritescope -- an instrument that costs approximately USD 2,000 to purchase or can be hired through inspection firms. The ferritescope survey should be conducted by the buyer's third-party inspector before the vessel surface treatment is applied, because pickling and passivation do not affect the ferrite reading but may affect surface accessibility.
A ferritescope survey that identifies locations of elevated ferrite reading not shown on the weld map is evidence of undisclosed welds. Each such location should be investigated: is there a documented repair weld record, was the weld performed by a qualified welder to an approved procedure, and was post-weld heat treatment performed and documented? A location that cannot be answered affirmatively requires treatment as a new weld -- weld procedure qualification, welder qualification, NDT, and heat treatment documentation.
The South Korean reactor vessel's corrosion problem was not a material failure. It was a documentation failure that created an uninspected metallurgical condition in a corrosion-critical application. The ferritescope survey that would have identified the repair weld locations before delivery was not specified in the inspection plan. Including it in future inspection plans for stainless steel equipment from Chinese fabricators costs four hours of inspector time. The cost of the alternative has been demonstrated.
Corrosion failures in Chinese stainless steel and alloy equipment frequently initiate at weld locations that do not appear on the as-built weld maps. The discrepancy is the failure cause.
A chemical plant in South Korea installed a Chinese-manufactured stainless steel reactor vessel in 2018. The vessel was fabricated from 316L stainless steel, ASME Section VIII Division 1 compliant, U-stamp certified by a third-party inspection body. After thirty months of operation handling a mildly corrosive process fluid, localized corrosion pitting appeared at three locations on the vessel shell.
The metallurgical investigation of the corrosion sites established that the pitting had initiated at weld heat-affected zones. The unusual finding: the as-built weld map for the vessel showed no welds at two of the three corrosion initiation sites. The heat-affected zone microstructure visible in the metallographic sections indicated that a weld had been made and subsequently ground flush, removing the weld bead but not the heat-affected zone in the base metal.
The fabricator had made repair welds at these locations during fabrication -- probably to address surface defects in the plate that were discovered during assembly -- and had ground them flush without documenting them in the weld record. The heat-affected zone sensitization that resulted from the weld thermal cycle created a metallurgical condition that accelerated corrosion in the process environment.
Why Undisclosed Welds Are More Common Than Buyers Assume
Stainless steel fabrication involves a production environment where plate surface defects, dimensional adjustments, and assembly fit-up issues frequently require localized repair welding during the fabrication process. These repair welds are a normal part of fabrication. What distinguishes professional fabrication practice from problematic fabrication practice is whether the repair welds are documented, inspected, and heat-treated in accordance with the same procedures that govern the primary welds.
In Chinese pressure vessel fabrication -- including fabrication under ASME U-stamp certification -- the documentation of repair welds is sometimes incomplete. The repair weld itself may be performed by a qualified welder using an approved procedure, satisfying the mechanical requirement. The failure to document the repair weld in the weld record creates a gap in the as-built condition record that has two consequences: the NDT coverage plan may not include the repair weld location (because it was not known to exist), and the heat-affected zone condition at the repair weld is not evaluated in the quality record.
For austenitic stainless steels, the heat-affected zone sensitization created by welding -- where chromium carbides precipitate at grain boundaries, reducing the local chromium content available for corrosion resistance -- is the mechanism through which weld locations become corrosion initiation sites. Solution annealing after welding can restore the corrosion resistance. In a complete weld record, the post-weld heat treatment of all welds is documented. In a record with undisclosed repair welds, the solution annealing of those welds is unverified.
The Inspection Protocol That Surfaces Undisclosed Welds
The most effective detection mechanism for undisclosed repair welds in stainless steel fabrication is ferrite measurement at the completion of fabrication, before the vessel is internally pickled and passivated. The delta ferrite content of austenitic stainless steel welds and heat-affected zones differs measurably from the base metal. A ferrite measurement survey of the vessel surface, conducted with a portable ferritescope, will reveal weld-affected zones as areas of elevated ferrite reading even where the surface has been ground flush.
This inspection takes approximately four hours for a standard pressure vessel and requires a ferritescope -- an instrument that costs approximately USD 2,000 to purchase or can be hired through inspection firms. The ferritescope survey should be conducted by the buyer's third-party inspector before the vessel surface treatment is applied, because pickling and passivation do not affect the ferrite reading but may affect surface accessibility.
A ferritescope survey that identifies locations of elevated ferrite reading not shown on the weld map is evidence of undisclosed welds. Each such location should be investigated: is there a documented repair weld record, was the weld performed by a qualified welder to an approved procedure, and was post-weld heat treatment performed and documented? A location that cannot be answered affirmatively requires treatment as a new weld -- weld procedure qualification, welder qualification, NDT, and heat treatment documentation.
The South Korean reactor vessel's corrosion problem was not a material failure. It was a documentation failure that created an uninspected metallurgical condition in a corrosion-critical application. The ferritescope survey that would have identified the repair weld locations before delivery was not specified in the inspection plan. Including it in future inspection plans for stainless steel equipment from Chinese fabricators costs four hours of inspector time. The cost of the alternative has been demonstrated.