Preferential corrosion of carbon steel elbow
Preferential corrosion of carbon steel elbow
(OP)
I am examining a carbon steel 6" ASTM A234 WCB standard wall elbow that failed on a gland water line (contains slightly alkaline condensate at approx 85°C.). We found the problem when a fine jet of water was spraying out of the elbow.
It appears to have failed from the outside due possibly to a nearby sulphuric acid leak. I can see some metal loss generally on the outsdie of the elbow and a little bit of preferential corrosion at the butt welds, but no major wall thickness loss. However the failure appears to be from deep fissuring in two areas on opposite sides of the elbow. See photos. It seems to have preferentially gone for these two areas.
I know the external acid corrosion is a little unusual, but has anyone seen prefential attack on an elbow before? I'm wondering whether it has selectively attacked because of a feature of the forging process, forging defect or perhaps a function of the stresses in the pressurised fitting.
I would welcome any views / experience you may have.
It appears to have failed from the outside due possibly to a nearby sulphuric acid leak. I can see some metal loss generally on the outsdie of the elbow and a little bit of preferential corrosion at the butt welds, but no major wall thickness loss. However the failure appears to be from deep fissuring in two areas on opposite sides of the elbow. See photos. It seems to have preferentially gone for these two areas.
I know the external acid corrosion is a little unusual, but has anyone seen prefential attack on an elbow before? I'm wondering whether it has selectively attacked because of a feature of the forging process, forging defect or perhaps a function of the stresses in the pressurised fitting.
I would welcome any views / experience you may have.





RE: Preferential corrosion of carbon steel elbow
Unless I am missing something, I see no real evidence of local external wastage related to corrosion attack where the wall would be thinned and failure from stress rupture in service. If the leaks occurred because of the cracks shown in the attached photograph, there could be two possibilities;
1. The elbow contained defects from fabrication that went undetected prior to service.
2. There could be corrosion fatigue or even caustic SCC cracking from the inside of the elbow and what you are observing is a through-wall crack that initiated and propagated from the ID surface.
I can't provide any more information. If you remove this elbow, look on the ID surface for clues.
RE: Preferential corrosion of carbon steel elbow
RE: Preferential corrosion of carbon steel elbow
RE: Preferential corrosion of carbon steel elbow
Transgranular or intergranular would make a difference of SCC or something else.
Look for oxide on the fracture surface, which may point mfg problems.
I have seen a failure like this due to thermal fatigue, but you don't seem to have the conditions for that.
All the more reason to due a full failure analysis.
RE: Preferential corrosion of carbon steel elbow
Corrosion in Caustic Solutions
"Low concentrations of caustic can be safely handled by carbon steel up to 180oF/82oC, where CSCC starts to become a risk factor, while the safe upper limit for a 50% solution is approximately 150oF/65oC, although cracking has occurred at temperatures as low as 120oF/48oC."
http://www.hghouston.com/naoh.html
http://www.hghouston.com/naoh_gph.html
S
Corrosion Prevention & Corrosion Control
RE: Preferential corrosion of carbon steel elbow
I'm not sure about SCC as a mechanism. If it were SCC, wouldn't the welds be prime sites for attack due to residual stresses? In fact where are the stresses from to cause cracking at the sides of the elbow? Also the caustic concentration would be very low, approx 0.5gpl.
I will get a detailed metallographic examination done on this and post the result when I get it.
RE: Preferential corrosion of carbon steel elbow
Cracking occurs along the neutral axis of bend. This is not atypical of cracks that I have seen from ID initiated corrosion fatigue or stress corrosion cracking. I absolutely agree a metallurgical analysis (chemistry as well a metallographic analysis) is required, not just a metallographic examination.
RE: Preferential corrosion of carbon steel elbow
By the photos and service description maybe one can think about caustic embrittlement. let us know about your metallographic examination.
luis