Improper Material/Tube Replacement

A 40 year-old boiler superheater tube was received for failure analysis, material verification and dissimilar-metal weld (DMW) analysis. The tube was specified as 2.0” OD x 0.438” MWT. The failure was just below a safe weld which was magnetic and verified (in the field) as T22 steel. However, this failure location was supposed to be stainless steel. A previous report from the same area indicated no signs of creep.   

All the sections were magnetic except the section between the DMW and the immediate butt weld, indicating it is stainless steel. OES testing was performed (labeled MV); the tube met the chemical requirements for ASTM SA-213 T11 Cr-Mo steel. The fracture was fish-mouthed in appearance and thick-edged, typical for a creep failure. Several longitudinal cracks were observed on the tube OD due to tube swelling. No hard phase was observed on the fracture face which precludes the possibility of short-term overheating. Ring samples were sectioned from the tube for dimensional and hardness measurements; the results (0.274-0.348”) were not consistent with 0.438” MWT. 


Poor fit-up was observed between the stainless steel and T11. The stainless steel tube should have been swaged to match the ID surface. Flow disruption may have occurred at this transition and caused a localized hot spot.

       Creep voids were observed along the carbon steel fracture face. This area was supposed to be stainless.                 

The tube failed due to the use of inferior material. Lower-grade 1½Cr- ½Mo T11 steel was used, resulting in severe creep damage. The allowable stresses and creep resistance are significantly lower for T11 material compared to TP 347H stainless. Poor fit-up may have caused flow disruption resulting in a localized hot spot. Therefore, the poor fit-up may have also contributed to the failure. The tube operating temperature was estimated at 1073 oF. This could be an underestimation due to broken oxide scale. The MWT is not consistent with the specified 0.438”; a relatively thinner wall was measured. The microstructure had transformed to spheroidized carbides in a ferrite matrix. Creep degradation was also observed away from the failure site. 

The DMW appears to be in poor condition. Severe creep damage was observed in the T11 side of the HAZ. Welding defects in the form of poor weld geometry and poor fit-up were observed. Fatigue cracks were observed in the HAZ at both the OD and ID surfaces.