Abstract
A stochastic model is developed to assess the susceptibility of series 300 stainless steels to stress corrosion cracking under sour service conditions. The model is based on the hypothesis that the critical conditions for crack initiation and propagation are satisfied in the metastable pit regime found in H2S-containing environments.
The model determines the probability of cracking by forecasting the most likely elapsed time to achieve sustained crack growth. The boundary conditions for the stochastic model are provided by experimental evaluation of the pit survival rate using a weak potentiostatic polarization up to 100 mV above the open circuit potential.
To validate the model, service conditions similar to actual field cases of stress corrosion cracking were simulated experimentally. The model forecasts were in agreement with typical field times to failure within 95% certainty in all of the evaluated cases.