A low alloy steel’s propensity for sulfide stress cracking in H2S frequently is evaluated based on
passing or failing NACE TM0177 Method-A and Method-D tests performed at a single (usually
maximum) value of material hardness. Here we explore a probabilistic assessment of cracking
tendency based on characterization of the mean Method-A threshold stress and the mean
Method-D stress intensity factor (KISSC) as continuous functions of hardness. For pressurized
pipe in H2S, the sensitivity to cracking failure is proposed to depend not only on the mean
threshold-vs.-hardness and the mean KISSC-vs.-hardness, but also significantly on multiple factors
which contribute to the scatter and standard deviation of cracking performance. The model
characterization of these factors enables assessment of the risk of failure due to a combination of
material, environment, and applied load. This concept is applied to the pressure performance
limit of pipe for cracking failure in H2S, and this is used to rank the importance of different
factors which contribute to the probability of pipe failure in a given H2S environment. This
approach identifies significant risk factors which usually are ignored in material testing at a
single hardness. The ranking of relevant factors can be used to guide testing. A case is made for
evaluating the risk of environmental cracking failure not at a single hardness point but rather as a
function of hardness varied over applicable ranges.
Keywords: sulfide stress cracking, SSC, DCB, pipe performance, probability, crack initiation,
crack propagation, KISSC, risk, hardness, scatter, NACE TM0177, Method-A, Method-D