In the late 1960s and early 1970s, the researchers of the NG-18 committee at the Battelle Institute in Columbus, Ohio completed a seminal study on the failure pressures of flaws in oil and gas pipelines. One of the key developments was the “log-secant” equation for the assessment of axial crack-like flaws. The model was later modified to improve its accuracy and precision.
The Gamma Exponent Model (GEM) was recently developed for assessment of axial crack-like flaws in pipelines. The developer recognized that the NG-18 log-secant model was theoretically derived on length, and then empirically corrected for depth. The new GEM was theoretically derived on depth, and then empirically corrected for length. The new model is similar in mathematical form to the original NG-18 log-secant model, but there are some key differences.
This work is a validation study of the GEM using axial crack failure pressure data from the industry literature. Laboratory tests with machined flaws, and hydrotest and in-service failures with natural metallurgical flaws, are also considered. The results of the GEM are compared to the equivalent failure predictions using other models. The strengths and limitations of the new model are discussed in the context of improved accuracy and precision for crack assessments.