The Gamma Exponent Model ("GEM") is a recently developed methodology for failure pressure prediction of flaws in oil and gas pipelines. The model has already shown varied applicability for different flaw types.
The focus of this work is axial metal loss flaws and axial stress corrosion cracking (SCC). The basic model is applicable to axial metal loss. The model is then modified to address issues associated with SCC. Specifically, this work demonstrates that a stress intensity factor relevant to periodic crack arrays, as observed in SCC colonies, can be used to account for crack-shielding. This results in more accurate failure predictions.
The model is validated against available laboratory, hydrotest and in-service failure data. Statistical analyses are performed to demonstrate the mathematical form of the new model are valid. Failure pressure predictions are shown to be comparable or better than current industry models.
Failure pressure calculations are an important part of all oil and gas pipeline operators' integrity management programs. Analysts must be able to predict the maximum pressure a pipeline can sustain in service and apply a suitable safety factor to account for uncertainties. Pipelines in service will have known (or unknown) distributions of flaws that must be included in these failure pressure predictions. Flaws may be crack-like, such as long seam or stress corrosion cracks (SCC), metal loss, such as corrosion or gouges, or deformation, such as dents or ovality. The focus of this work is axial corrosion and SCC crack-like flaws.
The failure pressure predictions, and their reliability, have a direct impact on the effectiveness and efficiency of the operators' integrity management programs. If failure pressure predictions are non-conservative, the operator may excavate and repair fewer features than are necessary to maintain pipeline integrity and this could result in an ineffective and unsafe integrity program. However, if failure pressure predictions are overly-conservative, the operator may excavate and repair more features than are necessary and this could result in an inefficient and financially costly integrity program.
