A time-delayed failure due to stress-activated creep (cold-creep) is a failure that occurs under a constant load and with no growth due corrosion, fatigue or some other environmentally assisted time-dependent degradation mechanism. A time-delayed failure is prevented by reducing the pressure.
ASME B31.4 and B31.8 recommend a 20 percent reduction, to 80 percent of the pressure at the time of damage or discovery. T/PM/P/11 Management Procedure for Inspection, assessment and repair of damaged (non-leaking) steel pipelines, an internal procedure used by National Grid, specifies a 15 percent reduction. The guidance in ASME B31.4 and B31.8, and in T/PM/P/11, is directly or indirectly based on the results of tests on the long term stability of defects conducted by the Battelle Memorial Institute and British Gas Corporation in the 1960s and 70s. The line pipe steels were Grades X52 or X60, and the full-size equivalent Charpy V-notch impact energy (where reported) did not exceed 35 J. The tests indicated that the threshold for a time-delayed failure was approximately 85–95% SAPF (straightaway-pressure-to-failure).
The strength and toughness of line pipe steels has significantly increased over the decades due to developments in steel-making and processing. The question then is whether an empirical threshold based on tests on lower strength and lower toughness steels is applicable to higher strength and higher toughness steels.
In the Tripartite Project, the Australian Pipelines and Gas Association (APGA), the European Pipeline Research Group (EPRG) and the Pipeline Research Council International (PRCI) collaborated in conducting full-scale six step-load-hold tests on higher strength and higher toughness steels. Companion papers present the other aspects of this multi-year project.
An empirical threshold for a time-delayed failure is estimated using the results of the six step-load-hold tests. That estimate is also informed by the other published small and full-scale tests (on lower strength and lower toughness steels). The Ductile Flaw Growth Model is used to infer the effect of strength and toughness on the threshold for a time-delayed failure.
A 15 percent pressure reduction, to 85 percent of the pressure at the time of damage (or of the maximum pressure that has occurred since the time of damage), is considered to be sufficient to prevent a time-delayed failure due to stress-activated creep in lower and higher toughness, in lower and higher strength, and in older and newer line pipe steels.