PRCI Report 205

L51683e

Battelle Memorial Institute

Need: The great majority of instances of stress corrosion cracking in high pressure gas pipelines are associated with the propagation of intergranular cracks, but several instances of transgranular cracking have now been experienced. The latter have involved service as well as hydro test failures, and transgranular cracks frequently have been found in the region of dents in those lines that display this mode of cracking. In all of those respects the transgranular and intergranular forms are identical, but the difference in crack morphology suggests that the mechanism of growth, and therefore the conditions under which the two forms occur, are different. Since the steels involved and the operating pressures are not essentially different for pipelines displaying the two different modes, the implication would appear to be that the different mechanisms result from variations in the environmental conditions.

Benefit: Field data relating to the transgranular stress corrosion cracking of high pressure gas pipelines indicates that the solutions found between coating and pipe in regions where cracks have occurred have low concentrations of dissolved salts and pH's of about 6.5, distinguishing them from those that are associated with the more common intergranular form of cracking. The lower pH solutions involved with transgranular cracking are the result of the presence of free carbon dioxide and the lack of significant cathodic current arriving at the pipe surface in those locations. Slow strain rate and cyclic loading tests employing solutions simulating those found in the field have shown that transgranular stress corrosion cracks can be generated under laboratory conditions and that they have the same characteristics as are observed in samples from lines that have developed cracks in service. Those similarities extend to the data from both field and laboratory sources showing considerable scatter, indicating the need for appropriate statistical treatment. It is suggested that the mechanism of transgranular cracking involves dissolution and the ingress of hydrogen to the steel and that implies that some of the concepts that have been employed in modeling the intergranular form of cracking will need to be modified for the transgranular case.

Result: The mechanism that has been suggested as possibly being involved with low pH cracking has important roles for pitting and the entry of hydrogen into the steel, neither of which plays a critical, or any, part in high pH cracking and therefore implying considerable difference in modeling. However, the evidence for the involvement of pitting and hydrogen leaves much to be desired and is often indirect. If the marked difference in the cyclic loading tests between the responses of the original surface of the pipe and polished surfaces is due to the former containing incipient pits, then it should be possible to artificially pit polished surfaces and cause them to develop cracks. That would constitute a reasonable test of the importance, or otherwise, of pitting. The importance of hydrogen in the crack growth process is supported by various observations, but direct measurement of the propensity for the low pH solutions to discharge hydrogen and for the latter to enter the steel in appropriate circumstances, e.g. without the application of low potentials, would be more convincing.

Product Details

Published:

09/01/1992

Number of Pages:

76

File Size:

1 file , 2 MB