Crack on Offshore Pipeline
Standar
Stress corrosion cracking can be a serious threat to the integrity of natural gas and petroleum pipelines. The pipeline industry responded to this threat by performing a comprehensive research program to determine the cause(s) of the failures and investigate various techniques for preventing future failures. A relatively concise list of discoveries has had a measurable impact on mitigation of the stress corrosion cracking threat.
Stress corrosion cracking can be a serious threat to the integrity of natural gas and petroleum pipelines. Photo: DNV
Starting with the first recognized stress corrosion cracking failure in 1965, the intergranular form of cracking (also known as high-pH SCC) was investigated to identify the causative agent and the controlling metallurgical, environmental, and stress related factors. In the 1980s, a second, transgranular form of stress corrosion cracking (near neutral pH SCC) was discovered in Canada, resulting in a similar scope of research activities designed to develop mitigation methods for this form of cracking. The information developed in these research programs has been incorporated into pipeline integrity management programs.
Three techniques
There are three common techniques used for management of the integrity of pipelines subject to stress corrosion cracking and other time dependent threats; hydrostatic testing, direct assessment, and in-line inspection (ILI). Directly following the initial stress corrosion cracking failures of gas transmission pipelines in the 1960s, hydrostatic testing was the primary tool used to confirm the integrity of the affected pipelines and prevent additional failures. The pipelines were pressure tested with water at pressures significantly higher than the operating pressure in order to remove any near critical flaws. This technique has a number of limitations. Very few, if any, stress corrosion cracking flaws are removed, and the pipeline must be taken out of service for testing. In dry climates, obtaining adequate sources of water can be a challenge, while freezing of the water can be an issue in winter months or northern climates. For liquid petroleum pipelines, the water must be extensively treated prior to discharge back into the environment. There also is a finite probability of a phenomenon known as a pressure reversal occurring, where the failure pressure after a hydrostatic test is lower than the maximum test pressure, as a result of subcritical crack growth during the hydrostatic test.
There are three common techniques used for management of the integrity of pipelines subject to stress corrosion cracking and other time dependent threats; hydrostatic testing, direct assessment, and in-line inspection (ILI). Directly following the initial stress corrosion cracking failures of gas transmission pipelines in the 1960s, hydrostatic testing was the primary tool used to confirm the integrity of the affected pipelines and prevent additional failures. The pipelines were pressure tested with water at pressures significantly higher than the operating pressure in order to remove any near critical flaws. This technique has a number of limitations. Very few, if any, stress corrosion cracking flaws are removed, and the pipeline must be taken out of service for testing. In dry climates, obtaining adequate sources of water can be a challenge, while freezing of the water can be an issue in winter months or northern climates. For liquid petroleum pipelines, the water must be extensively treated prior to discharge back into the environment. There also is a finite probability of a phenomenon known as a pressure reversal occurring, where the failure pressure after a hydrostatic test is lower than the maximum test pressure, as a result of subcritical crack growth during the hydrostatic test.
Alternatives to hydrostatic testing
Because of these limitations, there has been significant interest within the pipeline community in the development of alternatives to hydrostatic testing. One alternative is SCC Direct Assessment (SCCDA). The first recommended practice for SCCDA was issued in 2004 (NACE Standard RP0204-2004). SCCDA is a structured process intended to assist pipeline companies in assessing the extent of stress corrosion cracking on buried pipelines, thus contributing to their efforts to improve safety by reducing the impact of external stress corrosion cracking on pipeline integrity. The term is somewhat of a misnomer in that the process is much more extensive than simply examining the pipeline for evidence of stress corrosion cracking.
Because of these limitations, there has been significant interest within the pipeline community in the development of alternatives to hydrostatic testing. One alternative is SCC Direct Assessment (SCCDA). The first recommended practice for SCCDA was issued in 2004 (NACE Standard RP0204-2004). SCCDA is a structured process intended to assist pipeline companies in assessing the extent of stress corrosion cracking on buried pipelines, thus contributing to their efforts to improve safety by reducing the impact of external stress corrosion cracking on pipeline integrity. The term is somewhat of a misnomer in that the process is much more extensive than simply examining the pipeline for evidence of stress corrosion cracking.
The first step in the process (pre-assessment) involves the collection of existing information on the pipeline that can be used to assess the likelihood that the pipeline is susceptible to stress corrosion cracking. The applied research described above has formed the basis for establishing the critical information that should be collected. In the case of high-pH stress corrosion cracking, the initial selection of the most susceptible segments is based on five factors; operating stress (>60% of the specified minimum yield strength), temperature (>100°F), distance from compressor station (< 20 miles), pipeline age (>10 years) and coating type (other than fusion bonded epoxy (FBE)). In the case of near neutral pH SCC, there are four factors, excluding operating temperature.
Other types of information on the pipeline can be used for the selection of dig sites in the chosen segments. Again, much of this information is based on the applied research performed and includes factors such as topography, drainage, and soil type (near neutral pH SCC), the magnitude and frequency of cyclic pressure fluctuations, the specific coating type, including the girth weld coating, surface preparation for the coating, coating condition and prior history on the pipeline. A significant issue with SCCDA is that it is not capable of reliably identifying the location or locations of the most severe stress corrosion cracking on a pipeline segment. Accordingly, it is not necessarily a replacement for hydrostatic testing or in-line inspection in all instances. The pre-assessment phase of SCCDA may indicate that a particular pipeline segment is not likely to be susceptible to stress corrosion cracking and therefore, other threats are of a more immediate concern. An example would be a newer pipeline with an FBE coating. On the other hand, ILI, hydrostatic testing, or even pipe replacement may be warranted if extensive, severe stress corrosion cracking is found.
In-line inspection is the third technique used to manage time-dependent threats on operation pipelines. There is a long history of using magnetic flux leakage and, to a lesser extent, ultrasonic tools to address internal and extern corrosion threats on transmission pipelines. Around the time the NACE SCCDA recommended practice was being developed, there was a growing consensus within the pipeline industry that the new generation of crack detection tools would eliminate the need for hydrostatic testing or any of the elements of SCCDA. Integrity management would consist of four elements; find the cracks, size the cracks, assess the cracks and repair the cracks.
Unfortunately, the experience with crack detection tools has not been as desirable as had been hoped. Over the past few years, there have been at least two service failures of pipelines in which the operators were using crack detection tools for integrity management. In the case of one failure in Carmichael MS, the cause and contributing factors could not be confirmed because of extensive fire damage to the failed pipe. A portion of the fracture did propagate along the seam weld and the pipeline had a history of seam weld defects. The pipe section that contained the failure also had been previously hydrostatically tested at a pressure much higher than the failure pressure. This information, taken together, suggests that a likely cause of failure was the growth of a seam weld defect in service. In the second case, a failure near Marshall MI, there was evidence that fatigue cracks grew from a colony of pre-existing stress corrosion cracks.
Accuracy of crack sizing
The problem with the current generation of crack detection tools appears to be related to the accuracy of crack sizing. The tools are very good at finding crack-like features, and the length accuracy for the features also typically is good. The problem is the depth accuracy. The feature calls are usually binned by depth into several depths; e.g. <15% of wall thickness, 15–30% of wall thickness, 30–45% of wall thickness and >45% of wall thickness. Currently, there does not appear to be sufficient accuracy in the binning process for integrity management purposes. The failure pressures in pipelines containing cracks are much more sensitive to depth than to length of the flaws. The problem could be related to the precision and accuracy of the tool, or the analytical process for analyzing the tool data.
The problem with the current generation of crack detection tools appears to be related to the accuracy of crack sizing. The tools are very good at finding crack-like features, and the length accuracy for the features also typically is good. The problem is the depth accuracy. The feature calls are usually binned by depth into several depths; e.g. <15% of wall thickness, 15–30% of wall thickness, 30–45% of wall thickness and >45% of wall thickness. Currently, there does not appear to be sufficient accuracy in the binning process for integrity management purposes. The failure pressures in pipelines containing cracks are much more sensitive to depth than to length of the flaws. The problem could be related to the precision and accuracy of the tool, or the analytical process for analyzing the tool data.
Because of these sizing issues, pipeline operators are sometimes required to perform a confirmatory hydrostatic test on portions of their system to demonstrate the performance of the crack detection tool. In some cases, pipeline operators also may be faced with a situation where a large number (thousands) of features are found on a segment of a pipeline. The elements of SCCDA can then be used, in conjunction with the standard crack sizing, to identify which feature are most likely to be an integrity threat and should be excavated. These elements are also used to prioritize pipeline systems for in-line inspection. The point is that fundamental understanding of the SCC process, developed through applied research, is still of significant value in integrity management.
As previously described, the information developed in the research programs is also used by pipeline companies to operate their pipelines in a way that reduces the risk of stress corrosion cracking failures. Where there is a risk of high pH SCC, after-coolers are installed at compressor stations to reduce outlet temperatures. Operators may control pressure fluctuations to reduce the risk of both forms of stress corrosion cracking. Cathodic protection systems are enhanced, monitored, and maintained to keep the pipe to soil potential out of the potential range for cracking. Finally, this research information is used by prudent operators in construction activities to reduce the risk of stress corrosion cracking in the future. Near white surface preparation procedures and coatings, such as FBE, are selected to minimize stress corrosion cracking initiation. Girth weld coating systems are selected to minimize shielding. The pipeline systems are designed to minimize operating temperatures and cyclic pressures.
Future trends
The detection and sizing capabilities of the crack detection tools will undoubtedly improve, but there will always be a need for elements of SCCDA, developed through applied research, to prioritize pipeline segments for inspection. Depending on the pace of improvements in the precision and accuracy of these systems, there may continue to be a necessity, for the near future, to use elements of SCCDA to help identify which features are most likely to be SCC threats.
The detection and sizing capabilities of the crack detection tools will undoubtedly improve, but there will always be a need for elements of SCCDA, developed through applied research, to prioritize pipeline segments for inspection. Depending on the pace of improvements in the precision and accuracy of these systems, there may continue to be a necessity, for the near future, to use elements of SCCDA to help identify which features are most likely to be SCC threats.
Although not discussed in this article, there are other stress corrosion cracking threats to operating pipelines. Ethanol is used in almost all of the gasoline consumed in the United States as an oxygenating agent and octane booster. The current blend limit is 10% ethanol, but this may be increased to 15% in the near future. Denatured fuel grade ethanol is transported to terminals where it is blended with gasoline. Pipelines are the most efficient method to transport the ethanol but the threat of internal stress corrosion cracking has prevented general transportation via pipelines. Research is ongoing to address this threat.
A new internal stress corrosion cracking threat from alcohol was recently identified. Methanol is sometimes used as a drying agent for pipelines directly following hydrostatic testing. In northern climates, neat methanol is also sometimes used for hydrostatic testing to avoid freezing. There is growing evidence that this methanol has caused internal stress corrosion cracking of some pipelines; research is ongoing to address this threat.
Text: John A. Beavers, DNV
Source :