by Vlad Semiga, BMT Fleet Technology Ltd, Senior Pipeline Integrity Specialist
There are commonly held beliefs that the internal operational pressure fluctuations of a gas pipeline are less likely to result in fatigue cracking than a liquid pipeline, and that internal-pressure-induced fatigue is not an issue for most gas pipeline systems. As a result BMT Fleet Technology, on behalf of the Interstate Natural Gas Association of America (INGAA), 1 has developed an defensible engineering analysis process to investigate whether these observations remain valid for a gas pipeline system.
The analysis process takes the following into consideration:
- the pipeline’s operation, be it transmission, bi-directional flow at storage facilities
- existing features, such as dents or cracks
- material properties, such as pipe vintage and grade
- overall pipeline geometry.
The assessment supports gas pipeline operators by completing integrity-verification programs (IVP) which define pipeline segments that can reasonably be considered to be at risk of fatigue damage accumulation, and those which are not. Historically speaking, there have been very few incidents of cyclic-pressure-induced fatigue related failures in gas pipeline systems.2 There are a number of reasons for this, including:
- The general operation of a gas pipeline results in few large-amplitude internal pressure cycles being applied to the system – for example, continuous product shipping without the use of batching operations.
- The product being shipped is compressible in nature, meaning that changes in gas pressure occurring at the compressor station are dampened and do not propagate a significant distance along the pipeline.
Pressure spectrum severity
The fatigue life of a structural system is determined by considering the severity of cyclic loading, the geometry of the system supporting the cyclic load – for example pipe and including anomalies – and, in some instances, material properties. A significant difference between the fatigue susceptibility of gas and liquid pipelines can lie in the severity of the cyclic operating line pressure history.
Often the operation of a pipeline is categorised based on its maximum operating pressure (MOP), and while this approach might be suitable when categorising a pipeline from a general static-strength point of view, it does not provide an accurate indication of the pressure-induced cyclic-fatigue severity of operations. The MOP does provide a general limit of the maximum cyclic pressure that the pipeline can experience.
However, if a pipeline operates at a high MOP – 70 per cent specified minimum yield strength (SMYS) – and in a continuous manner with few pressure drops, it could be less susceptible to cyclic pressure-induced fatigue than a pipeline that operated at a lower maximum pressure – 30 per cent SMYS – that sees frequent pressure drops (e.g. shutdowns).
Therefore, when assessing or categorising a pipeline’s susceptibility to fatigue, the operational usage of the pipeline must be considered. The most direct way to accomplish this is to base the assessment on an actual detailed pressure time history of the pipeline, such as that provided by the pipeline supervisory control and data acquisition (SCADA) system.
To efficiently employ a SCADA report, or other pressure time history in a fatigue-life calculation, the complex variable-amplitude time history is represented statistically as frequency histogram of constant-amplitude pressure-change events. The process most widely used process to do this is rainflow counting3.
In this process, the output of a cycle count analysis is a histogram of applied pressure ranges and the associated number of cycles at each pressure range. When considering fatigue-crack growth using a fracture-mechanics’ approach in the absence of a threshold, every pressure cycle induces appreciable crack growth. The relative severity of various pressure-time histories is not easily discernable from the pressure-time history or the rainflow count histogram.
For this reason, BMT developed a spectrum-severity-indicator (SSI) characterisation technique that considers the number of cycles of a given pressure range required to grow a crack the same amount as the actual pressure-time history over one year. An example of this approach is illustrated in Figure 1, where the SSI is the number of 13 ksi hoop-stress cycles required to grow a crack the same amount as one year of the actual pressure-time history.
The higher the number of 13 ksi hoop-stress cycles associated with a time history, the more aggressive the spectrum is from a cyclic pressure (i.e. fatigue life) point of view.
In general, for a pipeline, the spectrum severity associated with the discharge or suction pressure-time history will be different. As such, when assessing the susceptibility of a gas pipeline to cyclic-pressure-induced fatigue, it is conservative to base the assessment on the most extreme of the discharge or suction pressure time history severities from the compressor stations bounding the pipeline segment.
Observed operation spectrum severities
To appreciate the magnitude and range of SSIs that can be observed in gas pipelines operating line pressure, 103 pressure histories were collected from gas pipeline systems considered by their operators to be representative of aggressive, moderate, and benign cyclic operations. Rainflow cycle counting was carried out on each of the pressure time histories and SSI’s were estimated.
Among the pipelines considered were:
- 81 continuous operational lines
- 30 bi-directional and 56 uni-directional lines
- mostly transmission pipelines, with three lines used in storage fields and ten in mixed operational use
- lines constructed between 1910–2010 (majority 1950–1970)
- diameters from 6.75 to 42 inch
- wall thicknesses from 0.156–0.844 inch
- yield strength from 35–70 ksi
(majority 52 ksi) - mean to yield pressure ratio of 10–80 per cent (majority 50–70 per cent).
The results, which consider the SSI magnitudes for these gas pipeline pressure time histories, illustrate the range of cyclic-pressure operation time histories (see Figure 2). This indicates that for the pipeline operations considered, the SSI ranged from 20 to 340 equivalent 13 ksi cycles per year. In comparison, a liquid pipeline with a severe operational-time history has been observed to have an SSI of 3,000 equivalent 13 ksi cycles per year.
Pipeline feature fatigue life limit
In the INGAA-sponsored study, the results of these fracture-mechanics’ based fatigue-life calculations was presented as a series of tables and graphs that identify features that can survive either greater than 100, or greater than 200, years of service based upon the pipeline-internal-pressure SSI. Based upon the sensitivities explored, the allowable initial crack size for a given fatigue life is primarily a function of the SSI and the pipe-wall thickness. Both the OD and the pipe grade were considered to have small effects on the allowable initial crack size.
Therefore, a family of fatigue limit curves were developed for four pipe wall thicknesses of 0.156, 0.25, 0.312, and 0.5 inch. These fatigue-limit curves applied to seven spectrum severities and two fatigue-life criteria (100 and 200 years). As an example, the 100 year fatigue life curves for the 0.25 inch pipe wall thickness with an axial crack is presented in Figure 3.
Assessment of the fatigue susceptibility for thicknesses other than those developed can be interpolated or conservatively considered by applying the criterion developed for the next-lower thickness. Once the pipeline operator determines the maximum size of features (e.g. cracks) contained in their system, based upon inspection or pressure testing, this criterion can be used to demonstrate that the pipeline system features will not be susceptible to pressure induced fatigue and will have a long life.
Concluding remarks
A key aspect of any pipeline integrity management and verification program is to identify threats to a pipeline’s integrity. As with other integrity threats, the risk of fatigue must be understood and characterised correctly by a pipeline operator in order to prioritise responses and minimise the chance of it impacting the integrity of a system.
The work described in this article provides the tools to consider the susceptibility of a pipeline segment to fatigue. These criteria would be of particular interest to gas pipelines where the operational internal pressure fluctuation severity is not severe and consideration of fatigue may be eliminated as an integrity threat.
The fatigue assessment criteria were developed around the concept of characterising the operational-pressure spectrum severity in terms of an SSI. This parameter characterises the fatigue severity of the pipeline operating-pressure spectrum, allowing for comparison and ranking of operating conditions and pipeline segments.
The internal cyclic pressure induced fatigue-severity criteria which have been developed, conservatively considers axial crack-like features and plain dent features, and provides a means of demonstrating if a gas pipeline is susceptible to internal cyclic pressure load induced fatigue damage accumulation. The approach presented considers the presence of pipe-wall anomalies and can be conservatively applied to other feature types (e.g. circumferential cracks, SCC, and dents with corrosion).
Acknowledgments
The author gratefully acknowledges the technical direction and support of the INGAA and its member companies in the completion of this work. Some data and concepts considered in this work were developed and influenced by results of Pipeline Research Council International research. The fatigue-life-assessment work presented in this article was completed by the team at BMT Fleet Technology, including V. Semiga, A. Dinovitzer, Dr S. Tiku, and Dr A. Eshraghi.
References
1. V. Semiga, ‘Evaluation of Fatigue in Gas Pipelines’, 11th International Pipeline Conference, Paper IPC2016-64486
2. INGAA, ‘Historic Review of Gas Pipeline Fatigue Induced Failures’, INGAA 2015
3. American Society for Testing and Materials, ‘Standard Practices for Cycle Counting in Fatigue Analysis’, ASTM E1049-85 (Re-approved 1997).
This article was featured in the June edition of Pipelines International. To view the magazine on your PC, Mac, tablet, or mobile device, click here.
If you have a technical paper you would like featured in Pipelines International contact Assistant Editor Nick Lovering at nlovering@gs-press.com.au