Recognising that the industry was moving to higher design factors and that high strength, hot-coated pipe (the hot coating process can change the material properties) was becoming an industry standard, the research was extended to develop software that would enable designers to conduct a simulated strength test on a pipeline test section, using actual data on the pipe elevation, the strength of individual pipe and the stress-strain relationship associated with the pipe. This software (PipeStrain) was tested and used by the pipeline industry.
The Australian Standard for field pressure testing (AS2885.5) has been revised (2012). The revision recognised that:
- Most pipelines being constructed in Australia have adopted a design factor of 0.8, requiring a minimum strength test pressure equivalent to 100 per cent of the pipe specified minimum yield strength (SMYS). This means that all pipe below the high point will be exposed to more than 100 per cent SMYS, creating the potential for some lower strength pipes to yield or even burst.
- The offset volume pressurisation endpoint method introduced with AS1978 is not capable of identifying the onset of yield in a few pipes, particularly with high yield to tensile ratio (Y/T) materials.
Consequently, the standard has redefined the strength test end points, based on the potential for any pipe to be exposed to a risk of plastic strain (yielding) during testing.
The first method allows the designer to take advantage of the actual yield strength of the pipe population (based on mill or other physical strength tests), and to define the maximum pressure at the low point of a test section by the lowest yield strength of pipe in the pipe population. When the test requirements and the pipe strength allows this method, the test section can be designed with confidence that no pipe will be exposed to unacceptable plastic strain.
The second method requires analysis of the pipe test section to predict the strain in each pipe in the test section, using as-built construction records and the strength and stress-strain data for the pipe. In order to provide the industry with a tool to undertake this analysis, the original PipeStrain software has been revised to improve usability and also to incorporate a strain limit; based on the results of extensive full-scale burst tests undertaken by the European Pipeline Research Group (EPRG) and incorporating the results of burst tests conducted by Australian researchers.
PipeStrain has been revised through an Energy Pipelines co-operative research centre project. The PipeStrain software can predict the behaviour of a pipeline section during hydrotesting (the pressure-volume plot), and predicts the hoop strains in each pipe in the section, to establish that no hoop strain is excessive. The software will be released for limited industry testing to confirm its performance and functionality. Once approved, it will be licensed to industry users on application, by APIA for use in conjunction with AS2885.5-2012.
Pipeline modelling
The behaviour of a pipeline in a field hydrostatic test is complex. Factors that affect the pipeline response are:
- The strain rate of the hydrotest is lower than strain rate used to measure the pipe yield strength measurement lowering the yield point [2];
- The difference between the restraint condition during the hydrotest and yield strength determination changes pipe yield behaviour [3];
- Material properties and thickness which vary with each pipe along the pipeline [4];
- Elevation differences that lead to a different elevation pressure head for each pipe superimposed upon the test pressure;
- Changes in Poisson’s ratio that affect stress and yielding as plasticity progresses [5,6]; and,
- The change that may occur in yield strength caused by ageing during the coating process.
The outcomes of this modelling provide an understanding of pipeline hydrostatic strength test behaviour. These factors have been re-assessed and incorporated in a revised version of PipeStrain.
The revised PipeStrain application will introduce the following changes to the previous versions:
- The input files and any output file are now in Excel format;
- Each pipe’s stress-strain curve will be adjusted for the pipe’s tensile and yield strengths;
- Output files will be calculated for the elevation where the test pressures are being read; and,
- A permissible maximum strain for each pipe will be calculated. This is based upon the pipe’s yield-to-tensile ratio and, in accordance with the revised AS2885.5-2012, is one-third of the burst strain established by extensive EPRG research.
Material properties
The material properties (the yield and tensile strengths, and a characteristic stress-strain curve taken from post-coated pipe), location, and elevation of each pipe must be known for the pipe in the test section. The accuracy of the predicted behaviour of the section depends on the accuracy with which the properties are known, particularly for higher grade pipe where the lower strain-hardening behaviour greatly amplifies any errors in the predicted strain.
Accurate modelling requires yield strengths and tensile strengths to be determined for each batch of pipe (usually each heat). It also requires accurate stress-strain curves taken from post-coated pipe, which faithfully represent the straining behaviour of the full circumference of the pipe (rather than just some part of it). Ring-expansion testing provides accurate stress-strain behaviour; this is rarely possible on large diameter pipe. The commonly available round-bar testing provides spurious yield strength values and a distorted estimate of the stress-strain curve. Flattened bar tensile testing is also inaccurate.
Pipe is often protected with a fusion-bonded epoxy (FBE) coating. This involves heating the pipe to a temperature of approximately 250°Celsius as part of the application process and may lead to an increase in yield strength due to strain ageing. After coating, many pipe exhibit an increase in yield strength of more than 10 per cent. The yield-to-tensile ratio may increase and the rupture ductility may be significantly reduced.
If pipe is to be subject to high-level hydrostatic testing, appropriate data has to be procured, either by agreement with the manufacturer, or by testing after delivery of the pipe. It is worth drawing attention to the need for careful thought when specifying the material tests, the need for matched samples to establish data on the yield strength increase with heating, and the need for stress-strain data from coated material. To avoid excessive cost and disruption, this data should be specified as a deliverable from the pipe supplier and also after coating. AS2885.1-2012 and Appendix P in AS2885.5-2012 provide guidance on the data requirements and its analysis.
Strain limit
With the ability to calculate strains for each pipe, the question of what is a rational strain limit arises.
Pipeline steels exhibit a strain-to-failure under tensile testing of more than 20 per cent. In the tensile test, the strain at the ultimate tensile strength is the uniform strain which is often less than 10 per cent.
In a pressurised pipeline the average hoop strain before failure seen in testing may be 2 per cent or less, a small fraction of the total elongation in a uniaxial tensile test. The hoop strain-to-failure is strongly influenced by the yield-to-tensile ratio, so low failure strains can be experienced in high yield-to-tensile ratio pipeline steels.
For high-level testing, AS2885.5-2012 nominates the allowable strain as the EPRG/3 limit; that is one third of the correlation curve for the lower bound of data for circumferential plastic strain (Figure 2) at burst for the yield-to-tensile strength ratio for each pipe. This can be expressed by: Strain limit (plastic, %) = -127.15R3 + 297.24R2-236.12R+65.22 where R is the yield-to-tensile ratio of each coated pipe. For lower strength pipe, where total strain may be greater than 2 per cent, the strain capacity of the coating must also be considered.
Conclusion
PipeStrain software reduces the risk of damage to pipes and/or coatings during field hydrostatic testing while providing a means to utilise the reserve ability of the pipe to safely withstand the plastic deformation required by the field pressure test when the design factor is 0.8.
The software has two uses:
- During pipeline design; designers can apply the pipe design wall thicknesses, expected strength distribution and typical stress-strain characteristics to the pipeline profile to refine the pipeline design (locating heavy wall thickness pipe at low points), or refine the locations of hydrostatic test sections. The simulation provides understanding of the reserve ductility of the pipeline, and avoids risk of damage to coatings.
- Prior to conducting the hydrostatic test, the designer and/or the hydrostatic test engineer can apply the as-built pipeline profile, pipe wall thicknesses, pipe strength and strain characteristics and predict the behaviour expected from the test section.
There are three key outcomes:
- The engineering software prediction will clearly identify the location(s) of low-strength or thin-wall pipes that have a potential to experience excessive plastic strain during the test. This will benefit the pipeline design and allow a test section to be checked prior to pressurisation.
- The analysis will allow the test section design to be changed if required to eliminate pipes where the strain limit is exceeded, and where the pipe properties permit, allow increased elevation differences to be accommodated, possibly eliminating some test sections.
- A simulated P-V Plot will be generated for monitoring during the actual field pressure test.
Information on how to access PipeStrain will shortly be available from APIA.
Co-author Dr Michael Law will be presenting on “˜High Design Factors, High Level Pressure Testing, Use of PipeStrain Software and Data Requirements’ at the 2012 APIA Convention at 4:00pm on Monday 15 October.
References
1. Gaessler H, Vogt G (1989) “Influence of yield to tensile ratio on the safety of pipelines” 3R International V28 April, pp 165 – 172
2. Gunn K (1980) “Strain Rate and the Yield Stress of Pipe Steels” Canadian Metallurgical Quarterly V19, pp 59 – 77
3. Benham P, Crawford R (1987) “Mechanics of Engineering Materials” Longman Scientific and Technical, New York
4. Jiao G, Sotberg T et al (1997) “The SUPERB project: linepipe statistical properties and implications in design of offshore pipelines” ASME Int. Conf. On Offshore Mechanics and Arctic Engineering, Yokohama
5. Simo J, Taylor R (1985) “Consistent tangent operators for rate-independent elastoplasticity” Computer methods in applied mechanics and engineering V 48, pp 101-118
6. MacKenzie P et al (1986) ” Measurement of Poisson’s ratio through the elastic-plastic transition” Strain, Feb 1986 pp 13 – 19
7. Liessem A Graef MK, Knauf G, Marewski U “Influence of thermal treatment on mechanical properties of UOE linepipe” Pipeline Technology Conference, Oostend, Belgium May 2004, pp 1263-1281.
