7th Pipeline Technology Conference 2012

Approach to assessing Pipeline Displacements Dr. Ulrich Marewski Open Grid Europe, Gladbecker Strasse 404, 45326 Essen, Germany Wolfgang Schmidt, TÜV Rheinland Industrie Service GmbH, Am Grauen Stein, 51105 Köln, Germany Abstract High-pressure gas pipelines are usually designed for a specific internal pressure as the governing load parameter. Apart from the internal pressure, however, pipelines may also be subject to other significant loads. These additional loads often lead to additional longitudinal stresses in the pipeline. So far, the relevant German technical codes and standards for pipelines do not provide a concept for assessing longitudinal and/or bending stresses. International standards and publications dealing with additional longitudinal or bending stresses often refer to failure criteria that are either based on the concept of critical strain or critical stress, but neither of these concepts has been transposed directly into the German codes and standards. Open Grid Europe GmbH - the largest TSO in Germany - recently started using an IMU (Inertial Measurement Unit) as part of its pipeline pigging operations, which allows detecting bending strain in the pipeline. This makes it possible to also consider longitudinal stresses in addition to the internal pipeline pressure when it comes to assessing the integrity of a pipeline. One question that remains concerns the limiting criterion to be applied in the case of multi-axial loads. The aim of this project therefore was to propose a sufficiently conservative approach to assessing bending strain. 1.

Additional stresses on high-pressure gas pipelines

The possible causes of additional stresses on high-pressure gas pipelines described in the informative annexes B to F of EN 1594 [1] ("Pipelines for a maximum operating pressure over 16 bar") are -

Settlement Mining subsidence Frost heave Landslide areas Areas with high seismic risk

(Annex B) (Annex C) (Annex D) (Annex E) (Annex F)

Settlement can occur especially in areas where the pipeline has underlying layers of soft soil. Settlement may also occur where the pipeline crosses an elevated structure such as a road or dyke (Figure 1).

The amount of settlement will depend on the thickness, depth and type of the soft soil layers as well as the height and age of the elevated structure. It usually takes many years for the settlement process to come to an end. Where additional loads are known to exist, Open Grid Europe usually installs strain gauges connected to wireless long-distance data transmission equipment (Figure 2). Whenever predefined bending limits are exceeded, the company takes appropriate action such as cutting the pipe to reduce the loads. Pipelines in mining areas, on or near dykes and in areas with soft soils are routinely monitored.

1.1

IMUs inspection and strain assessment

Open Grid Europe has been using pigs for regular inspections of high-pressure gas pipelines for over 10 years. Inline inspection tools include MFL pigs and calliper pigs along with a module to record the line’s geographic coordinates. The recording accuracy for the coordinates is usually better than ± 1 m relative to the pipeline's precise location (Figure 3). However, this accuracy is not sufficient to detect any displacement or bending of pipes or pipeline sections. The IMU (Inertial Measurement Unit) used to determine the coordinates nevertheless allows any bending of a pipe or pipeline section to be quantified. This is done by calculating the curvature of the pipeline using signals received from the gyroscopic instruments and accelerometers as well as the distance measured by the onboard odometer [2]. The curvatures determined this way are then translated into bending strains (and stresses). Bending strain is shown where it exceeds  = 0.125 %, which corresponds to a bending stress of approx. 250 Mpa with an accuracy of  = 0.02 % (40 Mpa). The assessment only relies on bending strains – i.e. not on additional longitudinal tensile stresses – acting on pipeline sections only just completed, i.e. not in prefabricated bends or field bends. As a rule, all bending strains are recorded, which also includes bending resulting from any elastic installation of the pipeline. Since strain assessments are essentially an evaluation of data recorded, they can be conducted on any pipeline inspected by pigging, provided the pig used was fitted with an IMU module. Open Grid Europe is therefore assessing its newly pigged pipelines in addition to all pipelines inspected in previous years using inspection tools with an onboard IMU module (Figure 5). In the run-up to the project, the accuracy of the results of the strain assessment was cross-checked against a series of separate measurement surveys. This involved measuring the pipeline displacements derived from the strain assessment in a vertical and horizontal direction at several locations using highly accurate surveying methods (Figure 6). It was found that the pipe displacements derived from the strain assessment showed a very high accuracy similar in its order of magnitude to the accuracy of the geodetic survey. The displacements indicated were confirmed with an accuracy of a few centimetres, and the bending strains derived by far exceeded the accuracy specified above.

1.2

Displacement analysis

For the first of five high-pressure gas pipeline sections for which a strain assessment was performed, the conditions at site and the possible causes of the bending strain detected were analysed in detail. Overall, the analysis included a pipeline length of some 500 km with roughly 300 detected cases of bending strain of  > 0.125 % (Table 1). Most of the bending strain cases were close to the threshold value of  = 0.125 %; the number of cases with a higher bending strain decreased in the shape of an exponential distribution. The highest bending detected was  = 0.33 %, which corresponds to a longitudinal bending stress in the region of the material's yield strength (Figure 8). By far the largest number of bending strain cases were found to be due to installation-related causes, with bending strain frequently occurring in the area of naturally steep terrain, on spool pieces or underwater crossings. In addition, bending strain was also mainly found along sections installed in soft soils (close to flowing or stagnant waters) and/or roads (Figure 9). Again, it was found that the causes were most probably installation-related (heavy construction vehicles driving over the buried pipeline, etc.). In most cases, the bending strain had developed in a vertical direction, which meant that it had been caused by vertical loads.

2.

Approaches to assessing deformation

None of the relevant German codes and standards addresses the subject of pipeline design on the basis of critical strain. However, in its informative annexes, EN 1594 provides some general information on how to deal with additional longitudinal loads. Internationally, pipelines are today designed and assessed according to strain-based concepts in a multitude of ways. While onshore pipelines in America are usually based of a strain-based design, offshore pipelines are frequently designed according to the Norwegian standard OS-F101 [3], which allows longitudinal strains of over 1 % (Figure 10). For offshore pipelines laid from reels, longitudinal strains of up to 2 % can even be expected. A conversion by calculation of the minimum bending radius allowed according to DVGW G463 [4] for field bends produced by cold bending into longitudinal strain gives a maximum admissible longitudinal strain of 2.5 %. EN 12732 [5] (Annex G2) contains a recommendation for assessing defective girth welds on pipes (EPRG recommendation). Defects on girth welds smaller than 3 mm (in the direction of the wall thickness) and shorter than 7 times the wall thickness are regarded as admissible if the longitudinal tensile strain of the pipe is smaller than  = 0.5 % (as per level 2). By contrast, the minimum bending radius allowed during the elastic installation of a pipeline calculated according to DVGW G463 only gives a longitudinal strain of  = 0.071 % (for STE 480). Even the elastostatic design taking account of the "stress" limit state according to EN 1594 (Annex G) only provides admissible axial bending strains of around  = 0.2 % in situations of linear-elastic

material behaviour, which are far lower than the detected bending strains described above. However, the section on elastostatic design for the "strain" limit state also states that axial strains larger than  = 0.5 % are acceptable if the corresponding proof is provided, i.e. that in normal cases, strains of  < 0.5 % should thus be admissible. An assessment of additional axial load should, however, also take into account that apart from bending strain on pipelines subject to an internal pressure, there may also be other tensile strain components (Figure 11). The obstructed transversal contraction resulting from the internal pressure as well as a possible shrinking of the pipeline string due to thermal temperature changes may cause additional longitudinal tensile strain. In the case of major bending (e.g. following landslides) the resulting elongation of the pipeline string should also be taken into consideration. Since pipes are structures with relatively thin walls, they should not be seen as being able to withstand longitudinal strain and longitudinal pressures equally well (Figure 12). Depending on the level of the internal pressure, there is also the chance that instable buckling or kinking is initiated on the pipe's shell in compression [6]-[12]. Further information on limit values (Figure 13) is given in EN 1594, Annex G in the Section on the "deformation" limit state [13].

3.

Summary

Open Grid Europe analyses bending strains on its high-pressure gas pipelines as part of regular pig runs. The bending strains detected were verified both in qualitative and quantitative terms. They were found to be largely attributable to circumstances which occurred during pipeline installation. The concept proposed for assessing bending strains is based on EN 1594 (Annex G) involving the use of axial limit strains, with different limit values having to be observed for the tensile and compression strain components. The concept is valid only for defect-free pipes. In special cases (additional wall thickness reductions, buckling, large ovalities, etc.), case-by-case proof (finite element calculations etc.) will be required. The suggested approach was developed in cooperation with TÜV Rheinland.

References [1]

DIN EN 1594, Gasversorgungssysteme – Rohrleitungen für einen maximal zulässigen Betriebsdruck über 16 bar – Funktionale Anforderungen. Juni 2009

[2]

Fischer, I. (GE Energy Oil and Gas): Abschlussbericht Krümmungs- und Dehnungsuntersuchung im Auftrag von E.ON Ruhrgas, 2008

[3]

DNV Offshore standard DNV-OS-F101, Submarine Pipeline Systems; Det Norske Veritas, Norwegen, October 2010

[4]

DVGW Arbeitsblatt G 463, Gasleitungen aus Stahlrohren für einen Betriebsdruck > 16 bar – Errichtung, Dezember 2001

[5]

DIN EN 12732, Schweißen von Rohrleitungen aus Stahl – Funktionale Anforderungen, September 2000

[6]

Vitali, L.; Torselletti, E.; Marchesani, F.; Bruschi, R.: Strain based design for land high grade pipelines in harsh environments. Snamprogetti, Italien.

[7]

Hauch, S.; Bai, Y.: Bending Moment Capacity of Groove Corroded Pipes. American Bureau of Shipping Houston, Texas, USA. 2000

[8]

Iflefel, I. B.; Moffat, D. G.; Mistry, J.: The interaction of pressure and bending on a dented pipe. International Journal of Pressure Vessels and Piping 82, 2005, pp 761-769.

[9]

Brüggemann, H.; Schaumann, P.; Keindorf, C.: Elasto-platsiches Tragverhalten von Stahlrohren unter Innendruck und Biegung. 3R international (44), Heft 7, 2005

[10] Prion, H.G.L. and Birkemoe, P. C., “Beam-Column Behavior of Fabricated Steel Tubular Members,” Journal of Structural Engineering, Vol. 118, No. 5, pp. 12131232 (May 1992) [11] Sherman, D. R., “Flexural Tests of Fabricated Pipe Beams,” Third International Colloquium, Stability of Metal Structures, Toronto, Canada (May 1983) [12] Dorey, A. B., Murray, D. W., Cheng, R. J. J.: Critical Buckling Strain Equations for Energy Pipelines – A Parametric Study. Journal of Offshore Mechanics and Arctic Engineering, Vol. 128, pp 248 – 255, August 2006 [13] Gresnigt, A. M.: Plastic design of buried steel pipelines in settlement areas. Heron, Vol. 31 No. 4, 1986

Tables and Figures Table 1: Statistic of detected displacements from 5 pipelines Pipeline A B C D E Amount of detected displacements 73 52 63 43 74 Length of pipeline (km) 114 56 73 29 235 Displacements per km 0,64 0,93 0,86 1,48 0,31

Total 305 507 0,85

Figure 1: Settlement monitoring on a dyke by long-distance data transmission (Open Grid Europe)

Figure 2: Online analysis of strain conditions on a pipeline in a mining area (Open Grid Europe)

Figure 3: Marker box used to reference coordinates and measured deviation of pig coordinates (Open Grid Europe)

Figure 4: Inertial Measurement Unit with 3 gyroscopic instruments (rotation measurement) and 3 accelerometers to detect displacements (GE PII Pipetronix)

Figure 5: Evaluation of IMU inspection

Figure 6: Verification of detected displacements

Figure 7: Comparison of displacements detected by pig with results of survey

Figure 8: Distribution of detected bending strains at different pipelines

Figure 9: Locations of the largest strains

Figure 10: Acceptable strain limits according different standards

Figure 11: Proof of resistance against tensile strain (example)

Figure 12: Longitudinal strains in a pipe subjected to bending loads

Figure 13: Critical strains