EWGAE 2010

Vienna, 8th to 10th September

Buckling detection within subsea pipeline laying using Acoustic

Emission technique

M. F. SHEHADEH, M. Abdel GELIEL, A. EL-ARABY

Arab Academy for Science, Technology and maritime Transport (AASTMT), Engineering College, Marine
Engineering Dept., Alexandria, Egypt; [email protected]

Keywords: Pipeline deformation, Buckling, Critical loads, Signal Conditioning

Abstract

The most critical factor on operation of subsea pipeline installation is control of stresses (e.g.
buckling, bending, fatigue etc.) in pipeline as it travels from the pipeline laying barge to the sea
floor. Hence, online monitoring for pipeline laying is an important task for early detection for
abnormal situation that may lead to catastrophic accident. This paper focuses on early detection of
buckling stresses produced by the subsea pipeline at the laying operation using AE technique. The
sensing configurations and signal analysis are carried out experimentally into a pipeline system in the
lab. The results have shown that AE technique can be used for early detection of pipes buckling and
bending.

Introduction

Pipes are of major importance for transport of liquids and gases mainly for natural gas and oil
either in shore or offshore. Subsea pipeline assembly is laid from a vessel into a body of water by
passing the pipeline through the water in a substantially vertical path for a substantial portion of its
descending travel.
The source of the acoustic emission energy is the elastic stress field in the material. Without
stress, there is no emission. Therefore, an acoustic emission (AE) inspection is usually carried out
during a controlled loading of the structure. This can be a proof load before service, a controlled
variation of load while the structure is in service, a fatigue test, a buckling test, or a complex loading
program. AE inspection is used because it gives valuable additional information about the
performance of the structure under load. Other times, AE inspection is selected for reasons of
economy or safety, and a special loading procedure is arranged to meet the needs of the AE test [1].
To keep up with the discovery of deepwater oil and gas fields, the J-lay system for pipeline
installation was invented. This method has been developed as an alternative method to install a
pipeline in very deep waters [2]. The pipeline leaves the vessel from a nearly vertical position, as
shown schematically in Fig.1 on the way down to the seabed; it acquires the characteristic J-shape
from which the name J-lay is derived. This case generates complex loading conditions which
influence the pipeline design.
The loads experienced by the pipe during such deepwater J-lay are as illustrated schematically in
Fig.1. High tension and relatively small external pressure close to the surface of the sea,
progressively increasing pressure and decreasing tension further down the long suspended section,
high external pressure and bending in the sagbend, and essentially hydrostatic pressure on a
flat seabed [3]. The structural integrity of the pipeline moving from the launching ramp to the touch
down point has to be controlled in real time [4]. As the AE is defined in ASTM E1316 terminology
for Nondestructive Examinations by the class of phenomena whereby transient elastic waves are
generated by the rapid release of energy from localized sources within a material, this method can be
used for real time monitoring at the laying operation.
At the pipeline laying process buckling stress, due to failure in installation system and/or rough
sea condition, switches the pipe from the straight, stiff configuration to the bent one that has very
small stiffness the transformation take place at a certain load which is known as the critical buckling
load. The most familiar example of structural buckling is the column. A column loaded within its
design specifications is straight and carries axial compression by undergoing membrane
deformations on reaching a critical load; a bent configuration becomes energetically.
The diagnosis of pipelines such as buckling and bending is carried out using Fault Detection and
Diagnosis (FDD) techniques [5]. FDD techniques are classified into model based and signal based
[6]. The selection of an FDD technique is based on the system characteristics and available
information. Since detection of pipeline buckling, in this paper, is done using AE signal, detection of
buckling, here, is classified as a signal based FDD method. Signal based FDD methods vary from
simple method such as limit checking and trend checking to a very advanced method based on new
techniques such as Principle Component Analysis (PMC) 7], Wavelet Analysis (WA) [8] etc. This
paper focuses on using the AE method for early detection of the critical pipeline buckles load.

Fig. 1. Schematic representation of J-lay installation and associated

pipeline loading [3]

Acoustic Emission system

Two USB AE node acquisition systems have been used in this work. The USB AE node has
internal AE preamplifier along with the capability for powering a low power line of external
preamplifiers and/or Integral Preamplifier sensors, which give the ability to select between the use of
the internal and external preamplifier via software. Two sensors types WSa are used, the “a” stands
for Alpha series of sensors. These sensors have built-in amplifiers and a fairly flat frequency
response but with two bands of relatively high sensitivity at around 100 kHz and 1000 kHz and they
have an operating temperature from -65 C to175 C. The sensors are 19 mm in diameter, 21 mm high
and 32 gm, the surface was kept smooth and clean and silicone grease was used as couplet to fill any
gaps caused by surface roughness and to eliminate air gaps which might otherwise impair AE
transmission.
Series of experiments were carried out to investigate the propagation of AE from simulated
(pencil lead break) and real (buckling) sources to an array of sensors. The overall purpose was to
evaluate the potential of AE monitoring to be used in real pipe laying systems. Given the size of
actual pipelines, full scale testing is difficult and can be prohibitively expensive. The approach taken
in most of the research studies use scaling in order to reduce the size of the specimens tested.
Dimensional analysis considerations easily show that a key geometric non-dimensional variable is
the diameter to thickness ratio (D/t). Therefore, experimental results obtained using small scale tubes
and pipes are transferable to actual large diameter structures with the same D/t. Regarding material
properties, the elastic modulus/yield strength ratio and the hardening exponent are often key non-
dimensional variables [3]. When scaling is adopted in analyzing structural instabilities, it is also
important to differentiate between ones that are elastic and those that occur in the plastic range of the
material. A seamless steel (ASTM A106 Grade B) pipe was used in all experiments. The pipe was
1.5m long, with a diameter of 50.8 mm (2”) and wall thickness of 4.5mm.

Wave speed estimation

The wave speed of AE propagation in structures can be treated theoretically as a variety of pure
modes implying simple sources and structures, even though the theoretical approach considers the
structure to be a homogeneous elastic continuum and without attenuation effects, which may not be
applicable to pipes in real applications [9].The experiment shown schematically in Fig.2 was held to
experimentally estimate the AE wave speed, where the first sensor was mounted 40 cm from one end
of the pipe, while the other sensor was incrementally moved axially along the pipe. The data were
sampled at 10 MSPS. Pencil lead breaks were performed on the surface of the pipe beside the trigger
sensor, as shown in Fig.2. The spacing between the sensor (S1) and the second sensor (S2) is
increasing with a constant interval of 20 cm for all of the locations. A total of 4 positions were used,
the experiment being repeated 5 times at each position. A common equation used for linear
estimation [10]:

x=½ (D - v. ∆t) (1)

where D is the distance between sensors; v is the constant wave velocity; ∆t is the time difference
between hits; and x is the source location measured from the first hit sensor. For each AE test the
time differences (∆t) in ms was recorded, the averages of wave speeds were calculated manually and
found to be 5000 m/s.

Experimental Procedures

Universal testing machine (i.e. max load 294kN) was used to compress the pipe of nominal
length of 60cm. The pipe was prepared for destructive testing by pined joint at both ends as shown in
Fig.3. An axial compression load was applied incrementally until the pipe is totally deformed. The
critical buckling load (Pcr) can be calculated by formula:

Pcr= π2 E I / L2 (2)
where Pcr is the critical load; E is the modulus of elasticity; I is the second moment of area; and L is
the effective length.

Trigger
Sensor

Fig. 2: Schematic layout of positions for source-sensor distances

The abnormal situation of pipeline is detected if its magnitude or/and energy exceeds certain
threshold [6] determined experimentally. Two sensors were positioned 10 cm away axially from the
mid-span. The AE signal was recorded with sampling rate of 10MHz, 40dB as a threshold and the
accusation timing was set up as 300µs for Peak Definition Time (PDT), 600µs for Hit Definition
Time (HDT) 4 and 1000µs for Hit Lookout Time (HLT) are timing parameters of the signal
measurement process. A proper setting of the PDT ensures correct identification of the signal peak
for rise time and peak amplitude measurements. Proper setting of the HDT ensures that each AE
signal from the structure is reported as one and only one hit. With proper setting of the HLT,
spurious measurements during the signal decay are avoided and data acquisition speed can be
increased.

Results and Discussions

The analyses are based mostly on signal processing [5] techniques on sensor 1 using commercial
software. The main objective of the analysis was to understand the behavior of the pipeline under
axial compression loading along with early detection of buckling. The loads increased gradually to
check the behavior of AE signal during the elastic and plastic mode of pipelines operation as shown
in Fig.3. The two channels have the typical distribution with time difference which are based on the
sensors position.
To correlate the collapse behavior with AE parameters, for elastic steel structures three typical
modes were considered. The load and the energy distribution during the entire test are shown in
Fig.4. The deformation mode is changing from the elastic type (zone A), to plastic type (zone B);
with the progress of collapse in (zone C). From the previous discussion, the curve may
hypothetically be divided into three main zones:
Zone A: Elastic deformation zone.
Zone B: Plastic deformation.
Zone C: collapse zone.
Since, Zone B is the most important stage of the test period which the material transforms from
elastic to plastic states. As shown in Fig.4, the critical buckling load has been clearly observed and
achieved at the highest AE signal energy (i.e. 116154 N). In practice another characteristic is used
more often that is called the amplitude distribution of pulses, shown in Fig.5, which is the AE
activity during the whole test, again, the highest AE amplitude between the critical buckling load and
pipe collapse.
 AE sensors

Fig.3. Typical libratory test of pipe buckling

Load
140000 Energy 80000

Critical
Load 70000
120000

60000
100000
50000

Energy (V .S)
80000
Load (N)

40000

2
60000 Zone A Zone B Zone C 30000

20000
40000
10000

20000
0

0 -10000
0 100 200 300 400 500 600
Time (S)

Fig.4. The load-curve and the energy evolutions of entire AE signal with time

The signal flow can be characterized (i.e. finger print) by the mean frequency of events, this
characteristics is directly related to the condition and type of deformation, as shown in Fig. 6. Fig.7
shows the AE hits and the amplitude for the hypothetical three zones. Zone A has the highest AE hits
with small amplitudes, while Zone B has highest amplitudes. Zone C where collapse takes place and
the hits increased although it has small amplitude. The origination, motion and growth of defects are
accompanied by changes in the microstructure and stress-strain state. Given this, the elastic energy is
redistributed that results in AE signal emission. In general, AE activity accompanied with plastic
deformation growth is high and referred to as primary AE activity, and it can be clearly observed
from Fig.8b (i.e. the highest AE signal amplitude and counts). The AE signal amplitude and counts is
relatively low, as shown in Figs.8a and 8c, for elastic and collapse modes referred to as secondary
AE activity.
Conclusions

Early diagnosis of pipelines enhances the pipeline dependability, decreases the loss of production
and increases the system safety. This paper focuses on early buckling detection of pipelines. During
pipeline deformation three different modes have been defined. The pipeline deformed in three
consecutive stages and can be clearly observed by the AE technique. The AE signals distributions are
sensitive to all stages of pipe deformation. Hence, buckling is at highest AE Activity and can be
monitored using AE technique. Studying the effects of different faults and internal/external
environment of pipelines on AE signal and the modeling are the main points for future work.

120

Collapse
Critical load
100

80
Amplitude(dB)

60

40

Threshold
20

0
-50 50 150 250 350 450 550 650
Time (S)

Fig. 5 The entire test amplitude of AE signal with time

.

Fig. 6a AE power spectrum for zone A Fig. 6b AE power spectrum for zone B

Fig. 6c AE power spectrum for zone C

Fig. 7a AE Hits (y) Vs Amplitude (x) for zone A Fig. 7b AE Hits (y) Vs Amplitude (x) for
zone B

Fig. 7c AE Hits (y) Vs Amplitude (x) for zone C

Fig. 8a. Secondary AE activity for zone A Fig. 8b. Primary AE activity for zone B

Fig. 8c. Secondary AE activity for zone C

References

1. K. M. Halford: Acoustic Emission in structural Health Monitoring, Key Engineering Materials,

Vol. 413-414, 2009, p 15-28.
2. S. Kyriakides, E. Corona: Offshore pipeline book, Vol. I Buckling and Collapse, 2007, Elsevier,
USA.
3. B. Guo, S. Song, J. Chacko, A.Ghalambor: Mechanics of offshore pipeline book, 2005, Elsevier,
USA.
4. M. Callegari, C.B. CariniI, S. LENCI, E. TORSELLETTI, L. VITALI: 16th AIMETA Congress
of Theoretical and Applied Mechanics, Ferrara, 2003, Italy.
5. J. Patton, P. Frank, and R. Clark, Issues of Fault Diagnosis for Dynamic System. Springer, 2000.
6. R. Isermann, Fault Diagnosis system: An introduction from fault detection to fault tolerant.
Springer, 2005
7. H. Guo, J. A. Crossman, Y. L. Murphey, and M. Coleman, “Automotive signal diagnostics using
Wavelets and machine learning,” IEEE Trans. Vehicular Technology, vol. 49, no. 5, pp. 1650-
1662, Sept. 2000.
8. S. Yoon, J.n Landry, N. Kettaneh, W. Pepe, and S. Wold, “Multivariate process monitoring and
early fault detection (MSPC) using PCA and PLS,” in Proc. Plant Automation and Decision
Support Conf., Hyatt Regency, San Antonio, Rexas, Sept. 21-24, 2003.
9. M. Shehadeh, J A Steel, and R L Reuben, "Acoustic emission source location for steel pipe and
pipeline applications: the role of arrival time estimation," J. Process Mechanical Engineering,
Vol. 220, 2006, p 121-13.
10. I. G. Scott, "Basic Acoustic Emission", Gordon and Breach, 1991, New York.