acoustic reflectometry for blockage detection in … · 2017. 10. 20. · acoustic reflectometry...
TRANSCRIPT
ACOUSTIC REFLECTOMETRY FOR BLOCKAGE DETECTION IN PIPELINE
JOSÉ L. A,VIDAL
Petrobrás Research Center - CENPES/PDEP/TOOL
Av.Horácio de Macedo 950- Cidade Universitária – 21941-915 -Rio de Janeiro-RJ
E-mail:[email protected]
LUCIANA L,SILVA
Ocean Engineering Department, Federal University of Rio de Janeiro, COPPE/UFRJ
Cidade Universitária – Bloco C - Ilha do Fundão— 21945-970- Rio de Janeiro-RJ
E-mail: [email protected]
Abstract-- Flow assurance is an important aspect of offshore, particularly deepwater pipeline design and operation, since one of
the critical issues is the eventual initiation and growth of hydrate or paraffin blockages under certain conditions. Ideally, operators
would benefit from online information regarding position and extent of an eventual blockage in a pipeline.The aim of this work is
to apply acoustic technology to design and make a prototype that can be used in a pipe to efficiently identify and measure
blockages. The technique uses a short duration sound pulse that is injected into the pipe. When the acoustic pulse encounters an
impedance discontinuity, a portion is reflected back towards the acoustic source and microphones or hydrophones. Analysis of the
measured signal reflections can provide valuable data related to location and size of the blockages. The experiments were
numerically reproduced with good correlation proving the potential of the technique.
Keywords—Blockage Detection, Pipeline, Flow Assurance, Hydrate, Acoustic Pulse, Reflectometry, Acoustic Technic
1 Introduction
Hydrate, paraffin wax and incrustation deposition in
oil and gas transport pipelines are an important challenge
for the development of deepwater environments. These
types of deposition always occurs along the inner wall of
the pipeline, which reduces the internal diameter and
eventually results in the blockage (N.X. Thanh,1999;
K.A. Papadopoulou,2008). During the service life,
pipelines should not fail because such failures could lead
to environmental, human, and economic costs. In this
context, flow assurance is critical for deepwater projects
in order to avoid pipeline blockages. Therefore detection
and localization of the formation of blockages at its early
stage is very important for remedial actions before
catastrophic effects occur. In recent years, detection of
blockage in pipelines has been an attractive area of
research and several methods have been proposed (K.A.
Papadopoulou,2008; O.A. Olugboji,2011). The
conventional methods for blockage detection include
flow pressure monitoring detection, thermographic,
radiographic methods, and others (K.A.
Papadopoulou,2008; O.A. Olugboji,2011; X. Chen,2007;
M.S. Ghidaoui,2005). Flow pressure monitoring
detection uses quick-acting valve to generate a water
hammer and the the reflections produced by any
blockage can be recorded. This method
has its disadvantages, like the high transient pressure
may cause considerable damage to the pipeline.
Thermographic and radiographic methods typically
require the use of a Remotely Operated Vehicle (ROV)
and needs to be near the location and these methods are
expensive and time-consuming operations.
The acoustic pulse reflectometry, a non-invasive
measurement technique and remotely operated, has been
shown to be an appropriate tool to solve this problem.
The technique is operated by injecting a small acoustic
signal into the fluid within the pipeline to remotely
detect blockages, and several components, such as
valves, flanges and T-sections located at long distances,
based on the analysis of reflected acoustic signals, as
these reflections carry sufficient information to identify
blockages and components (K.A. Papadopoulou,2008;
B.J. Forbes, 2003; A. Tolstoy, 2009; G.O. Albors,2007).
When the acoustic pulse encounters an impedance
discontinuity, a portion of the pulse is transmitted past
the discontinuity and a portion is reflected back towards
the acoustic source. Acoustic reflections are then
measured and analyzed to estimate the location of
impedance discontinuities within the systems under
investigation, also a sound wave is travelling inside a
pipe and the wavelength is large compared with the
diameter of the tube, wave propagation is very nearly
one-dimensional. That is, waves inside the pipe can be
considered to be plane waves. These waves can travel for
large distances with low attenuation (J.M. Hovem,2007).
Anais do XX Congresso Brasileiro de Automática Belo Horizonte, MG, 20 a 24 de Setembro de 2014
3916
This paper provides details of the results that have been
obtained from experiments tests and numerical model
that have been designed to determine the capability of
using the acoustic technique to detect blockages in
pipeline systems.
2 Acoustic Technique Background
The three dimensional acoustic wave equation
expressed in cylindrical coordinates is (P.M. Morse,1968;
N.R. Harris,1997):
∇2𝑃 =1
𝑟
𝜕
𝜕𝑟(𝑟
𝜕𝑃
𝜕𝑟) +
1
𝑟2(𝜕2𝑃
𝜕𝜃2) +
𝜕2𝑃
𝜕𝑧2=
1
𝑐2𝜕2𝑃
𝜕𝑡2 (1)
where P is acoustic pressure, t is time, c is the speed of
sound and r, and z are respectively radial, angular and
longitudinal variables. Assuming a variable separable
solution, P(r, , z, t)=R(r)()Z(z)T(t). The acoustic wave
equation for cylindrical symmetry can be solved to give a
general solution:
P(𝑟, 𝑧, 𝑡) = 𝐴 J (𝑘 𝑟)e
( ) (2)
where J (𝑘𝑟) is a Bessel function of zero order, kb is a
constant and 𝑘 is the exponential is a propagation
constant (wave number). For a pipe of radius a, a
boundary condition can be applied at r=a, In the case of
rigid wall the condition to be applied is that the particle
velocity normal to the wall is null, (𝜕𝑃 𝜕𝑟) = 0⁄ . When
analysing sound propagation in pipes it can often be
assumed that the wave front is uniform throughout the
cross-section. This assumption is limited to those cases
where the frequency f is below the cut-on frequency for
the first non-plane mode. For circular cross-sections this
limits given by the first zero of J1′ , the derivative of the
Bessel function of the first kind, as given by
𝑓 <1.84𝑐
2𝜋𝑟 (3)
For a pipe with internal diameter of 4'' filled the air, the
cut-off frequency is 1.96kHz.
3 Methodology
To test the acoustic system to locate and measure
blockages, tests were carried out in a small-scale pipeline
located in the Subsea Technology Laboratory at the
Federal University of Rio de Janeiro. An experimental
setup with a pipe of 4” internal diameter and length of
approximately 95m was constructed and composed of
several spools sections, each 500mm long, located in
different positions of pipeline to insert blockages, see
Figures 1 and 2. The tests can be performed in air and/or
water using different sizes of blockages and in different
positions. Paraffin blockages were made to simulate
different stages of obstruction, which covered
approximately 10, 25, 50 and 75% of the pipe's area, as
showed in Figure 3. For each were made three different
lengths 100mm 250mm and 500mm. The signal
generated by an acoustic pulse generator was passed
through an amplifier to drive a loudspeaker located at
one end of the pipeline, which transmitted the signal into
the fluid in the pipe. One microphone (BSWA model
MPA416) was mounted at a distance of 6.4 m from the
acoustic emitter. This microphone captures the incident
pulses and reflections returning from the blockages
under test, and the measurements were recorded using a
fit-for-purpose data acquisition system with sampling
rate of 100kS/s per channel. The ideal frequency content
of the acoustic signal injected into the pipe is the subject
of current research. However, in this work a short period
square pulse, with a width of 1ms, has been used.
Figure 1: Pipeline Layout (pipe of 4” internal diameter and
length of 100m). Above is schematic view and below is frontal
view with dimensions.
Figure 2: Zoom in the region of spool for insertion of
blockages.
Anais do XX Congresso Brasileiro de Automática Belo Horizonte, MG, 20 a 24 de Setembro de 2014
3917
Figure 3: Different sizes of paraffin blockages with 10, 25, 50
and 75% of the pipe's area. The percentage in this figure refers
to the percentage of the pipeline area that will be blocked.
In parallel with the experimental program, finite
element model using ABAQUS (version 6.9-1) were
developed using three-dimensional finite element model
that simulate acoustic wave propagation. The internal
fluid was discretized with 3-D acoustic elements
(AC3D4), see Figure 4. This element is defined by four
nodes with one degree of freedom at each node: pressure.
The reference pressure is used to calculate the sound
pressure level. The acoustic pressure in the fluid is
determined by the wave equation
1
𝑐2𝜕2𝑃
𝜕𝑡2− ∇2𝑃 = 0 (4)
P is the sound pressure, t is the time and c the sound
speed. There are some assumptions: the fluid is
compressible, the fluid is inviscid (no viscous
dissipation) and the density is uniform throughout the
fluid. The appropriate size of the elements is chosen in a
manner so that the propagating waves are spatially
resolved. In (F. Mosera,1999 ), it is recommended that
more than 10 nodes per wavelength be used, can be
expressed as
𝑙𝑒𝑙𝑒𝑚 =𝜆𝑚 𝑛
10 (5)
where 𝑙𝑒𝑙𝑒𝑚 is the element length and 𝜆𝑚 𝑛 is the shortest
wavelength of interest. The Figure 4 shows, for better
visualization, a zoom of the fluid mesh. The blockages
were simulated using a reduction of the cross-
sectional area of the pipe. We assume only plane waves
propagate through the fluid and in the case the outer
boundary of the internal fluid is represented by a non-
reflecting acoustic impedance. The acoustic pressure–
time input (from an acoustic emitter) used in the
numerical analysis was obtained from experimental tests.
Figure 4: Finite element mesh with zoom in the region of the
blockage.
4 Experimental Results
The laboratory results are presented for pipeline filled
with air. A typical normalized intensity plot with distance
for a pipe with no blockage is shown in Figure 5 (unit of
pressure Pa). It shows a significant reflected intensity
caused by the end of at 88 m from microphone, with no
significant reflections along the rest of the pipe. The
Figure 6 shows the results for the same pipe but with a
blockage of 75% of area and length of 500mm that is
placed in different positions on the pipe, 12.5m, 31.5m,
41.0m, 54.5m, 83.0m from the microphone. In all cases,
it is possible to locate the blockage and the farther the
blockage, the lower the intensity of the reflected signal
due to attenuation factor, which quantifies the rate of
exponential decay that occurs with the propagation. It is
also possible to see the multiple reflections from tests
performed on blockages closer to the microphone.
10% 25% 50% 75%
Anais do XX Congresso Brasileiro de Automática Belo Horizonte, MG, 20 a 24 de Setembro de 2014
3918
Figure 5: Response with the pipe without blockages.
Figure 6: Response of the pipe with blockage of 75% of
area and length of 500mm placed in different positions
on the pipe, 12.5m, 31.5m, 41.0m, 54.5m, and 83.0m
from the microphone.
Tests were performed with all the blockages placed in the
same position in the pipe 12.5m from microphone.
Figure 7(a) shows the normalized pressure results due to
reflection from blockages. All obstructions have the
same length of 500mm but simulate different stages of
blockage (10, 25, 50 and 75% of the area’s pipe). When
comparing results of different blockages, while the effect
of each obstruction is visible, the differences between
obstructions are only very slightly visible. Figure 7(b)
shows the results in the frequency domain after
performing the FFT of the signals. The differentiation
between them is more pronounced when analyzes the
signals in the frequency domain. The frequency contents
of all is approximately the same between 50Hz to 750
Hz, it shows that as the size of the blockage increases,
the amplitude of the reflected signal increases. This is to
be expected as the length of the blockage will also be
increasing, as can be seen in Figure 8. The blockages
have the same area (10%) and different lengths of
100mm, 250mm and 500mm.
Figure 7: Difference signal for varying blockages area. (a)
Response in time space and (b) frequency analysis of
blockage signals.
Anais do XX Congresso Brasileiro de Automática Belo Horizonte, MG, 20 a 24 de Setembro de 2014
3919
Figure 8: Difference signal for varying blockages length. (a)
Response in time space and (b) frequency analysis of blockage
signals.
The laboratory experiments show that the acoustic
technique is a promising method for detecting pipe
different stages of blockages. The waves propagate at the
local speed of sound and the difference in arrival time
between transmitted waves and reflected waves is used
to calculate the distance from the microphone to the
blockage. The localization errors in all experiments were
between 0.15 and 1.05%. The uncertainty in sound of
speed measurement of the pulse signal can cause the
uncertainty of the location of blockage.
5 Numerical Results
Figure 9 illustrates a sequence of acoustic wave
propagation for a pipe with blockage (75% and 500mm).
The Figure 9(a) shows the behavior of acoustic wave
propagation before it reaches the blockage and the Figure
9(b) shows the result of the wave after passing through
the blockage. It is possible to see that a portion of the
incident wave is transmitted and part is reflected.
Figure 9: Sequence of acoustic wave propagation for a pipe
with blockage (75% and 500mm): (a) acoustic wave before it
reaches the blockage and (b) transmitted and reflected wave
after passing through the blockage.
Therefore, a comparison of numerical results obtained
with experimental results was conducted. Figure 10
shows some results of the comparison between the
numerical and experimental results for the cases where
the blockage is located 12.5m from the microphone and
the sizes of obstructions vary as: (a) 10% and 250mm,
(b) 25% and 500mm, (c) 50% and 250mm and (d) 75%
and 100mm. Results show that the numerical model can
be used to adequately capture the experimental observed
behavior. The numerical model can be used to make
predictions of the behavior of acoustic propagation
different pipelines systems as well as different types of
fluid.
(a)
(b)
Anais do XX Congresso Brasileiro de Automática Belo Horizonte, MG, 20 a 24 de Setembro de 2014
3920
Figure 10: Comparison between the numerical and
experimental results for the cases where the blockage is located
12.5m from the microphone and the sizes of obstructions vary
as: (a) 10% and 250mm, (b) 25% and 500mm, (c) 50% and
250mm and (d) 75% and 100mm.
6 Prototype for Field Testing
After carried out the study to verify the feasibility of
the acoustic technique for detecting blockages, a
prototype was developed for oil and gas industry. High
power acoustic emitters can be mounted as shown in the
Figure 12. This technology will normally be required to
survey long lengths of pipeline and therefore low
frequency signals will be required. The specific signal
that is used should be selected according to the pipe
configuration with a compromise on survey length and
accuracy.
Figure 12: Acoustic emitters mounting arrangement.
The Subsea Technology Laboratory - Federal
University of Rio de Janeiro and Petrobrás jointly
conducted a field test using the acoustic monitoring
system. The tested water-filled pipeline
of 12” internal diameter and length of approximately 5
km. The tests were performed with flow in the pipeline
and at pressurized condition (150 psig). Figure 1 shows a
map showing pipeline system in red color.
Figure 12: Tested pipeline system.
Anais do XX Congresso Brasileiro de Automática Belo Horizonte, MG, 20 a 24 de Setembro de 2014
3921
During these tests, the gas gun was located
approximately 240m of water collection station. Figure
13 shows how the acoustic system was connected to the
pipeline.
Figure 12: Acoustic system connection.
The Figure 13 shows the recorded data assuming a
speed of sound of approximately 1,450 m/s. The incident
signal has 25 ms of duration (highlighted area 1). In the
range between 25 ms and 67 ms it is possible to identify
multiple reflections from the interface between the
mechanical assembly of acoustic system to the pipe
(highlighted area 2). Between 67 ms and 336 ms there is
no significant reflections, just the background noise from
flow (highlighted area 3). After 336 ms it is possible to
see a significant difference in the reflected signals
indicating the existence of any obstacle. The distance
between the test point and the water collector station is
the same seen by the acoustic system, highlighted signal
showed in area 4 (Figure 13).
Figure 13: Field test response.
7 Conclusions
This paper deals with the analysis of acoustic
technique that can be used in a pipeline to efficiently
identify and measure blockages during operation on line
(no shutdown). Experimental tests and numerical
simulations are carried out considering acoustic
propagation behaviors in pipeline showing their
influence on system dynamics. The laboratory and field
results show that the technique can be used for gas pipes
and water pipes with large diameter. Results show that
the system enables remote and continuous monitoring of
pipeline and provides an indication of location and the
relative size of blockage. Tests indicate that the proposed
approach is capable of identifying blockages as small as
10% of the pipe area. It is possible that smaller
blockages can be identified, but tests were not performed
with blockages less than 10% of the area's pipe, but the
numerical model shows that it is possible to identify
obstructions with 1% of pipe area. The prototype for oil
and gas industry showed good results for water-filled
pipeline and it would appear to be a suitable to perform
detections.
Acknowledgments
The authors would like to thank the financial support
from CENPES/PETROBRAS.
References
N.X. Thanh, M. Hsieh, R.P. Philp, “Waxes and
asphaltenes in crude oils,” Organic Geoche- mistry,
Vol. 30, pp. 119-132, 1999.
K.A. Papadopoulou, M.N. Shamout, B. Lennox, D.
Mackay, A.R. Taylor, J.T. Turner, and X. Wang, "An
evaluation of acoustic reflectometry for leakage and
blockage detection", Proc. IMechE Vol. 222 Part C: J.
Mechanical Engineering Science, 2008.
O.A. Olugboji, “Development of an Impact Monitoring
System for Petroleum Pipelines”, AU J.T. Vol. 15(2),
pp. 115-120, 2011.
X. Chen, S.P.E. Ying Tsang, and H.Q. Zhang, and T.X.
Chen, “Pressure-Wave Propagation Technique for
Blockage Detection in Subsea Flowlines”, SPE
Annual Technical Conference and Exhibition,
Anaheim, California, U.S.A, 2007.
M.S. Ghidaoui, M. Zhao, D.A. McInnis and D.H.
Axworthy, “A Review of Water Hammer Theory and
Practice”, Applied Mechanics Reviews, Vol. 58,
2005.
Anais do XX Congresso Brasileiro de Automática Belo Horizonte, MG, 20 a 24 de Setembro de 2014
3922
B.J. Forbes, D.B. Sharp, J.A. Kemp, A. Li, “Singular
System Methods in Acoustic Pulse Reflectometry”,
Acta Acustica United with Acustica Vol. 89, pp. 743
– 753, 2003.
A. Tolstoy, K.V. Horoshenkov, M.T. Bin Ali, “Detecting
pipe changes via acoustic matched field processing”,
Applied Acoustics 70 pp. 695–702, 2009.
G.O. Albors, A.M. Kyle, G.R. Wodicka, Fellow, E.J.
Juan, “Computer Simulation Tool for Predicting
Sound Propagation in Air-Filled Tubes with Acoustic
Impedance Discontinuities”, Proceedings of the 29th
Annual International Conference of the IEEE EMBS
Cité Internationale, Lyon, France, pp. 23-26, 2007.
J.M. Hovem, “Underwater acoustics: Propagation,
devices and systems”, Journal of Electroceramics,
Vol. 19 (4), pp. 339– 347, 2007.
P.M. Morse, “Theoretical Acoustics” McGraw-Hill, Inc.,
Printed in the United States of America, 1968.
N.R. Harris, M. Hill, J.D. Turner, “Practical aspects of
modal propagation through water filled pipes”,
Ultrasonics Vol. 35, pp. 73-78, 1997.
F. Mosera, L.J. Jacobs, J. Qub, “Modeling elastic wave
propagation in waveguides with the finite element
method”, NDT&E International, Vol. 32, pp. 225–
234, 1999.
Anais do XX Congresso Brasileiro de Automática Belo Horizonte, MG, 20 a 24 de Setembro de 2014
3923