Enhanced Pipeline Monitoring With Fiber Optic Sensors · PTC 2011, Enhanced Pipeline Monitoring With Fiber Optic Sensors, J.Frings Page 3 of 12 The same standard optical fibers (typically

Page 1 of 12

Enhanced Pipeline Monitoring with Fiber Optic Sensors

Jochen Frings,

ILF Consulting Engineers, Germany

[email protected]

Abstract

Pipelines are efficient, highly reliable and safe means of transportation. However, although the

number of leaks could be reduced since the early 70’s of the last century due to improved design

and maintenance procedures as well as improved materials, leaks still appear. Most of these are

originated by external causes such as digging excavators or slope movements despite intensive

pipeline right of way surveillance by foot, car and out of the air.

These events are a clear sign for a monitoring gap. Due to the highly distributed nature of pipelines

with classical technology very high investments for point sensors including power and

communication facilities would have been necessary to allow complete coverage with real time

monitoring capabilities along the pipeline route.

The technical evolution of fiber optic sensing technologies allows closing large parts of this

monitoring gap. With a maximum active sensor length of up to 30 kilometer and local resolution

down to one meter these distributed fiber optic sensors are apt to detect various external leak

causes and actual leak locations by sensing temperature, strain and vibrations and even sound.

After a short introduction to the technical approaches for distributed fiber optic sensing, an

overview on pipeline related applications in the field of leakage detection, third party activity

monitoring, ground movement detection and integrity monitoring will be presented.

Finally insight to some engineering aspects of fiber optic sensing applications will be given.

1 INTRODUCTION

Pipelines are part of the backbone for modern communities’ lifestyle and are

absolutely indispensable for transportation of water, gas, oil and all kinds of

products.

Faults in these systems do not only result in service outages and financial losses but

bear the potential of spillages causing environmental pollution or even disastrous

accidents.

Due to this, governments, engineering companies and industry associations have

developed design, operation and maintenance standards for pipelines (e.g. TRFL

(19) in Germany) based on which the number of leaks could be reduced drastically

6th Pipeline Technology Conference 2011

PTC 2011, Enhanced Pipeline Monitoring With Fiber Optic Sensors, J.Frings

Page 2 of 12

since the 60’s and early 70’s of the last century. As a result pipelines today are

highly reliable and safe means of transportation.

However leaks and disastrous events still appear. Statistically about 50% of all leaks

were caused by third party activities according to the European gas and oil

transportation industries’ recent yearly reports (1) and (2) followed by construction /

material failure, corrosion and ground movement.

In conclusion most leaks could be avoided, if the third party activities could be

detected in time to intervene before the actual leak appears. Despite often massive

efforts to monitor the pipeline’s right of way by walking, driving and flying along the

right of way complete coverage of the pipeline right of way with real time monitoring

equipment has been impossible for a long time. This is because of the prohibitive

cost for installation, operation and maintenance of literally thousands of point

sensors like vibration or motion detectors and video cameras which would have to be

installed including power supply and communication facilities along the pipeline route

with classic technology.

These and many more issues can be resolved with the help of distributed fiber optic

sensor cables, which are sensible over their complete length up to the range of 30

kilometers and which are able to detect temperatures, strain, vibration and sound

with high location accuracy and absolute resolution. Since fiber optic cables are

insensitive to EMC, designed for harsh environments and independence of additional

field power supply or additional communication installations they are optimally suited

for highly distributed pipeline monitoring applications.

Based on these advantages distributed fiber optic sensing has been applied

successfully in a variety of applications already and thus can be considered to be a

field proven technology.

However, the technical evolution of distributed fiber optic sensing is still ongoing.

New products with improved sensing technology and new signal analysis algorithms

allow in combination with permanently increasing computing performance higher

sensibility, detection speed and thus new applications.

In the following a short introduction to the distributed sensing technology will be

provided, before an overview on several applications of the technology is presented

and some engineering aspects are discussed.

2 FIBER OPTIC CABLES ARE DISTRIBUTED SENSORS

Fiber optic cables are standard equipment for transmission of voice, video and other

data and are frequently installed along pipelines and often used to enable

communication between and remote control of individual stations of the system.


Page 3 of 12

The same standard optical fibers (typically single mode) are suitable to measure

several physical effects with high absolute and local accuracy.

2.1 Scattering

Fiber optic cables are typically designed such that scattering effects are minimized to

maximize transmission distance and data rate. However, it could be shown that

some scattering effects of injected laser light depend on the fiber optic cable ambient

conditions (temperature [T], strain [ε]) (3) as shown in Figure 1.

1) Rayleigh scattering:

Elastic scattering of light based on density and composition fluctuations within

the cable material. Scattering itself is not sensible to ambient conditions, but

used for fiber integrity sensing and interferometric sensing applications.

2) Raman scattering:

Inelastic scattering of photons due to molecular vibration within the fiber

material. The magnitude of the molecular vibration and the scattered signal is

influenced by the environmental temperature.

3) Brillouin scattering:

Based on time dependent density variations of the fiber material. The

wavelength of the scattered signal is depending on the ambient temperature

and the strain or vibration of the optical fiber.

To measure the Raman and Brillouin scattering effects advanced and specialized

optical time domain reflectometers (OTDR) are applied. These measurement

devices send short laser pulses into the fiber and analyze the time-distance related

reflection/scattering signals with regards to frequency and amplitude of the desired

scattering effect. In consequence it becomes possible to measure strain and

temperature along the fiber, as shown in Figure 2.

Figure 1: Scattering effects in fibre optic cables caused by temperature [T] or strain [ε], (3)


Page 4 of 12

Figure 2: Temperature and strain profile along optical fiber (4)

Multiple products are available on the market. Typical temperature resolution is in

the range of 0,10K, while strain resolution can be in the area of 20με, both with a

local resolution in the range of 1m, while the absolute ranges largely depend on the

cable construction. In all cases improvement of resolution corresponds to increased

time for measurement and hence both have to be adapted application specific.

Maximum sensor lengths for single mode fiber based Raman and Brillouin systems

typically are in the range of 20 to 30 kilometers, while multi mode fiber based Raman

systems typically have a reach of up to 8 kilometers

While for Brillouin scattering temperature measurement can be implemented with

standard telecom cable constructions (loose tube) which decouple the fiber from

external strain as much as possible, for strain measurements temperature

compensation has to be implemented. This can be achieved by using two fibers in

parallel: One coupled to e.g. the pipeline to measure strain changes; the other fiber

installed nearby and strain relieved to measure the temperature.

Based on strain sensing various vendors offer also detection of vibration (changing

strain).

3.2. Interferometers

For more than a century interferometers have been a well known solution to detect

very small changes of distances and the interferometer principles have been

successfully applied to fiber optic measurement configurations since the early days

of fiber optics.

With the advent of modern fiber optic components and using the constantly

improving computing performance for improved measurement signal analysis

research projects and several vendors developed configurations that can work as

distributed microphones/hydrophones with high sensibility and good location

accuracy.

For example a fiber optic configuration according to the Mach-Zehnder

Interferometer can detect sound waves or vibrations by analyzing signal interference


Page 5 of 12

between two separate sensor fibers (5) and thus can act as a hydrophone (see

Figure 3).

Figure 3: System block diagram presenting fibre optic interferometer

The interferometer can manage a sensing distance up to 40 km and its sensibility is

at least up to 3 m radius around the fiber optic cable.

According to (18) another system sends two accurately timed pulses and analyses

the interference of the Rayleigh scattering signals and thus is able to detect sounds.

Sensitive cable length of up to 50 kilometers is claimed to be possible.

Intelligent signal analysis is necessary to identify and separate farming machines,

underground construction works, digging, tapping and other events which are subject

of a specific training phase.

Another approach to overcome the weak locating capabilities of standard

interferometers is described in (17) were an interferometer is combined with Brillouin

instrument. While the interferometer allows precise analysis of the event, the

Brillouin instrument allows to locate the event precisely.

3 PIPELINE APPLICATIONS FOR DISTRIBUTED FIBER OPTIC SENSORS

Based on the above it becomes clear that distributed fiber optic sensors are almost

ideal for many types of pipeline monitoring applications and several of these

applications have been implemented during recent years all over the industry.

Unfortunately it is not possible to give a complete overview and thus the following

should be considered as examples only.

3.1 Leak Detection

Loss of transported medium due to pipeline leaks typically results into one or more of

the following detectable effects:

1. Local cooling due to Joule-Thomson effect (high pressure gas pipelines)


Page 6 of 12

2. Soil temperature change due to temperature difference between soil and

emanated fluids and due to evaporation effects.

3. Especially in high pressure applications the emanating medium generates

detectable sounds.

Based on Raman or Brillouin scattering effects the temperature changes can be

detected, if the medium temperature is different from the soil temperature. Hence

distributed temperature sensing has been reported to be applied for natural gas,

brine, phenol, sulfur, LNG, crude oil and other mediums and allows detecting even

very small leaks (for example (3),(6),(7)). Compared to the conventional intrinsic

Pipeline Monitoring methods this approach has the additional advantage to be

completely independent of any process conditions.

Even the periodical opening and closing of small leaks in gas pipelines due to

freezing effects can be identified with modern signal analysis methods.

For offshore pipelines the application of leak sound detection is reported in (6) based

on a Brillouin strain measurement system.

Distributed temperature sensing is used in all cases to improve the performance of

computational monitoring systems. This is not only due to the fact that distributed

temperature sensing has been installed at pipelines which already had

computational monitoring systems. Although distributed temperature sensing is a

well proven technology that has shown to be able to detect very small leaks in short

time, it is very hard to calculate the minimum detectable leak size or to guarantee a

maximum detection time which in many cases are necessary to receive pipeline

operation licenses.

3.2 Ground Movement Detection and Structural Health Monitoring

Geohazards like earthquakes, landslides and surface subsidence result into ground

movement and thus put additional stress on the pipelines, tunnels and other

underground infrastructures. Distributed fiber optic strain sensors have been applied

in two ways to identify the endangering ground movements:

Strain sensing fibers have been attached directly to the pipeline walls to

measure the walls’ strain changes and to conclude on the consequential

movements and deformations (8), (9), (10). In case a sudden strain increase

is detected, the pipe internal pressure can be reduced to reduce the total

stress and such to reduce the risk and/or effects of a leak.

Strain sensing fiber optic cables are installed in parallel and close to the

infrastructure (11). This method allows covering large route sections due to

simplified cable installation method.


Page 7 of 12

While pipeline sections know to have an increased risk of ground movements have

been monitored with point sensors already during recent years, distributed fiber optic

monitoring can be installed along longer stretches of the pipeline. Thus also strain

changes due to planned ground works (e.g. trenchless installation of crossing

pipelines and cables) can be monitored.

3.3 Third Party Activities

The majority of all reported pipeline leak incidents has been caused by third party

activities including construction and agricultural works, illegal tapping and intentional

damaging. By applying distributed strain sensing (e.g. (6),(7)) or interferometer

based hydrophones (e.g.(5), (18), (21)) along the pipeline or other buried

infrastructure, it becomes possible to detect approaching heavy earth working

machines, actual digging (manual or machine supported), metallic contact with the

pipeline and other sound and vibration signals. For example the system installed

along the BTC pipeline (21) could detect a third party activity including manual

digging. Because knowing the exact location of the event, immediate response could

prevent illegal tapping and consequential environmental and financial damages even

during test operation.

3.4 Fire Detection

Distributed temperature sensing with fiber optic cables is used as heat detector for

fire detection (13) in tunnels. Because the cable is sensitive along its complete

length and because the temperature can be detected within a wide range it becomes

possible to determine the fire position and development very detailed. In

consequence ventilation and other fire fighting measurements can be coordinated

efficiently.

3.5 Power Cable and Transformer Monitoring

Power cable isolation (XLPE) typically is rated for an operating temperature of 90oC.

Especially in power cable tunnels, when cables are bundled and mounted to cable

trays it is possible that this temperature is exceeded in high load situations. To this

end distributed fibre optic temperature sensors have been installed inside the cable

isolation (e.g. (14), (15)), so that a direct temperature assessment becomes possible

with high local resolution and thus countermeasures can be initiated easily.


Page 8 of 12

3.6 Status Monitoring of Water Mains

Pre-stressed concrete cylinder pipes (PCCP) are widely used in water mains.

However in several cases PCCP did not show to be as durable as expected, which

resulted in several water main ruptures with considerable damages and a large

number of smaller defects resulting into water losses. It has been shown (16) that

the stability of PCCP correlates with the number of broken wires inside the pipe. As

the breaking wire emanates a special sound this can be detected by an acoustic

sensitive fibre optic sensor installed inside the pipeline. Based on the pipe book a

software package can than count the number of broken wires and can issue an

alarm in case the total number of broken wires or the number of wires broken within

a certain time period exceeds a limit.

Since the wire break event is only very short the pulsed Brillouin based acoustic

detectors often do not detect sufficient information for clear identification of the wire

break event. On the other side interferometers such as Sagnac- or Michelson

interferometer analyse the signal continuously and such receive all available

information about the wire break sound - but they are weak in locating the signal.

Thus a combined interferometer and Brillouin detector is described in (17) and has

been applied to several water mains in the United States of America.

3.7 Pig Position Detection

Systems designed to detect sounds of third party activities are also apt to detect the

sounds created by pigs according to e.g. (22).

4 SOME ENGINEERING ASPECTS OF DISTRIBUTED SENSING

Since standard telecom cables are optimized for long distance signal transmission

with protection of all fibers against strain (e.g. jelly field tubes) or against

environmental influences (e.g. multiple sheath layers), for measurement purposes

these are not always the best solution. For example strain relieved fibers are not

able to measure strain and thick multilayer sheaths increase measurement delay for

temperature. In consequence various measurement cable constructions are

available in the market. For example there are cables with specialized profiles to

allow gluing the cable to any type of structure such as steel bridges or concrete

tunnels. Other cables are specialized for high temperature measurements and more.


Page 9 of 12

Figure 4: Application specific sensor location along pipeline

Of course the sensor position decides about the correctness of the measurement.

Typical locations of the sensor cables relative to for example a pipeline are indicated

in Figure 4. For liquid leak detection the sensor should be installed below the pipe

and on top for leak detection of most gaseous mediums, while pipeline strain

obviously can be measured only, if the fiber is directly bonded to the pipeline. Typical

gas and telecom cable positions are suitable for third party interference as well as for

ground movement detection. For different applications similar considerations apply.

To achieve a clear leak detection signal at least 2 to 5oK temperature difference

between the fluid in the pipeline and the soil temperature around the sensor are

required, although datasheets state to have accuracy down to 0,1oK. It has to be

proven that this minimum temperature difference can be guaranteed throughout the

whole year despite all seasonal temperature changes and for all pipeline operation

states to avoid blind periods.

While this is normally no problem for e.g. high pressure gas pipelines (due to the

strong Joule-Thomson effect), detailed seasonal analysis of soil temperature profiles

is necessary for typical crude and product pipelines, even in case the transported

medium has a much higher temperature than the not influenced soil temperature.

For simplification only a single season scenario (winter, operating pipeline) is shown

in the figure below. It shows the results of our analysis for a crude oil pipeline under

the assumption of 33oC medium temperature and -10oC air temperature after several

weeks of continuous operation and allows estimation of the minimum necessary

distance between the pipeline wall and the sensor cable.


Page 10 of 12

Figure 5: Soil temperature profile around oil pipeline (winter)

For the required temperature difference a distance of more than half a meter from

the bottom of the pipe has to be obeyed.

The effectiveness of most distributed fiber optic sensor systems is based not only on

the sensor design and installation but also on the signal analysis program design.

These pattern analysis mechanisms have to be trained carefully to ensure maximum

detection rates with minimum false alarm rates.

5 CONCLUSION

Distributed fibre optic sensing is a field proven technology for online monitoring of

temperature, strain, vibration and sound over long distances with high local

resolution that is apt to improve pipeline integrity, safety and security considerably.

Many different applications have already been implemented in the industry and

constantly ideas for new applications are being created.

Although the basic principles for distributed fibre optic sensing have been known for

more than two decades, the evolution of these technologies does not yet come to an

end and especially during recent years improvements could be made by using the

constantly increased computing power for improved analysis of the measurements.

Distributed fibre optic sensing systems installed along pipelines must be considered

as integral parts of pipeline instrumentation and application and installation of these

systems have to be properly engineered to get the desired results.


Page 11 of 12

6 BIBLIOGRAPHY

1. CONCAWE. Performance of European Cross Country Oil Pipelines, Statisctical

Summary of reported spillages. s.l. : www.concawe.org.

2. EGIG. EGIG Report. s.l. : www.egig.nl.

3. Pipeline Leakage Detection and Localization Using Distributed Fiber Optic

Sensing. D. Inaudi, B. Glisic, A. Figini, R. Walder. RIO Pipeline Conference, 2nd to

4th October 2007.

4. Waldner, R. Pipeline Leakage Detection and Localization Using distributed Fiber

Optic Sensing. Webinar. s.l. : Smartec, 2009.

5. Future Fiber Technologies. Secure Link: Installation Overview. s.l. :

www.fftsecurity.com, 2009.

6. Sensornet. Digital Pipeline Integrity Monitoring System - Application Guide.

7. Schlumberger. Integrated Pipeline Monitoring – Integrity. 2008.

8. Long-range Pipeline Monitoring by Distributed Fiber Optic Sensing. D. Inaudi, B.

Glisic. Calgary, Canada : s.n., 2006. 6th International Pipeline Conference.

9. Earthquake Detection and Safety System for Oil Pipeline. L. Griesser, M. Wieland,

R. Walder. December 2004, Pipeline & Gas Journal.

10. Daniele, Inaudi and Branko, Glisic. Overview of Fibre Optic Sensing Applications

to Structural Health Monitoring. 13th FIG Symposium on Deformation Measurement

and Analysis. Lisboa : s.n., 2008.

11. Omnisens. Pipeline Integrity Monitoring - Transandean Route, Case Study.

12. Soga, Kenichi. Monitoring and Assessment of Civil Engineering Infrastructure. A

Sensory World: Novel Sensor Technologies and Applications. Cambridge : University

of Cambridge, 2007.

13. Liu, Z. and Kim, A.K. Review of Recent Development in Fire Detection

Technologies. National Research Council Canada. 2003. NRCC-45699.

14. Sensornet. Power Case Study - Power Cable Tunnel Monitoring. 2009.

15. Glombitza, U. and Hoff, H. Fibre Optic Radar System for Fire Detection in Cable

Trays. 13th International Conference on Automatic Fire Detection. Duisburg : s.n.,

2004.

16. An Integrated Dynamic Approach to PCCP Integrity Management. J. Elliot, J.

Stieb, M. Holley. 2006. Pipelines - Service to the Owner.

17. Paulson, Peter O. Fiber Optic Sensor Method and Apparatus. US Patent

7,564,540 B2 United States, 21 Jul. 2009.

18. A. J. Rogers, S. E. Kanellopoulos, S. V. Shatalin. Detecting a Disturbance in the

Propagation of Light in an Optical Waveguide. US Patent 7,872,736 B2, United

States, 18. Jan. 2011.

19. TRFL- Technische Regel für Rohrfernleitungsanlagen nach §9 Absatz 5 der

Rohrfernleitungsverordnung. 2010.


Page 12 of 12

20. S. Scott, M. Barufett. Worldwide Assessment of Industry Leak Detection

Capabilities for Single and Multiphase Pipelines. Texas A&M University. s.l. : OTRC

Library Number: 8/03A120, 2003.

21. QinetiQ. QuinetiQ Optasense is selected to protect the BP-led Baku Tbilisi

Ceyhan Pipeline, press release 29. Sept. 2010, s.l. www.qinetiq.com

22. Optasense. Pipeline and Security Monitoring. s.l. www.quinetiq.com, 2010

Documents

Enhanced Pipeline Monitoring With Fiber Optic Sensors · PTC 2011, Enhanced Pipeline Monitoring With Fiber Optic Sensors, J.Frings Page 3 of 12 The same standard optical fibers (typically