Distributed Temperature Sensing:

Review of Technology and Applications

Abhisek Ukil, Senior Member, IEEE, Hubert Braendle, and Peter Krippner

Abstract—Distributed temperature sensors (DTS) measure tem- applications in the temperature and strain detection in cables.
peratures by means of optical fibers. Those optoelectronic devices Various power systems applications are discussed in Section IV.
provide a continuous profile of the temperature distribution along Section V reviews different upcoming applications. Major man-
the cable. Initiated in the 1980s, DTS systems have undergone sig- ufacturers of the DTS systems are mentioned in Section VI.
nificant improvements in the technology and the application sce-
nario over the last decades. The main measuring principles are Section VII discusses about the current and future trends in
based on detecting the back-scattering of light, e.g., detecting via DTS, followed by conclusions in Section VIII.
Rayleigh, Raman, and Brillouin principles. The application do-
mains span from traditional applications in the distributed tem-
perature or strain sensing in the cables, to the latest “smart grid” II. DTS TECHNOLOGY
initiative in the power systems, etc. In this paper, we present com-
DTS technology was invented more than 20 years back. The
parative reviews of the different DTS technologies, different appli-
cations, standard, and upcoming, different manufacturers. main measuring principles are based on detecting the back-scat-
tering of light, e.g., using the Rayleigh [1], Raman [2], and Bril-
Index Terms—Ampacity, anti-Stokes, Brillouin, cable tempera- louin [3] principles. This section provides a brief overview of
ture, distributed strain, DTS, fiber optic temperature sensing, fire
the different technologies.
detection, leakage detection, locomotive temperature, Raman scat-
tering, sag, Stokes, structural health, transformer hot spot.
A. Raman Scattering

I. INTRODUCTION Optical fibers are typically made from doped quartz glass, a
form of silicon dioxide (SiO ). With the amorphous solid struc-
ISTRIBUTED temperature sensors (DTS) measure tem- ture of SiO , the thermal effects along the fiber cause lattice
D peratures by means of optical fibers. Those optoelectronic
devices provide a continuous profile of the temperature distribu-
oscillations. As light falls on the thermally excited molecular
oscillations, the photons of the light particles and the electrons
tion along the fiber cable. of the molecule undergo an interaction, resulting in scattered
DTS systems were conceptualized in the 1980s, with signif- light, also known as Raman scattering. The scattered light has a
icant improvements in the technology and the application sce- spectral shift equivalent to the resonance frequency of the lattice
nario over the last decades. DTS systems are capable of mea- oscillation [2], [4].
suring temperature with a high degree of accuracy over sig- The back-scattered light contains three spectral components,
nificant distances. Typical figures are detection accuracy in the the Rayleigh scattering with wavelength of laser source, the
range of C at a resolution of 0.01 C to a spatial resolution Stokes component with the higher wavelength in which the
of 1 m, over measurement distances as long as 30 km. Being photons are generated, and the Anti-Stokes components with
insensitive to the electromagnetic interference (EMI), the fiber a lower wavelength. The intensity of the Anti-Stokes band is
optic-based DTS systems are of particular interest in the elec- temperature dependent, while the Stokes band is temperature
trical applications, e.g., power systems, cables, etc. insensitive. The ratio of the Anti-Stokes and the Stokes light
In this paper, we present a comparative review of the dif- intensities provides the local temperature measurement [5].
ferent DTS technologies and different applications, standard This is shown in Fig. 1. The approach was developed in the
and upcoming, relative to the time of writing. Quasi-continuous 1980s at Southampton University, U.K. [2].
methods like a multitude of fiber bragg gratings inscribed in a
fiber are not considered in this review. The remainder of the B. Brillouin Scattering
paper is organized as follows. Section II provides overviews
about the different technologies of DTS. Section III describes Brillouin scattering refers to the scattering of a light wave
by an acoustic wave due to a nonelastic interaction with the
acoustic phonos of the medium [3], [6]. The Brillouin scattering
Manuscript received March 22, 2011; revised June 22, 2011; accepted July produces both frequency down- (Stokes) and up-shifted (Anti-
08, 2011. This work was supported by the Sensors & Signal Processing (SSP)
program, ABB Corporate Research.
Stokes) light, given by
A. Ukil and H. Braendle are with ABB Corporate Research, Baden
Daettwil CH-5405, Switzerland (e-mail: abhisek.ukil@ch.abb.com; hu- (1)
bert.braendle@ch.abb.com).
P. Krippner is with ABB Corporate Research, Ladenburg, Germany (e-mail: where, is the Brillouin frequency shift, is the angular fre-
peter.krippner@de.abb.com).
quency shift, is the refractive index of the fiber, is the lon-
gitudinal acoustic velocity for the fiber, and is the free-space
wavelength of the pump light [7]. The Brillouin frequency shift
 TABLE I
COMPARISON OF DIFFERENT DTS TECHNOLOGIES

Spammer et al. [14] used a Sagnac interferometer merged

with a Michelson interferometer to detect distributed strain or
Fig. 1. DTS using Raman scattering derived by the ratio of the Anti-Stokes to temperature along the fiber cables. The two interferometers are
Stokes band intensities. illuminated by two light sources. The output of the Michelson
interferometer is proportional to the phase change caused by the
perturbation, while the Sagnac interferometer would be propor-
varies linearly with the strain and the temperature, as given by tional to the product of the phase change by the perturbation and
[7] the distance to it. The distance of the perturbation (e.g., temper-
ature or strain profile) could thus be obtained by dividing the
output of the Sagnac interferometer by the Michelson one [14].
(2)

For m, the strain and the temperature coefficients D. Use of Rare Earth Ions
of the Brillouin frequency shift, due to the strain and the
temperature , are measured as ( m) and Loss of the fiber at a wavelength on the edge of an absorp-
( MHz/K), respectively [8]. Thus, the Brillouin scat- tion band can be monitored by measuring the local fiber absorp-
tering-based technique can be used for sensing both distributed tion using backscattering techniques. Rare-earth materials like
temperature and strain, but not simultaneously both. The dif- holmium, erbium, ytterbium, and praesodymium show high sen-
ferent coefficients would be utilized to separate the effects, pro- sitivity for utilization in DTS systems. Yataghene et al. [15] re-
vided one is interested to measure either the temperature or the ported high thermal sensitivity of holmium-doped optical fibers.
strain, but not in a mixed-mode. Even though Parker et al. [7] Holmium ions show best thermal sensitivity in cryogenic tem-
studied possible ways to simultaneously detect temperature and perature range, around 650 nm [15].
strain using the Brillouin scattering, that is not quite standard Ko et al. [16] reported DTS systems using the temperature de-
yet. pendent gain of 1.48 m pumped erbium-doped fiber amplifier.
The Brillouin scattering-based approach was developed The reported temperature sensitivity of the gain is % C
mainly in the 1990s. The scattering effect can be seen in the for a gain of 28 dB at 20 C. The sensor configuration is sim-
time-domain, e.g., by Brillouin Optical-fiber Time Domain ilar to the optical time-domain reflectrometry, for a potentially
Analysis (BOTDA) [9]. Otherwise, the scattering effect can be cheaper DTS configuration [16].
done in the frequency domain, e.g., by Brillouin Optical-fiber
Frequency Domain Analysis (BOFDA) [10]. The BOFDA
usually results in higher spatial resolution [11]. E. Comparison

Table I shows a comparative summary of the different DTS

technologies developed more or less chronologically in time,
C. Other Fiber Optic Methods namely, the Rayleigh, Raman, and Brillouin scattering. From
Table I, even though the Rayleigh scattering shows the best
Sang et al. [12] used Rayleigh back-scattering to measure dis- accuracy, it is quite limited in terms of range of fiber length
tributed temperature in a nuclear reactor, using 2 m segments which is very important these days to monitor longer lengths of
of commercially available single mode optical fibers and fre- cables, e.g., in underground power distribution systems. From
quency domain reflectrometry. The Rayleigh scattering-based that point of view, the Brillouin scattering provides best length
method claims to measure with an accuracy of 0.6% full scale range, with highest temperature sensitivity and relatively good
with a spatial resolution of 1 cm up to 850 C [12]. measurement time. Brillouin scattering could also detect dis-
Chaube et al. [13] used BOTDA technique for dynamic strain tributed strain, which cannot be done by the other two methods.
measurement. However, traditional BOTDA technique requires However, one could not possibly measure the distributed tem-
time-durations in the range of minutes to make measurements. perature and strain simultaneously. Typically, the applications
Their approach aims to reduce the measurement time by new of the Brillouin scattering are either for distributed tempera-
signal configuration using multiple pump frequencies in the ture measurement or strain, not both simultaneously. Therefore,
form of a frequency-domain comb. They reported measuring Brillouin scattering would likely be a preferred choice in future
the strain distribution of a 120 m long SMF-28 fiber cable in developments, as a replacement for the Raman scattering as al-
256 s, or at a rate of 3.9 kHz [13]. ready seen in the commercial systems [11].
 Fig. 3. Temperature profiling of underground cables, showing temperature dis-
tribution along the cable in meter resolution taken at different time.

A Raman scattering-based DTS system was applied by Kawai

et al. [21] on 77 kV 400 mm XLPE cables to detect fault points
along the cable at an alarm threshold of 5 C. The optical fiber
was placed in a stainless steel sheathed tubular structure of the
Fig. 2. Power cables: (i) XLPE (420 kV), (ii-a) submarine XLPE (420 kV), cable.
(ii-b) submarine mass-impregnated (600 kV), (iii) HVDC Light (320 kV), and ABB studies on 150 kV cable transmission system were re-
(iv) HVDC Light submarine (320 kV). ported in [22], using Raman scattering-based DTS. Example
measurement, shown in Fig. 3, demonstrates the temperature
distribution along the cable with a resolution of 1 m taken at dif-
III. CABLE APPLICATIONS ferent time. The accuracy of the measured temperature was in
the range of C, with measurement time of 5–10 s depending
A. Power Cables and Utilizations on the distance [23]. The spatial positions between 533 to 609
High- and medium-voltage (HV and MV) cables are impor- m show temperature hot spots in the cable. This type of temper-
tant parts of the power transmission and distribution systems. ature hot spots limits the overall ampacity of the cable. How-
Starting from as early as 1883, until date cables for different ever, accurate temperature profiling can quantitatively evaluate
voltage levels are available, e.g., from 36 kV (MV) to extra high the limits of the ampacity which is often set empirically without
voltage (EHV) levels like 420 kV or 500 kV for HVDC appli- knowing it. Thus the cable’s ampacity does not get utilized to
cations [17]. Power cables are increasingly becoming important its full potential [24]. Typical layout configurations of fiber in-
for more and more underground power distribution systems as tegrated cables are shown in [25].
they are considered safe, robust, and optimized for space com-
pared to overhead lines [18]. Fig. 2 shows different classes of C. Other Cable Related Applications
HV cables [17]. Tayama et al. [26] used DTS systems to detect anchor
damage and defacement of wire armor in a 6.6 kV XLPE sub-
B. Cable Temperature and Ampacity marine cable laid in the seabed between Kata and Tomogashima
The maximum current (Amp) carrying capacity of an under- in Japan. Different optical fibers were used for monitoring the
ground cable transmission or distribution system is limited by cable temperature (fiber put in the cable core) and the mechan-
the maximum allowable conductor and cable surface temper- ical damage (fiber put in the outer bedding of core).
ature. This is termed as ampacity [19]. The surrounding soil The optimal usage of DTS systems for cables depend on the
composition, density, and moisture content have influence on conductor temperature as well as prediction of the maximum
the heat dissipation capability of the cable. Thus, even if the allowable ampacity towards safe overloading specification. In
cable has good quality and thermal performance, the inherent in- the direction towards the latter, Li [27] used DTS systems to
homogeneous ambient conditions cause thermal unbalance and estimate the soil thermal parameters from the measurement of
hot spots along underground cables. Thus, continuous thermal the surface temperatures of the cables. From the DTS mea-
profiling along the length of the cable is of particular interest. surements, soil parameters like thermal diffusivity, conductivity,
And DTS is inherently applicable in this context. etc. are estimated using finite-element methods and the gra-
Nakamura et al. [20] reported temperature monitoring of 275 dient-based optimization method.
kV EHV 1 2500 mm aluminium-sheated XLPE cable joints
in underground power transmission lines. They made compara- IV. POWER SYSTEMS APPLICATIONS
tive surface temperature measurements using DTS (Raman scat- As already mentioned in the previous section, long range DTS
tering based) and several thermocouples along the cable. Real- was developed essentially due to the need of monitoring long
time measurements using the DTS system showed significant undersea cables, requiring spatial resolution around 10 m [28].
accuracy within the acquisition time frame of 1 min. This section reviews other power systems applications.
Fig. 4. Fiber optic distributed temperature measurements along the windings
in power transformer.

A. Transformer Monitoring
Transformers are one of the most important as well as ex-
pensive equipments in the power systems. Not being a rotating
element, cooling of the transformer windings is very impor-
tant, which is often done via oil, air (fan), or combined ways.
However, nonuniform localized temperature rise can cause rapid
thermal degradation of insulation. Analytical methods of esti-
mating the temperature distribution and prediction of the hot
spots are discussed in [29] and [30]. Nevertheless, DTS systems Fig. 5. Temperature measurements along the windings in a traction transformer
are quite relevant in this aspect to provide a real temperature in locomotive.
distribution along the transformer winding. Currently, DTS is a
tool for validation of thermal models during transformer devel-
opment and for use in type tests rather than a broad usage of C. Switchgear Monitoring
online monitoring system. Boiarski et al. [32] reported monitoring of distributed tem-
Studies [22], [25], [31] report successful tests using DTS sys- perature in other power plant devices, e.g., switchgear at R.E.
tems (Raman scattering based) to measure the temperature of a Burger plant, West Virginia. Commercially available DTS
22 MVA, oil-cooled power transformer. The fiber was housed system was installed in a 2.4 kV switchgear cell. Starting at
in a v-grove (mm range) with paper isolation in the copper wire 14.7 m, the fiber was attached to the lower finger cluster in
which was used in the winding, the total run length being in the the switchgear with a 2.1 m fiber loop between the two cluster
range of 1000 m. Thus, the DTS was able to measure the exact regions [32]. The DTS system used Rayleigh backscattering.
temperature profile in the transformer winding. The tempera- From the temperature distribution at an accuracy of C,
ture distribution along the winding of the 22 MVA transformer it could be noticed that the maximum temperature in the
is shown in Fig. 4. From Fig. 4, the variations of the temper- switchgear occurs near the top part of the bottom cluster [32].
ature can be noticed, especially the higher temperatures at the
both ends of the winding. This was due to the fact that at the D. Rotating Machine Monitoring
end part of the winding, the inhomogeneity of the magnetic field A DTS system was used to monitor the rotor winding tem-
causes increased eddy current losses. The temperature fluctua- perature of a synchronous condenser at Pacific Gas and Electric
tions noticeable throughout the winding shows the nonuniform substation in San Mateo, California [32]. Four fiber optic sen-
reach of the cooling oil to the inner and the outer depths due to sors, epoxied to prevent vibration effects, were inserted in the
physical layout of the winding. Nevertheless, the detailed tem- separate V-block regions of the rotor. The distributed tempera-
perature profile provides a precise condition monitoring of the ture measurements were in agreement with the simulated data
transformer winding. [32].
DTS system-based monitoring of rotor winding temperature
B. Traction Transformer Monitoring in a boiler feed-pump motor was investigated at the Hudson Av-
Monitoring of the temperature in a traction transformer was enue power station, Con Edison Company, New York [32]. The
also reported in [22] and [31]. For this, DTS system was used DTS system successfully monitored the winding temperature at
to monitor the temperature variations of the winding in a trac- different loading conditions of the motor. It also measured stator
tion transformer in a locomotive (Lok 2000, type 460, no. 107) and air passage temperature rise above ambient as a function of
in Switzerland. Temperature was monitored in the transformer the motor current [32].
winding as the 607 ton locomotive traveled in a route Zurich-
Gotthard-Chiasso in Switzerland, with variations in altitudes. E. Pipework Monitoring in Generating Plants
This gave different loading conditions of the transformer. The In Pressurized Fluidized Bed Combustion (PFBC) power
measurements are shown in Fig. 5. In Fig. 5, the mid-part be- generating plants, hot output gases ( C) from different
tween Zurich and Chiasso has hilly regions which yielded in reactors are brought together, passing through a ceramic filter,
higher temperatures. to drive the turbines for power generation [28]. The temperature
of the corrosive gases flowing under high pressure significantly temperature might cause faster aging as well as cause the con-
exceed the operating limit of the pipes which are insulated ductors to elongate, exceeding the sag safety limits above the
internally. Therefore, it becomes critically important to monitor ground. Thus, sag monitoring is a very important aspect, typ-
the temperature of the pipeline to detect early any problem ically performed by measuring the surface temperature of the
in the insulation, welded joints, seals, etc. DTS systems were phase conductor, measuring conductor tension, measuring using
successfully applied in a fiber loop of approximately 4000 m to image-processing software, using global positioning systems
monitor the temperature. The specially designed high-temper- (GPS), measuring the angle of the conductor at the pole, etc.
ature fiber optic sensor was laid along the steel surfaces. This [36].
would be similar to about 3000 point temperature measure- Utilizing the sag measurements, Olsen and Edwards [36] pro-
ments [28]. posed a lowcost method to indirectly estimate the average con-
ductor core temperature. The principle is based on attaching two
F. Surface Monitoring of Reformer Vessel ends of a grounded wire of high electrical resistance to an ap-
Reformer vessel is a brick-lined steel vessel in which a con- propriate location on each of the two transmission-line towers,
version process is carried out at high temperature ( C) then measuring the induced current on the wire by the transmis-
and pressure. If there are any failures in the internal burner, the sion-line conductors [36].
flame could get disoriented, causing serious damage to the re- The average conductor core temperature could be esti-
fractory bricks. A DTS system was applied to monitor the entire mated by the following relation [36]:
surface of the vessel for hot spots. A total of about 1700 m fiber
loop was used to successfully monitor the temperature of the (3)
surface of the vessel, indicating temperature variations in the
range of C across different parts of the vessel [28]. where is the reference temperature for which the reference
G. Monitoring of Boiler Furnace sag is known, is the measured sag, are the con-
ductor’s weight per unit length, cross-sectional area, effective
Lee et al. [33] reported the development of a DTS system for modulus of elasticity, and temperature coefficient of linear ex-
real time monitoring of high temperature in boiler furnace in pansion, respectively, and is the length of the tower span
power plants. The interesting parameters are the spatial and the [36].
temporal distributions of high temperatures within a gas, coal, Overall, in the overhead line monitoring applications, the
or oil-fired boiler. These play an important role in assessing and DTS systems have to compete against model-based solutions
controlling the polluting sources like NOx. which use only a few temperature measurement points, line
H. Monitoring of Overhead Transmission Lines sag measurement, etc. Here as well, currently DTS is more
a tool during development rather than for online monitoring,
Overhead lines are an indispensable part of the power trans- applications like load balancing, etc. for cost reasons.
mission and distribution (T&D) systems. For modernizing the
aging T&D systems, the utility industry is putting a lot of focus V. OTHER APPLICATIONS
on reliable, flexible, and efficient systems in the “smart grid”
initiative. The ampacity standard of the overhead lines are one A. Structure Monitoring
of the most important factors that influence such initiative.
In T&D system, the ampacity of the overhead lines is gov- Lanticq et al. [37] reported design and validation experiments
erned by the current versus temperature characteristics [34]. to monitor strain and temperature in concrete structures using
Therefore, it becomes highly important to estimate the tempera- fiber optic cables based on Brillouin scattering. They used both
ture distribution along the lines, which would be also influenced standard single-mode fiber and polarization maintaining fiber
by weather and environmental factors like ambient temperature, embedded into a reinforced concrete beam, and the Brillouin
wind speed, any snow condition, etc. Nevertheless, the proper time-domain method (BOTDA). The acceptable strain measure-
understanding of the temperature profiles along the overhead ment results show promise towards continuous health moni-
lines would enable the operators to extend the current carrying toring in structural environments.
capacity within the safety limit. The temperature distribution in This is more and more needed to prevent accidents like the
the overhead conductors would also influence the sag and ten- sudden collapse of I-35W Mississippi river bridge in August
sion of the conductors. 2007 [38]. Continuous health monitoring sensors laid along the
The work by Klein et al. [35] reports development of a re- concrete structures could indicate incipient damages, difficult
liable software-based sag, tension, and ampacity measurement to pick up by human inspection, leading to catastrophic failures
system, called “STAMP”. The program provides conductor [39].
temperature, sag, and tension in real time, taking into ac-
count weather conditions like wind speed, wind direction, air B. Leakage Detection
temperature, etc. The software-based approach is aimed at In civil engineering, health monitoring of dams and dikes is
providing lowcost solution. In their work, Klein et al. made use very important to identify early any leakage in order to pre-
of single-mode DTS system as a benchmark to calibrate and vent catastrophic failures like flooding, etc. The conventional
compare the accuracy of the software program. They reported methods are based on manual inspection which is rather unre-
95% confidence levels in estimating the temperature, sag, and liable. Studies by Khan et al. [40], [41] report the utilization of
tension, vis-á-vis the DTS-based measurements [35]. DTS systems to monitor leakage.
As mentioned above, the amount of power flow in a con- The principle relies on the fact that due to leakage, significant
ductor is limited by the ampacity of the same. Excessive high water flow causes temperature change between the canal water
and the ground. The DTS system then effectively provides the VI. MANUFACTURERS
distributed temperature profile along the dam.
However, the change in temperature could also be caused by The list of manufacturers of DTS systems is a growing one.
other factors like seasonal variation, precipitations, etc. There- Some of the leading manufacturers are mentioned below. How-
fore, source separation analysis [40] is also required to reliably ever, this is not an exhaustive list, and the order does not signify
use the thermometric data from the DTS. any relative measure.
Nikles et al. [42] used Brillouin scattering-based DTS sys- • Sensa® [47]: DTS systems like SUT-® family, DTS-®
tems to perform leakage detection in a 55 km brine pipeline in family for applications like power T&D, leakage detection
Berlin, Germany. The reported accuracy is 1 C over the 55 km in oil and gas, fire detection, etc.
with a measuring time in less than 10 min [42]. • omnisens® [48]: distributed monitoring systems for tem-
Sensornet case-studies [51] report utilization of DTS systems perature, strain, fatigue, etc., like DITEST-® LTM®, AIM®,
to detect leakage in gas and water pipelines for the Korean Na- SHM®, DSM®, STA-R®, DLIGHT series®, etc.
tional Gas authority, monitoring of a 57 km Ethylene pipeline in • es&s® [49]: DTS system like DiTeSt® for distributed tem-
northern Germany, monitoring any leakage in a liquefied natural perature and strain measurement.
gas (LNG) pipeline in an oil and gas company in Texas, etc. • LIOS TECHNOLOGY® [50]: stand alone DTS systems,
integrated Real Time Thermal Rating (RTTR) package.
• sensornet® [51]: fiber optic sensors and digital monitoring
C. Application in Oil and Gas systems.
DTS systems are also applied in the oil and gas industry. In • SensorTran® [52]: DTS product families like ASTRA®,
oil/gas well, DTS installations are typically categorized as re- CENTUARUS®, GEMINI®, NEPTUNE®, CABLES®.
trievable, semipermanent, and permanent. Details about such in- • Weatherford® [53]: optical DTS, optical pass-thru pres-
stallations can be referred to in [43]. sure/temperature gauge.
DTS is used to monitor the temperature log of the well. The • AP SENSING® [54]: DTS system like EN54-5®.
temperature log could then be associated with the effects of • Promore Engineering Inc.® [55]: reservoir monitoring sys-
the liquid flow when a well is shut in. Gas entry into a channel tems.
causes a cooling effect which can be detected in the static • sabeus® [56]: Field Sense™ MPT temperature sensing sys-
rathole. The DTS temperature log could also provide informa- tems, BHPt pressure sensing systems.
tion on the cooling effect in a water injection well. DTS systems • LUNA Technologies® [57]: Distributed Sensing System™
are also used to monitor wells on a pad or production platform, (DSS) 4300 for making distributed measurements of tem-
steam breakthrough in a production well. DTS systems could perature and strain.
also be used to detect the start up of electric submersible pumps • HALLIBURTON® [58].
(ESP) and motor [43]. • MAXIM® [59].
Besides the manufacturer companies, there are some non-
profit professional societies as well.
D. Fire Detection • The Fiber Optic Association Inc. [60].
• Subsea Fiber Optic Monitoring Group [61].
Automatic fire detection is an important topic to prevent asset
• IEEE Photonics Society [62].
damage and human casualty. Automatic fire detection systems
are based on optical smoke detector, ionization detector, infrared
or ultraviolet flame detection, etc. One of the most promising VII. DISCUSSIONS AND FUTURE TRENDS
technologies is linear optical fire detection, based on the DTS
principle, e.g., Raman scattering [44]. Using frequency domain The following comments are cited.
reflectrometry in a silica glass fiber, the study [44] reported 1) The cost of DTS systems and installations vary greatly
tracking of rapid changes in the temperature profile to detect for different applications and settings. For example, in the
fire and localize the seat of the fire within a building with 1 m power systems and cable applications often the voltage
resolution. level, type of cable, distance, etc. are the influencing fac-
tors. Typical costs could be in the range of USD 100 000. In
E. Applications in Mines oil and gas applications, also the cost depends highly on the
well configurations, distance, applications, etc., typically
For safe electrical operation, in mines, shuttle car trailing ca- being in the range of USD 50 000–150 000 [43]. Therefore,
bles should be operated below the safety limit which is about the cost of the DTS systems is still on the higher side.
90 C [45]. Operation in temperature condition more than that 2) Therefore, one obvious future need would be to have
might cause premature insulation failure. Therefore, tempera- cost-effective DTS solutions maybe in the range of USD
ture monitoring under dynamic condition is very important. 10 000–20 000, or less.
Dubaniewicz et al. [45] reported utilization of DTS systems 3) Despite the relatively higher cost, the real-value of DTS
embedded within the metallic conductors to measure tempera- systems is to provide distributed temperature and/or strain
tures at 1 m intervals along the entire cable length, at an accuracy measurement. This is often useful for safety critical or dif-
of C. ficult to access applications, involving equipments which
In addition to the distributed temperature, Dubaniewicz et al. themselves are highly expensive, e.g., HV power trans-
[46] also reported use of fiber optic technology to monitor at- former, subsea oil and gas applications, etc. Getting a di-
mospheric related parameters like methane (CH ) and carbon rect and continuous measurement of the temperature pro-
monoxide (CO). file to prevent potential damage to such equipments is a
 growing trend in continuous condition monitoring for pre- [2] J. P. Dakin, D. J. Pratt, G. W. Bibby, and J. N. Ross, “Distributed optical
dictive maintenance. fiber Raman temperature sensor using a semiconductor light source and
detector,” Electron. Lett., vol. 21, pp. 569–570, 1985.
4) Therefore, a better way of judging the value of DTS-based [3] X. Bao, D. J. Webb, and D. A. Jackson, “Combined distributed tem-
monitoring would be to introduce alternative business perature and strain sensor based on Brillouin loss in an optical fiber,”
models based on total cost of ownership (TCO) for ex- Opt. Lett., vol. 19, no. 2, pp. 141–143, 1994.
pensive devices. TCO would incorporate device costs, life [4] A. H. Hartog, “Distributed temperature sensing in solid-core fibers,”
Electron. Lett., vol. 21, pp. 1061–1062, 1985.
cycle and operating costs, costs including all additional [5] T. Shiota and T. Wada, “Distributed temperature sensors for single
infrastructure, maintenance, replacement in case of no mode fibers,” Proc. SPIE, vol. 1586, pp. 13–18, 1991.
monitoring, etc. [6] X. Bao, J. Dhliwayo, N. Heron, D. J. Webb, and D. A. Jackson, “Ex-
5) With the “smart grid” initiative, integration of renew- perimental and theoretical studies on a distributed temperature sensor
based on Brillouin scattering,” J. Lightwave Technol., vol. 13, no. 7,
ables and distributed energy sources like off-shore wind pp. 1340–1348, 1995.
farms, remote photovoltaics, etc. are foreseen. This would [7] T. R. Parker, M. Farhadiroushan, R. Feced, V. A. Handerek, and A. J.
imply increased utilizations of HVDC cables, etc. Syn- Rogers, “Simultaneous distributed measurement of strain and temper-
ature from noise-initiated Brillouin scattering in optical fibers,” IEEE
onymously, usage of the DTS systems for monitoring the J. Quantum Electron., vol. 34, no. 4, pp. 645–659, 1998.
cable health, optimal usage of the cable current carrying [8] D. Culverhouse, F. Farahi, C. N. Pannel, and D. A. Jackson, “Potential
capacity, better loading of the cables, overhead conductors of stimulated Brillouin scattering as sensing mechanism of distributed
for energy efficient operations would be increasingly temperature sensors,” Electron. Lett., vol. 25, pp. 913–914, 1989.
[9] K. Shimizu, T. Horiguchi, and Y. Koyamada, “Measurement of dis-
important. tributed strain and temperature in a branched optical fiber network by
6) Over the last decades, the DTS interface softwares have use of Brillouin optical time-domain reflectrometry,” Opt. Lett., vol.
evolved greatly with flexible graphical user interfaces, 20, no. 5, pp. 507–509, 1995.
faster acquisition time, interoperability amongst the dif- [10] D. Garus, T. Gogolla, K. Krebber, and F. Schliep, “Brillouin optical-
fiber frequency-domain analysis for distributed temperature and strain
ferent vendor equipments, etc. measurements,” J. Lightwave Technol., vol. 15, no. 4, pp. 654–662,
7) With the advancement of Brillouin scattering-based DTS 1997.
systems, structural health monitoring of concrete struc- [11] V. Lecoeuche, M. W. Hathaway, D. J. Webb, C. N. Pannell, and D. A.
tures, buildings, dams, etc. are possible. This is very impor- Jackson, “20-km distributed temperature sensor based on spontaneous
Brillouin scattering,” IEEE Photon. Technol. Lett., vol. 12, no. 10, pp.
tant to prevent catastrophic failures [38]. Also, continuous 1367–1369, 2000.
monitoring of leakages along water, oil, and gas pipelines [12] A. K. Sang, M. E. Froggatt, D. K. Gifford, S. T. Kreger, and B. D. Dick-
over several kms would be an advantage to improve the erson, “One centimeter spatial resolution temperature measurement in
a Nuclear reactor using rayleigh scatter in optical fiber,” IEEE Sensors
system or plant efficiency and reliability. J., vol. 8, no. 7, pp. 1375–1380, 2008.
8) The fiber optic-based DTS systems are inherently insensi- [13] P. Chaube, B. G. Colpitts, D. Jagannathan, and A. W. Brown, “Dis-
tive to electromagnetic interference, etc. This is of partic- tributed fiber-optic sensor for dynamic strain measurement,” IEEE Sen-
ular interest in the electrical applications, e.g., power sys- sors J., vol. 8, no. 7, pp. 1067–1072, 2008.
[14] S. J. Spammer, P. L. Swart, and A. A. Chtcherbakov, “Merged
tems, cables, etc. Sagnac-Michelson interferometer for distributed disturbance detec-
tion,” J. Lightwave Technol., vol. 15, no. 6, pp. 972–976, 1997.
VIII. CONCLUSION [15] A. Yataghene, M. Himbert, and A. Tardy, “Distributed temperature
sensor using holmium-doped optical fiber and spread-spectrum tech-
DTS systems measure temperatures by means of optical niques,” J. Lightwave Technol., vol. 15, no. 4, pp. 654–662, 1997.
fibers, detecting via Rayleigh, Raman, and Brillouin principles. [16] P. K. Y. Ko, S. Demokan, and H. Tam, “Distributed temperature
Using Brillouin scattering, it is also possible to measure the dis- sensing with Erbium-doped fiber amplifiers,” J. Lightwave Technol.,
vol. 14, no. 10, pp. 2236–2245, 1996.
tributed strain. DTS systems provide a continuous profile of the [17] Handbook, whitepapers ABB Cables, 2010. [Online]. Available: http://
temperature that is distributed along the cable or other device. www.abb.com/cables.
In this paper, we presented a comparative view of the different [18] K. T. Yoon and D. S. A. Teo, “Controlling and monitoring Singapore’s
standard DTS technologies along with nonstandard ones. On underground grid,” IEEE Comput. Appl. Power, vol. 12, no. 4, pp.
23–29, 1999.
application side, we reviewed distributed temperature sensing [19] B. L. Harshe and W. Z. Black, “Ampacity of cables in covered tray,”
in cables for better ampacity judgment. In the power systems, IEEE Trans. Power Del., vol. 12, no. 1, pp. 3–14, 1997.
monitoring of power transformer and traction transformer in [20] S. Nakamura, S. Morooka, and K. Kawasaki, “Conductor temperature
order to detect hot spots is very important. Monitoring of the monitoring system in underground power transmission XLPE cable
joints,” IEEE Trans. Power Del., vol. 7, no. 4, pp. 1688–1697, 1992.
power plant devices like switchgear, rotating machine wind- [21] T. Kawai et al., “A new approach to cable fault location using fiber
ings, boiler furnace, reformer vessel, etc. are reviewed as well. optic technology,” IEEE Trans. Power Del., vol. 10, pp. 85–91, 1995.
Leveraging advancement of Brillouin scattering, applications [22] H. Braendle, T. Rendenbach, and B. Weber, Fiber Optic Temperature
like structural health monitoring in concrete structures, leakage Measurements, Tech. Rep., no. TDT-D-90-01/2, ABB, 1993.
[23] Handbook ABB Cables, ABB Kable und Draht GmbH. Mannheim,
monitoring in dams, oil and gas application, fire detection, and Germany, 1993.
safety applications like in mines are reviewed. The number of [24] L. Goehlich, F. Donazzi, and R. Gaspari, “Monitoring of HV cables
companies providing DTS systems is growing ever so much, offers improved reliability and economy by means of power sensors,”
Power Eng. J., pp. 103–110, Jun. 2002.
some being listed herein. Overall, the DTS systems provide [25] G. Nokes, “Optimising power transmission and distribution networks
great advantages in terms of continuous profiling of distributed usig optical fibre distributed temperature sensing systems,” Power Eng.
temperature and/or strain for safety critical or difficult to access J., pp. 291–296, Dec. 1999.
applications. [26] H. Tayama, O. Fukuda, Y. Inoue, Y. Koike, and K. Yamamoto, “6.6
kV XLPE submarine cable with optical fiber sensors to detect anchor
damage and defacement of wire armor,” IEEE Trans. Power Del., vol.
REFERENCES 10, no. 4, pp. 1718–1723, 1995.
[1] A. H. Hartog, “A distributed temperature sensor based on liquid-core
optical fibers,” J. Lightwave Technol., vol. 1, pp. 498–509, 1983.
[27] H. J. Li, “Estimation of soil thermal parameters from surface temper- [54] AP SENSING®, Manuals. [Online]. Available: http://www.apsensing.
ature of underground cables and prediction of cable rating,” IEE Proc. com.
Trans. Gen. Trans. Distrib., vol. 152, no. 6, pp. 849–854, 2005. [55] Promore Engineering Inc.®, Manuals, whitepapers. [Online]. Avail-
[28] A. Hartog, “Distributed fibre-optic temperature sensors: Technology able: http://www.promore.com.
and applications in the power industry,” Power Eng. J., vol. 9, no. 3, [56] sabeus®, Manuals, whitepapers. [Online]. Available: http://www.
pp. 114–120, 1995. sabeus.com.
[29] M. K. Pradhan and T. S. Ramu, “Prediction of hottest spot temperature [57] LUNA Technologies®, Manuals, whitepapers. [Online]. Available:
(HST) in power and station transformers,” IEEE Trans. Power Del., http://www.lunatechnologies.com.
vol. 18, pp. 1275–1283, 2003. [58] HALLIBURTON®, Manuals, whitepapers. [Online]. Available: http://
[30] M. K. Pradhan and T. S. Ramu, “Estimation of hottest spot temperature www.halliburton.com.
(HST) in power transformers considering thermal inhomogeniety of [59] MAXIM®, Application Note 687. [Online]. Available: http://www.
the windings,” IEEE Trans. Power Del., vol. 19, no. 4, pp. 1704–1712, maxim-ic.com/an687.
2004. [60] The Fiber Optic Association, Inc., Tech. information, whitepapers.
[31] ABB Secheron Ltd., “Measurement of temperature profile in trans- [Online]. Available: http://www.thefoa.org.
former windings with optical distributed sensors,” in Proc. EPRI Conf., [61] Subsea Fiber Optic Monitoring Group, Publications, whitepapers. [On-
1997. line]. Available: http://www.seafom.com.
[32] A. A. Boiarski, G. Pilate, T. Fink, and N. Nilsson, “Temperature [62] IEEE Photonics Society, Journals, Conf. Publications. [Online]. Avail-
measurement in power plant equipment using distributed fiber optic able: http://www.photonicssociety.org.
sensing,” IEEE Trans. Power Del., vol. 10, no. 4, pp. 1771–1778, 1995.
[33] K. Y. Lee, B. H. Kim, and J. P. Velas, “Development of an intelligent
monitoring system with high temperature distributed fiber-optic sensor
for fossil-fuel power plants,” in Proc. IEEE PES General Meeting, Abhisek Ukil (S’05–M’06–SM’10) received the
2004, vol. 2, pp. 1989–1994. bachelor of electrical engineering degree from
[34] IEEE Standard for Calculation of the Current-Temperature Relation- the Jadavpur University, Calcutta, India, in 2000,
ship of Bare Overhead Conductors, IEEE Std. 738, NY, 2006. the M.Sc. degree in electronic systems and engi-
[35] K. M. Klein, P. L. Springer, and W. Z. Black, “Real-time ampacity and neering management from the University of Bolton,
ground clearance software for integration into smart grid technology,” Bolton, U.K., in 2004, and the Ph.D. degree from
IEEE Trans. Power Del., vol. 25, no. 3, pp. 1768–1777, 2010. the Tshwane University of Technology, Pretoria,
[36] R. G. Olsen and K. S. Edwards, “A new method for real-time moni- South Africa, in 2006, working on power systems
toring of high-voltage transmission line conductor sag,” IEEE Trans. disturbance analysis with Eskom.
Power Del., vol. 17, no. 4, pp. 1142–1152, 2002. From 2000 to 2002, he worked as a software
[37] V. Lanticq, M. Quiertant, E. Merliot, and S. Delepine-Lesoille, “Bril- engineer at InterraIT, India. After joining in 2006,
louin Sensing cable: Design and experimental validation,” IEEE Sen- currently he is a Principal Scientist at the Integrated Sensor Systems Group,
sors J., vol. 8, no. 7, pp. 1194–1201, 2008. ABB Corporate Research, Baden-Daettwil, Switzerland. He is author/coauthor
[38] S. Gupta, “Damaging bridges will cry for help before they fail,” New- of more than 40 refereed papers, a monograph Intelligent Systems and Signal
Scientist, pp. 18–19, Jan. 2011. Processing in Power Engineering (New York: Springer, 2007), two book
[39] S. A. Wade, K. T. Grattan, and B. McKinley, “Incorporation of fiber- chapters, and inventor/co-inventor of six patents. His research interests include
optic sensors in concrete specimens: Testing and evaluation,” IEEE signal processing, machine learning, power systems, embedded systems,
Sensors J., vol. 4, no. 1, pp. 127–134, 2004. sensors, and sensor-based systems.
[40] A. A. Khan, V. Vrabie, J. I. Mars, A. Girard, and G. DUrso, “A source
separation technique for processing of thermometric data from fiber-
optic DTS measurements for water leakage identification in Dikes,”
IEEE Sensors J., vol. 8, no. 7, pp. 1118–1129, 2008.
[41] A. A. Khan, V. Vrabie, J. I. Mars, A. Girard, and G. DUrso, “Automatic Hubert Braendle received the Ph.D. degree in
monitoring system for singularity detection in dikes by DTS data mea- physics from the University of Zurich, Zurich,
surement,” IEEE Trans. Instrum. Meas., vol. 59, no. 8, pp. 2167–2175, Switzerland, in 1974.
2010. From 1974 to 1978, he was a Postdoctoral Re-
[42] M. Nikles et al., “Leakage detection using fiber optics distributed tem- searcher in high-energy physics with the University
perature monitoring,” in Proc. SPIE Int. Symp. Smart Structure Mate- of California, Los Angeles, and with the Swiss
rial, San Diego, CA, 2004, pp. 18–25. Federal Institute of Technology Zurich, Zurich,
[43] J. J. Smolen and A. vd Spek, “Distributed temperature sensing—A DTS Switzerland. In 1978, he joined BBC (now ABB
primer for oil & gas production,” in Tech. Report, Shell, 2003. Switzerland, Ltd.), Baden, Switzerland. He is
[44] U. Glombitza and H. Hoff, “Fibre optic radar system for fire detection currently a Research Fellow at the Sensor Tech-
in cable trays,” in Proc. 13th Int. Conf. Automatic Fire Detection, Duis- nologies Group, ABB Corporate Research Center,
berg, Germany, 2004. Baden-Daettwil, Switzerland, where he is engaged in research on sensor
[45] T. H. Dubaniewicz, P. G. Kovalchik, L. W. Scott, and M. A. Fuller, technologies, optical sensing, and analytics.
“Distributed measurement of conductor temperatures in mine training
cables using fiber-optic technology,” IEEE Trans. Ind. Appl., vol. 34,
no. 2, pp. 395–398, 1998.
[46] T. H. Dubaniewicz, J. E. Chilton, and H. Dobroski, “Fiber optics for Peter Krippner received the Ing. degree in elec-
atmospheric mine monitoring,” IEEE Trans. Ind. Appl., vol. 29, no. 4, trical engineering from the Karlsruhe Institute of
pp. 749–754, 1993. Technology, Karlsruhe, Germany, in 1995, and
[47] Sensa®, Manuals, whitepapers. [Online]. Available: http://sensa.org. the Ph.D. degree in mechanical engineering from
[48] omnisens®, Manuals, whitepapers. [Online]. Available: http://www. the Karlsruhe Institute of Technology, Karlsruhe,
omnisens.ch. Germany, in 2000.
[49] es&s®, Manuals, whitepapers. [Online]. Available: http://www.esands. He was a group leader at the Karlsruhe Institute of
com. Technology dealing with micro optical sensing de-
[50] LIOS TECHNOLOGY®, Manuals, whitepapers. [Online]. Available: vices. He joined ABB Corporate Research, Laden-
http://www.lios-technology.com. burg, Germany, in 2000, working as senior principal
[51] sensornet®, Manuals, whitepapers. [Online]. Available: http://www. scientist in the area of field instruments and gas and
sensornet.co.uk. liquids analyzers with a focus on micro electro mechanical systems. Since 2007,
[52] SensorTran®, Manuals, whitepapers. [Online]. Available: http://www. he has been heading the global ABB R&D program for sensing and signal pro-
sensortran.com. cessing, covering the whole range of sensing technologies and signal processing
[53] Weatherford®, Manuals, whitepapers. [Online]. Available: http:// techniques to create innovative products as well diagnostic functionality in the
www.ep-solutions.com. area of condition monitoring.