2.1.1.

Fiber Bragg grating

FBG is a periodic and permanent modification of the core refractive index value (typically ${10}^{-5}$ to ${10}^{-3}$) along the optical fiber axis.^23–25 Two counter-propagating core modes will be coupled, which cause a Bragg reflection (shown in Fig. 2) and the Bragg wavelength can be expressed as

Fig. 2

Schematic of an FBG.

The main advantage of FBG is an absolute measurement and independent of the light fluctuating. Due to the narrowband reflective spectrum and the low insertion loss, FBG arrays are easily constructed along a single-mode fiber (SMF). The excellent multiplexing capability is beneficial for distributed sensing or quasidistributed sensing. Nowadays, FBGs play more and more important roles in sensing applications.

2.1.2.

Fiber-optic interferometer

Fiber-optic interferometer is a common technology in sensing areas. In most fiber-optic interferometers, light is split into two beams that propagate in different optical paths then combined again.²⁶ The output light contains information about the optical path difference and is displayed as a cosine function. The phase difference $\mathrm{\Delta }\mathrm{\Phi }$ can be described as

Based on different working principles, a fiber-optic interferometer can be categorized into a Mach–Zehnder interferometer (MZI), Michelson interferometer (MI), Fabry–Perot interferometer (FPI), and Sagnac interferometer (SI).

As shown in Fig. 3(a), an MZI works in a transmission mode. The light is launched into the fiber and then coupled into two fibers by a coupler. The other coupler is used to combine the light in two fibers to realize the interference. The in-line MZI-based sensors have been rapidly developed due to their compact configurations and designable functions. They usually operate with two modes and realized interference in one fiber, eliminating the need for couplers. Nowadays, some typical configurations have been designed using long-period grating (LPG), SMF, photonic crystal fiber (PCF), and tapered fibers.^27–31

Fig. 3

Schematic of fiber interferometers: (a) MZI, (b) MI, (c) FPI, and (d) SI.

The MI works in a reflection mode as shown in Fig. 3(b). The MI sensor generally required a mirror element to reflect the light at the end of fiber. Some typical in-line MI configurations are also designed using LPG, PCF, twin-core fiber, and so on.^32–35

An FPI is generally composed of two parallel reflecting surfaces separated by a certain distance (called an etalon).³⁶ Two reflected beams form an interference fringe. Two categories of FPI sensors, extrinsic FPI (EFPI) and intrinsic FPI (IFPI), are shown in Fig. 3(c) according to the reflectors outside or inside the fiber. The EFPI could obtain a high-finesse interference signal by utilizing high reflective mirrors.³⁷ The inside reflector of IFPI sensor could be formed by micromachining, FBG, chemical etching, and thin film deposition.^38–41 It is important to design an etalon length of FPI-based sensor to balance the requirements of sensing range and sensitivity in applications.

Different from the above fiber interferometers, an SI is formed by an optical fiber loop. As illustrated in Fig. 3(d), the input light is divided into two directions by a coupler and the clockwise and counter-clockwise propagating light combined again by the same coupler.⁴² The output spectrum contains the beating frequency of the two counter-propagating light if the fiber coil is rotated. Nowadays, SIs has been applied in various sensing applications and fiber-optic gyroscopes⁴³ may be the most significant.

2.1.3.

Distributed fiber-optic sensing technology

For distributed fiber-optic sensing technology, the fiber optical itself acts as a continuous array of sensors. The environmental perturbation affects the parameters of the core of fiber, such as optical length, diameter, and refractive index, which modulate the backscattered light, including its intensity, phase, and frequency. Monitor the backscattered light of the fiber, the surrounding environment perturbations can be figure out.

The backscattering can be categorized as Rayleigh, Brillouin, and Raman, which are shown in Fig. 4. Rayleigh scattering is caused by the inhomogeneity of the fiber core and the backscattered Rayleigh light has the same the wavelength with the input light.^44–47 Brillouin scattering is caused by an acoustic wave from lattice vibration, which is the interaction between the input light and the phonon.⁴⁷ The Brillouin backscattering is usually about 15 and 20 dB weaker than the Rayleigh backscattering and has a frequency shift $\sim 11\mathrm{GHz}$ for SMF around 1550 nm. This frequency shift is sensitive to temperature and strain.⁴⁸ Spontaneous Raman scattering is caused by the phonons⁴⁹ and about 10 dB weaker than the spontaneous Brillouin scattering. A frequency shift $\sim 13.0\mathrm{THz}$ with a wide bandwidth of $\sim 6\mathrm{THz}$ is much larger than that of the Brillouin scattering. The intensity of Stokes signal is temperature insensitive, whereas the anti-Stokes signal is temperature sensitive. Therefore, the Raman scattering light can be used to temperature measurement.⁵⁰ Some distributed sensing technologies based on the scattering mechanisms are introduced here briefly.

Fig. 4

Typical spontaneous scattering spectrum in an optical fiber.

OTDR

Since demonstrated by Barnoski and Jensen,⁵¹ OTDR has become a commercial instrument to characterize an optical fiber. The working principle of OTDR is illustrated in Fig. 5. A short optical pulse is launched into the fiber and monitoring the attenuation of the backscattered signal by a detector.^52,53 The Rayleigh backscattering light should be decayed exponentially with time. The external perturbations (e.g., bending, break-induced end face, or introduce a connector) applied on the fiber will be changed the attenuation at the perturbation location. The attenuation of Rayleigh backscattered light is used for distributed sensing along the fiber.

Fig. 5

The principle of operation of the OTDR.

The spatial resolution of an OTDR is defined as $c\tau /2n$. $c$ and $n$ are the light velocity in vacuum and refractive index of optical fiber, respectively. $\tau$ is the bandwidth of the input pulse. The lower the bandwidth of the input pulse, the higher the spatial resolution. However, it means lower pulse power, which decrease the signal-to-noise ratio (SNR) of OTDR and is a serious problem for long-range sensing. Therefore, it is important to optimize the bandwidth of pulse to satisfy the spatial resolution and the sensing range in application. In addition, the response time of the detector is usually several nanoseconds, which also limited the spatial resolution of OTDR.

$\mathrm{\Phi }$-OTDR

The $\mathrm{\Phi }$-OTDR technique was first proposed by Taylor and Lee⁵⁴ by detecting the intensity changes of the interferometric light and has become an effective tool for distributed vibration and intrusion sensing nowadays.^55–58 Different from OTDR, $\mathrm{\Phi }$-OTDR requires a highly coherent laser with narrow line width. The output signal of $\mathrm{\Phi }$-OTDR is modulated by the coherent interaction of numerous scattering centers within the pulse duration. The working principle of the $\mathrm{\Phi }$-OTDR is shown in Fig. 6. When a perturbation is applied on a sensing fiber, the phase will change at the position due to the change in refractive index and length of the fiber. This induces the intensity change corresponding to the time or the position. The backscattered Rayleigh light is launched into an interferometer system, such as an imbalanced MI (or MZI). The output interference signal is split into three signals differing by $2\pi /3$ phase delay by a $3\times 3$ coupler. Three PDs detect the signals and a data acquisition card saved the signals. Then the perturbation can be identified by proper demodulation algorithm. Similar to OTDR, the spatial resolution and the sensing range need to be balanced. Another challenge for $\mathrm{\Phi }$-OTDR is the polarization mismatch, which could reduce the probability of detecting perturbation points and result in wrong signals.

Fig. 6

The principle of operation of the $\mathrm{\Phi }$-OTDR.

B-OTDR and B-OTDA

The Brillouin scattering-based distributed sensing was focused on temperature and strain measurement.^59–62 Brillouin scattering-based sensors are categorized as the B-OTDR and B-OTDA.

B-OTDR is based on spontaneous Brillouin scattering and was initially introduced to enhance the working range of OTDR.⁶³ The operation principle of B-OTDR is shown in Fig. 7. A pulsed laser launched optical pulse into the sensing fiber and generate spontaneous Brillouin scattering. A local laser generates continuous wave (CW) light to act as the heterodyne light. The heterodyne light is mixed with the scattered light. The detector is receiving the coherent signal. Despite that the back scattered signal is weak, it can be work even if the fiber is broken, which is suited for industrial applications.⁶⁴ B-OTDA is based on stimulated Brillouin scattering. As shown in Fig. 8, a pulse laser and a CW laser launched light to both ends of the fiber. To detect the weak counter propagating Brillouin scattering signal, a coherent detection technique is adopted.⁶⁵ When the frequency difference of the CW signal and the input pulse is equal to the Brillouin frequency shift, stimulation of the Brillouin scattering occurs.^53,66 B-OTDA measures the gain of the two counter propagating waves and improves the performance. However, it does not work when the loop is broken at any point of the fiber.

Fig. 7

The principle of operation of the B-OTDR.

Fig. 8

The principle of operation of the BOTDA.

R-OTDR

Since being proposed in the 1980s,^50,67 Raman-based sensing systems have widely been used in distributed temperature measurement. The operation principle of R-OTDR is illustrated in Fig. 9. A pulsed laser source launched a pulse into the fiber. The detector used to measure the backscattered Stokes and anti-Stokes band responses over a roundtrip propagation. A section of fiber that knows the temperature is inserted between the coupler and sensing fiber for temperature calibration. In principle, there are some limitations with the types of fiber in R-OTDR systems. For multimode fiber, the intermodal dispersion will broaden the pulse signal and reduce the spatial resolution. For SMF, a high-performance laser source and anti-Stokes photodetector are required because the wavelength of anti-Stokes signal should be larger than the cut-off wavelength of the fiber ($\sim 1310\mathrm{nm}$ for standard SMF). For example, an R-OTDR sensing system based on 15-km SMFs requires the detector working long wavelengths, such as InGaAs photodetectors.⁶⁸

Fig. 9

The principle of operation of the R-OTDR.

2.2.1.

Partial discharge monitoring

Partial discharge is defined as a localized dielectric breakdown under high-voltage stress, which bridges the insulation between two conductors.⁶⁹ It is the main phenomenon that causes the degradation of insulation of high-voltage power transformers. To ensure the safety of power transmission, partial discharge measurements are an effective method of monitoring the condition of power transformers.

Partial discharge generates mechanical stress waves that propagate through the surrounding oil in the range of 100 to 300 kHz.²¹ To detect these waves, acoustic emission detection methods are the most used in partial discharge monitoring. Compared with ultrahigh frequency method, the acoustic detection method is immune to electromagnetic interference, and it is easy to locate the insulation defect with the detected signal. Optical fiber as a dielectric material can be easily placed directly into the power transformer to get close to the potential partial discharge source, enabling it to diagnose small defects. Many investigations demonstrated the excellent characteristics of optical fiber sensor in partial discharge monitoring.

Interferometric sensors have been widely investigated in partial discharge monitoring. Some researchers used MZI to detect partial discharge by acquiring acoustic signals.^70–73 Based on all fiber MZI and homodyne demodulation, 1.3-Pa acoustic pressure can be detected in a transformer.⁷² A multichannel heterodyne interferometer capable of processing four channels simultaneously is applied to locate the partial discharge position.⁷³ An MI-based sensor with a high SNR and flat response between 20 to 150 kHz is used to monitor the partial discharge.⁷⁴ EFPI sensors^75–78 have been used by some researchers to detect partial discharge due to its compact size and realize single-point measurement. It is need to note that at the same conditions, the EFPI configuration had a better performance compared with the IFPI sensor.⁷⁹ Due to truly path-matched interference mode, Sagnac interferometric sensor shows its stability against environmental temperature influences and attracted researchers’ interest. Based on the fiber-optic SI, internal acoustic pressure and vibration due to the partial discharge in the transformer was monitored.⁸⁰ A fiber-optic acoustic sensor was reported using a balanced Sagnac sensor and an EDFA-based fiber ring laser with the degree of polarization tunable function. The experimental results showed its feasibility to monitor high-frequency acoustic pressure to 300 kHz.⁸¹

FBG sensor^82–84 is also used widely because of its multiplexing capability and small size. Due to impulsive acoustic pressure generated during partial discharge, detecting the wavelength shift of FBGs is a solution for partial discharge monitoring.⁸² The intensity demodulation method based on FBGs and a wide band laser is also suitable for partial discharge monitoring.⁸³ This type of partial discharge detection system does not require either an FBG analyzer or a narrow band tunable laser source. However, the sensitivity of the system was not very high. FBGs combined with FPI configuration can improve the sensitivity of detecting the weak pressure.⁸⁴ Another method to improve the sensitivity is to use a phase shift FBG (PS-FBG).^85,86 Because the PS-FBG exhibits a sharp resonance, the ultrasonic sensitivity is improved compared with the FBG. It has been reported that the sensitivity of PS-FBG-based ultrasonic sensor is 8.46 dB higher than PZT sensor between 50 to 400 kHz.⁸⁶

2.2.2.

Temperature monitoring

The undesirable heat radiation in the transformer weakens the insulation of transformers, which causes faults in the high probability. The transformer losses are among the most important factors in rising of top oil and hot-spot temperature (HST). This causes rapid thermal degradation of insulation. HST, the highest temperature on the oil or winding, is the most important parameter of the transformer’s life calculation. Industry standards limit maximum allowable HSTs in transformers to 140°C with conventional oil insulation.²¹

Traditional calculation and measurement of temperature distribution inside power transformers including calculation formula for HST,⁸⁷ finite-element method,⁸⁸ analytical methods,^89,90 and diagnostic measurements.⁹¹ Optical fiber sensors are a better choice for temperature monitoring in power transformer for the reasons of electromagnetic interference and intrinsic safety. The Raman scattering-based sensing technology can realize a distributed temperature sensing (DTS) along the whole optical fiber. The temperature distribution along the winding of a 22-MVA in an oil-cooled power transformer was measured by DTS technology,^92,93 shown in Fig. 11. It is a trend for the temperature measurement in transformer.

Fig. 11

Fiber-optic distributed temperature measurements along the windings in a power transformer.⁹³

The FBG array is another choice to realize quasi-DTS in power transformer, which overcome the long acquisition time and the lower sensitivity of DTS system. An FBG temperature sensor achieved temperature measurement within the transformer winding.^94,95 Hot-spot measurements based on FBG sensors were realized on the power transformers of 25 kVA,⁹⁶ 154 kV,^97,98 $31.5\mathrm{MVA}/110\mathrm{kV}$,⁹⁹ and $35\mathrm{kV}/4000\mathrm{kVA}$.¹⁰⁰ Optical fiber sensors have much lower error deviation compared with that of Pt100 resistance thermometers.¹⁰¹

2.2.3.

Hydrogen monitoring

In a transformer, oil worked as insulation, coolant, as well as the condition indicator will generate gasses, such as ${\mathrm{H}}_2$, ethane (${\mathrm{C}}_2{\mathrm{H}}_6$), methane (${\mathrm{CH}}_4$), acetylene (${\mathrm{C}}_2{\mathrm{H}}_2$), ethylene (${\mathrm{C}}_2{\mathrm{H}}_4$), carbon monoxide (CO), and carbon dioxide (${\mathrm{CO}}_2$) when the electrical or thermal stresses are accumulated to a certain degree. The fault gases in the transformer oil will reduce its insulation. The gases abnormity is the first evidence of an incipient fault in a transformer. For example, the abnormal ${\mathrm{C}}_2{\mathrm{H}}_6$ and ${\mathrm{C}}_2{\mathrm{H}}_4$ concentrations often mean there is thermal fault $<300°\mathrm{C}$ and between 300°C to 700°C, respectively, while the abnormal ${\mathrm{H}}_2$ concentration means a partial discharge or a thermal fault between 150°C to 300°C.²¹ According to the IEEE standards,¹⁰² the concentrations for these key gases and the corresponding status of the transformers are listed in Table 1.

Table 1

Dissolved gas concentrations for the key gases and total dissolved combustible gas (TDCG).

The traditional methods to analyze the dissolved gases in transformers including the chromatography,^103,104 spectroscopy,^105,106 and gas sensors.^107–110 Some electrochemical sensors based on semiconducting metal oxides have been utilized to gas sensing.^107–110 Optical fiber gas sensors attract people’s interest due to its low cost, simple structure, and immunity to electromagnetic interference. Nowadays, optical fiber sensors have realized monitoring these gases of concern, such as ${\mathrm{CH}}_4$,^111–114 ${\mathrm{C}}_2{\mathrm{H}}_2$,^115,116 and ${\mathrm{H}}_2$, in transformer oil. Among these gases, ${\mathrm{H}}_2$ especially in Ref. 117, provides a clearer insight in the health condition of a transformer and allows for an early failure evaluation.

The ${\mathrm{H}}_2$ concentration sensors need to meet the following specifications to provide a continuous monitoring of the condition of power transformer. They should work from environment temperature to at least 110°C. The ${\mathrm{H}}_2$ concentration detection range should be from 100 to 1000 ppm. Selectivity toward ${\mathrm{H}}_2$ is crucial to avoid false alarms.¹¹⁸

The main types of fiber-optic sensing schemes, FBG-,^119–125 modified FBG-,^126–129 reflection,^130,131 and interferometer-¹³² based are all by coating Pb-film on fiber as shown in Figs. 12(a)–12(d), respectively. For FBGs coating with the Pd-based film, the expansion of the Pd-based thin film induced by the H absorption would introduce the stress on FBGs and their Bragg wavelength may shift, which is used as the sensing scheme. The thickness of Pd films shows different responses.¹¹⁹ An FBG hydrogen sensor based on Pd/Ag composite film displays better stability than that of the pure Pd film.^120,121 Optimizing the microstructure by reducing the grain size or thickness of the films can improve the sensor performance. Pd/Au film¹²² and Pd/Cr film¹²³-based FBGs hydrogen sensors were also reported. A polymer coating is added as an intermediate layer to improve the poor adhesion between the thick Pd film and optical fiber.^124,125 To improve the sensitivity, a D-shaped FBG sensor-coated with a Pd thin film by magnetron sputtering was reported to test dissolved ${\mathrm{H}}_2$ in transformer oil.¹²⁶ A high sensitivity of 1.96 ppm at every 1-pm wavelength shift was demonstrated in the range of 0 to 719.7 ppm of dissolved ${\mathrm{H}}_2$.

Fig. 12

Schemes of fiber-optic ${\mathrm{H}}_2$ sensors based on Pd-based films. (a) FBG,¹¹⁴ (b) D-shape FBG,¹²⁸ (c) reflection-based sensor,¹³¹ and (d) interferometer.¹³²

Some literature reported a side-polished FBG-based hydrogen sensor, whose scheme is illustrated in Fig. 12(b). Due to its intrinsically sensitive to curvature, side-polished FBG can improve the sensitivity of ${\mathrm{H}}_2$ sensing. Compared to a standard FBG coated with the same ${\mathrm{WO}}_3$-Pd film, the side-polished FBG can increase the sensitivity of the sensor by $>100\%$.¹²⁷ The sensitivity of a Pd/Ag composite film-based side-polished FBG sensor is about 11.4 times the conventional FBG sensors.¹²⁸ Like the structure of the side-polished FBG, the sensitivity of a $16\text{-}\mu \mathrm{m}$ chemically etched FBG-based hydrogen sensor is improved by $>30\%$.¹²⁹

For reflection-based sensors,^130,131 as shown in Fig. 12(c), the Pd film is coated at the end face of an optical fiber and its reflectivity is changed due to the ${\mathrm{H}}_2$ absorption. ${\mathrm{WO}}_3\text{-}{\mathrm{Pd}}_2\mathrm{Pt}\text{-}\mathrm{Pt}$ composite film¹³⁰ coated on the end face of the optical fiber exhibited a detection limit of 20 ppm in air at 25°C. The filtering function of ${\mathrm{Pd}}_2\mathrm{Pt}\text{-}\mathrm{Pt}$ catalyst layer provided a good selectivity against 2% ${\mathrm{CH}}_4$ and 1% CO. A polytetrafluorethylene-Pd (PTFE-Pd) thin film-based fiber-optic sensor¹³¹ was demonstrated to measure the dissolved ${\mathrm{H}}_2$ concentration. The results indicated that a larger range of ${\mathrm{H}}_2$ concentrations could be measured at higher temperatures, which are more relevant for the temperature of an operating power transformer.

A PCF modal interferometer-based ${\mathrm{H}}_2$ monitoring sensor for power transformer was also demonstrated.¹³² As shown in Fig. 12(d), when the $\mathrm{Pd}/{\mathrm{WO}}_3$ film coated on the PCF surface absorbs ${\mathrm{H}}_2$, the effective phase difference of interferometer was changed by the induced stress and the resonant wavelength of interference spectrum was shifted. Experiment results showed a linear correlation between the resonant wavelength and the concentration of ${\mathrm{H}}_2$ dissolved in the transformer oil and a sensitivity of about $0.109\mathrm{pm}/(\mu \mathrm{L}/\mathrm{L})$ with a response time of $\sim 30\min$.

In summary, partial discharge, temperature, and hydrogen abnormity induced by electrical and thermal stresses in transformers are the main early faults needed to be detected. As shown in Table 2, types of fiber-optic interferometers are suitable for partial discharge monitoring by detecting acoustic emission signals and FBG sensors with smaller size and stronger multiplexing capability are an alternative choice but with lower sensitivity. A distributed optical sensing system, mainly based on R-OTDR, is a solution in temperature monitoring with the accuracy of $\sim 1°\mathrm{C}$, the spatial resolution of meter-level and the measurement time of second-level. Quasidistributed sensing technology, FBGs, with less measurement points is more suitable for the requirements of higher resolution of centimeter level, higher measurement speed of kilohertz level, and higher accuracy of $\sim 1°\mathrm{C}$. Hydrogen in transformer oil has been monitored by Pb-film-coated optical fiber sensors. To improve their sensitivity and response, time optimizations are still focused on the film thickness, materials, and the structures of optical fiber sensors.

Table 2

Optical fiber sensing applied in the power transformers.

2.4.1.

Temperature monitoring

The temperature along the transmission lines is a crucial parameter because it affects the ampacity of overhead transmission lines. The common temperature measurement of transmission lines includes noncontact infrared technology¹⁴³ and direct measurement through the surface electronic thermometer.^144,145 However, these methods easily interact with strong electromagnetic interference from the transmission line and require additional power. Nowadays, optical fiber sensors become an effective method to measure the temperature distribution of the transmission lines. Generally speaking, two kinds of optical fiber temperature sensing schemes, the point sensing and distributed fiber-optic sensing, have been applied in transmission lines.

B-OTDR¹⁴⁶ based on the Brillouin frequency shift with temperature of optical fiber has been used to realize the online temperature monitoring of three-phase cables with OPPC/OPGW. The experimental results showed that the temperature difference between the three-phase cables and the optical fiber inside is 0.018°C. R-OTDR¹⁴⁷ system with the temperature accuracy of $\pm 1°\mathrm{C}$ was used to measure the temperature of 400/50 OPPC with different actual working conditions of current-carrying capacity, wind velocity, and environment temperature. B-OTDR-based distributed temperature sensor for localizing the lightning stroke-induced temperature change in OPGW was also reported.¹⁴⁸

FBGs^149,150 with portable systems were also used in temperature monitoring of transmission lines and they demonstrated a higher accuracy of $\sim 0.1°\mathrm{C}$ and greater stability. An FBG-based temperature measurement system was installed on No. 44 tower of 110-kV Zhengzou transmission line in Luzhou, China.¹⁵¹ The FBG system was a quasidistributed sensing technology and needed to inscribe the FBG in the optical fiber, which is not fully matched with OPPC/OPGW and limit its application.

2.4.2.

Icing monitoring

Heavy ice coating of overhead transmission lines, as shown in Fig. 15, imposes a serious threat to the safe operation of power grids, such as conductor breakage, insulator flashover, and tower collapse. Numerous theoretical icing models have been developed to create reliable tools for predicting the features of icing process.¹⁵²

Fig. 15

Icing of transmission lines.

Many methods including climatological data method, image method, load cell method, have been developed to monitor the ice load of overhead transmission lines. For the conventional electrical load cell used in ice monitoring, it is easily disturbed by strong electromagnetic interference, and it usually needs a power supply. The optical fiber load cell has been widely applied in monitoring the ice coating of overhead transmission lines without these problems.

Ogawa et al.¹⁵³ presented an FBG load cell for ice monitoring on overhead transmission lines and carried out the tension experiment to acquire the relationship between the load and strain. To solve the cross sensitivity of strain and temperature, an unforced FBG is usually used as a comparison sensor. An FBG-based sensing system realized online icing monitoring on the 44# tower of 110-kV Zhenzou power grid.¹⁵⁴ An overhead conductor tension sensor based on FBG^155,156 was connected to the tower and the insulator with metallic clamps to monitor the icing-induced tension of 110-kV high-tension transmission line in Zhaotong, Yunnan province. The experiment results indicated that the tension sensitivity of the FBG sensor was $9.8\mathrm{pm}/\mathrm{kN}$ and shows a good repeatability. A distributed online temperature and strain fiber sensing system based on the combination of B-OTDR and FBG were also proposed¹⁵⁷ and shown in Fig. 16. The B-OTDR sensing system is used to measure the temperature of overhead lines and the FBG sensing system is adopted to measure the tension of overhead lines.

Fig.16

Schematic diagram of galloping monitoring based on FBG sensors.¹⁵⁷

To improve the performance of FBG-based icing monitoring system, some structures were designed for load cell.^158–164 Elastic element with two near-elliptical-shaped concavities structure¹⁶¹ and coupled dual-beam (“S” beam) structure,¹⁶² which is shown in Figs. 17(a) and 17(b), respectively, can be worked in harsh environments with a good sensitivity and resolution. The shearing structure with additional grooves as elastic element of load cell was designed to detect the eccentric load and temperature simultaneously.^163,164 Two vertically FBGs were mounted onto the additional grooves to eliminate temperature effects on strain measurement without extra FBG.

Fig.17

(a) Elastic element with two near-elliptical-shaped concavities structure¹⁶¹ and (b) elastic element with coupled dual-beam (“S” beam) structure.¹⁶²

Some analysis models were also proposed by researchers.^165–167 A correction factor of gravity acceleration was added and greatly improved measurement performance in windy weather.¹⁶⁵ A strain difference model achieved absolute stress change values by measure the line sag difference constant of two points.^166,167 Experiment in Xuefeng mountain of Hunan shows that the strain difference model is superior to the traditional calculation method and can be applied on monitoring ice thickness of transmission lines.

Wydra et al.¹⁶⁸ presented a method that monitors the state of a transmission lines based on chirped FBG (CFBG) sensors. Compared to the uniform FBG, the linearly CFBG has a more considerable variation in wavelength, which would be used to improve the sensitivity and accuracy. In the presented CFBG sensor, temperature can be compensated by the spectrum shift.

2.4.3.

Galloping monitoring

Transmission line galloping is wind-induced vibration of both single and bundle overhead conductors with low frequency (typically 0.1 to 3 Hz) and large amplitude (about 5 to 300 times of the power transmission lines diameter), which usually occurs in a specific environment with strong winds and transmission lines icing and lasts for several hours. Galloping may reduce air gaps between conductors, occasionally causing flashovers and repeated power supply interruptions, which is really harmful for power grid.¹⁶⁹

Galloping has been studied for many years in theory and experiment, and many measures have been developed.^170,171 However, the galloping of transmission lines under different circumstances has not been accurately studied. The experimental method is an effective way to study the mathematical model of galloping or to verify the device to prevent flashovers.¹⁷² Therefore, there is a great need for an effective online galloping monitoring system. The commonly used galloping monitoring techniques are camera and electrical accelerometers. However, these methods are susceptible to high-voltage environment and require power supply for monitoring devices. Optical fiber sensing technology maybe overcomes these limitations and plays a more and more important role in real-time monitoring. Moreover, OPPC/OPGW applied widely in transmission lines accelerates the process.

Bjerkan¹⁷³ monitored the vibrations of a 160-m span of a 60-kV line by FBG sensors gluing on the phase conductor. Huang et al.¹⁷⁴ investigated an FBG strain sensor on the phase conductor to measure its tension. Rui et al. designed a two-dimensional FBG-based acceleration sensor to monitor the power transmission line galloping.¹⁷⁵ Chen et al.¹⁷⁶ used FBG-based strain sensor and temperature sensor to monitor the galloping of an overhead power transmission line. FBG sensors were installed on both ends of the OPGW cable, which can be seen from Fig. 18. The experiment results showed that the galloping amplitude could be obtained accurately in real time.

Fig.18

Schematic diagram of galloping monitoring based on FBG sensors.¹⁷⁶

Hao et al.¹⁵⁷ developed a distributed online temperature and strain fiber sensing system based on the combined B-OTDR and FBG technology. The temperature of the transmission lines was monitored by B-OTDR, whereas the tension of the transmission lines was measured by FBG sensing system. The experimental line simulated the actual transmission line of Yun-Guang $\pm 80\mathrm{kV}$ DC transmission project. The experiments showed that Brillouin scattering and the FBG reflected signal would not affect each other. The transmission line load variation could be measured effectively and accurately by the FBG tension sensor.

A dynamic tension detection system composed of FBG tension sensors, OPGW, and a wavelength interrogator, as shown in Fig. 19, was designed to detect the galloping of overhead transmission lines.¹⁷⁷ The FBG tension sensor was installed between an insulator string and power tower to monitor the dynamic tension of the phase conductor. A series of experiments at the State Grid Key Laboratory of Power Overhead Transmission Line Galloping proved its feasibility.

Fig.19

Schematic diagram of galloping monitoring system.¹⁷⁷

In summary, transmission lines are one of the most important parts in a power grid system but easily affected by weathering, such as icing and wind blowing. In addition, the excessive temperature rise will cause high-power energy loss and some accidents during high-voltage transmission process. Table 3 illustrated the optical fiber sensing technology applied on power towers and transmission lines. Two main fiber-optic sensing technologies applied in transmission lines are the FBG sensing and distributed sensing technologies. FBG sensors are more stable and cheaper and portable demodulation system. In contrast, the distributed sensing technology requires a highly demanding light source and a demodulator, but it can utilize the fiber in OPPC/OPGW directly without extra packaging and installation processes. Any point information along the OPPC/OPGW can be measured by distributed sensing technologies. The sensing cost is reduced dramatically and could be a trend for the health monitoring of long-distance transmission lines.

Table 3

Optical fiber sensing technology applied on power towers and transmission lines.