A Study on the Influence of End-Sheath Aging and Moisture Absorption on Abnormal Heating of Composite Insulators

Abstract

Abnormal heating of composite insulators of high-voltage transmission lines concentrate at its end, especially in a high-humidity environment. In order to study the influence of end-sheath aging and moisture absorption on abnormal heating of composite insulators, in this paper, we first discuss the appearance test, temperature rise test, and dielectric characteristic test conducted on 110 kV decommissioned composite insulators. Test results indicated the temperature rise in composite insulators increased with ambient humidity, but temperature rise was not severely affected by surface contamination of its shed and sheath; in dry environments, the dielectric constant and dielectric loss factor of high-voltage end sheaths are higher than of those of medium- and low-voltage end sheaths, and the loss effect becomes more severe after moisture absorption in a high-humidity environment. After the tests, the authors established a COMSOL simulation model of composite insulators, to analyze changes in the electric field and thermal field of the end sheath of composite insulators due to the coupling of electric and thermal fields. It was concluded that the dielectric constant of a high-voltage end sheath of the composite insulator increased after moisture absorption, distorting the partial electric field on the surface; meanwhile, the dielectric loss factor increased significantly after water molecules intruded into the aging layer of the sheath as polar molecules. Therefore, the dielectric loss (leakage conductance loss and lossy polarization loss) caused by aging and moisture absorption of the sheath surface under partially high field strength in the high-humidity environment was the leading cause of abnormal heating at the high-voltage end of composite insulators. The conclusion of this paper serves as an important reference for revealing the causes of abnormal heating of composite insulators in high-humidity environments and the influence mechanism of external factors on abnormal heating.

1. Introduction

Composite insulators are widely used in high-voltage transmission lines by virtue of their excellent pollution flashover resistance, high mechanical strength, and low cost [1,2,3,4,5]. By far, the number of composite insulators in China with 110 kV and above has exceeded 10 million, and their safe operation is highly significant for the stable operation of the power system in view of such a massive number of insulators.

Abnormal heating and other defects frequently occur with composite insulators in high-humidity areas where the length of service life and further heating may lead to internal insulation breakdown, broken strings, and other vicious faults [6,7,8]. According to incomplete statistics, there have been 13 abnormal fracture accidents of 500 kV composite insulators in China since the 1990s. Observations with an infrared thermometer before fracture indicated the high-voltage end of these composite insulators underwent abnormal heating [9]. At present, the generation mechanism of abnormal heating of composite insulators has not been sufficiently identified. Under certain environmental conditions, the surface contamination of the shed and sheath, the partial discharge caused by internal defects of the insulator, water intrusion, and other factors may lead to abnormal heating of the composite insulator [10,11,12,13,14]. In one study [10], abnormal heating at the composite insulator end was attributed to the polarization loss of surface contamination of the sheath under high field strength at the end, rather than an internal insulation defect, but the authors failed to consider the influence of moisture absorption after aging of silicone rubber sheaths on abnormal heating of composite insulators. According to the gray density and salt density tests of abnormal heating composite insulators in the literature [11], it was found that contamination was light, and the possibility of heating caused by contamination was excluded. The analysis of moisture absorption properties of silicone rubber indicated abnormal heating of composite insulators was caused by the increase in the dielectric constant of the sheath after moisture absorption, change in insulator electric field, and aggravation of polarization loss. However, that study has neglected the influence of the change in dielectric loss factor on abnormal heating after the sheath absorbs moisture. In another study in the literature [12], the authors deemed the dielectric loss tangent of composite insulator sheath increased with water molecule content in the sheath, leading to the abnormal heating of composite insulators. However, the study did not consider the influence of changes in the dielectric constant on the electric field at the end surface of the composite insulator sheath after aging and moisture absorption. The authors of [13,14] studied the infrared testing and electrical testing conducted for a composite insulator with abnormal heating, and test results showed a partial discharge caused by internal defects of the insulator was the leading cause of its abnormal heating. The above-mentioned studies were only experimental studies, focusing on theoretical analysis of experimental phenomena and laws, but did not cover simulation verification of corresponding theories during their cause analysis of abnormal heating of composite insulators.

In view of this, in this paper, we describe the infrared temperature rise test carried out for 110 kV decommissioned composite insulators, and to this end, the temperature rise properties of composite insulators at different ambient humidity levels were acquired. Through experimentation and theoretical analysis, we studied the influence of surface contamination and dielectric properties of the silicone rubber sheath on abnormal heating of composite insulators and revealed the cause of abnormal heating at the high-voltage end of 110 kV decommissioned composite insulators in high-humidity environments. Furthermore, the finite element analysis method was adopted to simulate and analyze the change in surface field strength of composite insulators before and after moisture absorption and the abnormal heating of composite insulators in high-humidity environments. The correctness of the analysis of abnormal heating of decommissioned composite insulators in this paper was verified. The conclusion of this paper serves as an important reference for revealing the causes of abnormal heating of composite insulators in high-humidity environments and the influence mechanism of external factors on abnormal heating.

2. Samples and Test Methods

2.1. Samples

The composite insulator is composed of high temperature vulcanized silicone rubber shed umbrella cover, glass fiber reinforced resin mandrel and hardware, as shown in Figure 1, where the silicone rubber umbrella cover not only bears the electrical load, but also protects the mandrel. It can be known from the preparation process of high temperature vulcanized silicone rubber that its main components are polydimethylsiloxane, flame retardant aluminum hydroxide and silica.

In this paper, two typical samples were selected from a batch of 110 kV composite insulators decommissioned due to abnormal heating, to carry out infrared temperature rise tests for the samples at different humidity levels. Specific parameters of the samples are listed in

Table 1. Parameters of samples.

Sample No.	Model	Structure Height (mm)	Number of Sheds (Big/Small Sheds)	Mandrel Diameter (Including Sheath) (mm)
#1	FXBW-110/100	124	27(14/13)	28
#2	FXBW-110/100	134	23(12/11)	28

. This batch of deteriorated composite insulators had operated for about 15 years; thus, their sheds and sheaths have been subjected to obvious aging. The two decommissioned 110 kV composite insulators are numbered #1 and #2, respectively. The two composite insulator samples featured no obvious crack marks on the shed and sheath, aged and white shed color, slight contamination, and no deteriorated internal interface and mandrel based on visual observation, with the appearance shown in Figure 2.

2.2. Test Scheme

The infrared temperature measurement tests of two insulators were carried out in a climatic chamber that was designed with the dimensions of 4.5 m × 4.5 m × 2.5 m, the adjusted relative humidity within the range of 30~95% RH (relative humidity), an error of ±5%, and the ambient temperature (20 ± 3) °C. Restricted by the climatic chamber height, the composite insulator samples were horizontally placed, and the high- and low-voltage ends of the insulator were supported by insulating struts with the sample insulation distance of 1.2 m above the floor. The test layout is presented in Figure 3.

A FLIR T1040 28 infrared imager (FLIR Systems Inc, Portland, OR, USA) was used for recording the temperature rise in composite insulators. The infrared thermal imaging camera features an IR resolution of 1024 × 768 pixels, the temperature ranged from −40 to +2000 °C, and its emissivity was set to be 90%.

Test steps are as follows:

• Before the test started, the composite insulator samples were placed in the climatic chamber, the mist amount of the ultrasonic humidifier (Wetwells, Hangzhou, China) was adjusted by rotating the knob so the ambient humidity reached the set value, and the samples were stabilized at this state for 30 min to ensure uniform distribution of humidity in the climatic chamber;
• When humidity was 30% RH (ambient humidity), the AC voltage with an effective value of 70 kV (110 kV × 1.1/$\sqrt{3}$) was applied to the composite insulator samples, and the temperature rise in the sample was recorded every 15 min. The temperature rises in the samples reached stable levels after the voltage was applied for 60 min, and the temperature rises in the samples were recorded;
• In an environment with 55% RH, the insulator samples were placed for 72 h for full moisture absorption, and the AC voltage with an effective value of 70 kV was applied to the samples. After the voltage was applied for 60 min, the temperature rise in the sample reached a stable level, which was recorded;
• Step 3 was repeated for the composite insulator samples in an environment with 75% RH and 95% RH.

3. Infrared Temperature Rise Test

The infrared temperature rise curve of samples can be obtained with Equation (1) according to DL/T664-2016 [15] as follows:

T = T₁ − T₀

(1)

where T₁ is the highest temperature of the samples after the voltage is applied for 60 min (unit: K); T₀ is the temperature of the ambient temperature reference body (unit: K).

In Figure 4, the temperature rise curves are illustrated for two insulator samples at different humidity levels after the voltage was applied for 60 min.

As shown in Figure 4, the temperature rise in two 110 kV insulators increased with the ambient humidity. In low-humidity environments (≤55% RH), the peak temperature rise in #1 and #2 samples was less than 0.5 K, and no obvious heating occurred. The peak temperature rise increased to about 1.0 K when humidity was 75% RH, and the peak temperature rise in the two samples reached 2.6 and 2.0 K, respectively, when humidity was 95% RH. The temperature rise in the two samples in a high-humidity environment was obviously higher than in a low-humidity environment. Preliminary analysis showed the reasons are as follows: (1) With the increase in humidity, the moisture absorption of the string contamination layer led to an increase in the leakage current and leakage conductance loss on the surface of the composite insulator, and then the increase in temperature; (2) microcracks and micropores appeared on the surface of the silicone rubber sheath after aging, and microdefects on the surface of the silicone rubber sheath accelerated the intrusion of external water molecules in a high-humidity environment. As polar molecules, water molecules underwent polarization loss under the action of an alternating electric field, thus increasing the temperature rise.

4. Cause Analysis of Sample Heating

4.1. Influence of High-Voltage End Contamination on Composite Insulator Heating

To explore the partial influence of surface contamination on abnormal heating of composite insulators, first, surface contamination of sample #1 with higher temperature rise was cleaned with a wet sponge, and it was placed indoors for 72 h natural drying. The sample was then placed in an environment with humidity of 95% RH for 72 h full moisture absorption; then, the infrared temperature rise test was conducted at 95% RH, and the temperature rise in the sample was observed. The results are presented in Figure 5.

The parts in front of three sheds and sheaths at the high-voltage end of sample #1 were recoated with grade IV contamination. The salt density was 0.3 mg/cm², and the ratio of diatomite to sodium chloride was 6:1 in accordance with GB/T 16434-1996 [16]. Figure 6 shows the temperature rise in sample #1 at 95% RH after recoating with grade IV contamination.

Table 2. Infrared inspection results of sample #1 under different test conditions.

Test Humidity	Test Conditions	Temperature Rise/K
High humidity (95% RH)	Without treatment	2.6
	Surface contamination cleaning	2.7
	The parts in front of three sheds and sheaths at the high-voltage end were recoated with grade IV contamination	2.6

lists the results of the temperature rise in the samples before and after surface contamination treatment of composite insulators. As shown in the table, sample #1 still encountered abnormal heating at the high-voltage end after surface contamination was cleaned, and the temperature rise was 2.7 K, indicating no significant increase, compared with 2.6 K before treatment. After the high-voltage end was coated with contamination, the temperature rise in sample #1 was 2.6 K, which was roughly the same as that after surface contamination was cleaned. Therefore, surface contamination was not the cause of abnormal heating of decommissioned composite insulators in this study.

4.2. Influence of Aging and Moisture Absorption of Sheath Surface on Dielectric Properties of Composite Insulators

The Novocontrol broadband dielectric spectrometer was used for measuring the dielectric properties of silicone rubber, and its equipment data sheet is shown in

Table 3. Equipment data sheet for Novocontrol broadband dielectric spectrometer.

The Main Parameters
Model	NOVOCONTROL
Temperature range	−160~400 °C
Frequency range	3 × 10−6~3 × 109 Hz
Impedance range	10 mΩ~100 TΩ
Capacitance range	1 fF~1 F
Phase difference accuracy	2 × 10−3
Tan(δ) accuracy	3 × 10−5
Measuring voltage	10−6~3 V
DC bias voltage range	±40 V

.

Silicone rubber materials were cut and acquired from the sheath surfaces of the high-, medium-, and low-voltage ends of sample #1, and a silicone rubber test piece with a flat surface and a thickness of about 1 mm was prepared with the materials. The test pieces were dried in a drying oven for 48 h and then placed in a constant temperature and humidity box with a relative humidity of 95 ± 3% RH and temperature of 20 ± 2 °C for 72 h. The dielectric properties of each test piece before and after moisture absorption were measured with the broadband dielectric spectrometer at a temperature of 20 °C and a frequency of 50 Hz. The dielectric constant and dielectric loss factors of the test piece at 50 Hz are listed in

Table 4. Dielectric constant and dielectric loss factors of test pieces before and after moisture absorption.

Location	Dielectric Constant		Dielectric Loss Factor
	Before Moisture Absorption	After Moisture Absorption	Before Moisture Absorption	After Moisture Absorption
High-voltage end	4.36	7.20	1.30%	12.90%
Medium-voltage end	4.16	6.30	0.60%	6.10%
Low-voltage end	4.18	6.71	0.59%	6.52%

.

As shown in Table 4, the dielectric properties of high-voltage silicone rubber samples were much higher than those of medium-voltage silicone rubber samples and low-voltage silicone rubber samples, and their dielectric constant and dielectric loss factors were 1.14 and 2.11 times those of medium-voltage silicone rubber samples, respectively, after moisture absorption. After the aging of composite silicone rubber, polar groups were generated inside the insulator, and defects such as micropores and microcracks appeared on the surface. Those microdefects provided convenient channels for water intrusion, and the aging of the sheath at the high-voltage end was more obvious than that at the medium- and low-voltage ends, leading to different dielectric properties among the high-, medium-, and low-voltage ends [17,18]. The dielectric constant and dielectric loss factors of the high-voltage silicone rubber material of sample #1 after moisture absorption further increased to 1.65 and 9.92 times that of materials in a dry environment. The reasons were described as follows: The content of water molecules in silicone rubber material increased; water molecules are a medium with high polarity, and its dielectric constant is 81 at 20 °C, while the silicone rubber is a medium with low polarity, and its dielectric constant is 3~5 at 20 °C [19]. The distribution of the electric field on the composite insulator surface was influenced greatly by the dielectric constant of the material. The increased dielectric constant of silicone rubber after moisture absorption led to the change in surface field strength at the high-voltage end of a composite insulator, the polarization loss increased significantly with the dielectric loss factor, resulting in abnormal heating at the composite insulator end.

5. Simulation Analysis of Causes of Abnormal Heating of Composite Insulators

5.1. Comparison of Electric Field at High-Voltage End of Composite Insulators before and after Moisture Absorption

The infrared temperature rise results in Section 3 indicate that the heating parts of the two decommissioned composite insulators were both at the high-voltage ends. Therefore, we explored the change in field strength distribution at the high-voltage end of the insulators after moisture was absorbed by silicone rubber. Based on the dielectric constant of silicone rubber after moisture absorption measured in Section 4.2, the field strength at the end surface of the insulator shed and sheath after moisture absorption was calculated using COMSOL—a finite element analysis simulation software program. Under normal operating conditions, the insulator voltage is AC power frequency voltage, and the wavelength of 50 Hz AC voltage is about 6 × 10⁶ m, which is much larger than the size of each part of overhead transmission lines. Therefore, it can be approximately considered that the field strength of the insulator is relatively stable all the time, and the electric field distribution of the insulator can be analyzed through electrostatic field analysis [20]. A 110 kV composite insulator model was built by using SOLIDWORKS, a 3D modeling software program, and the applied voltage of the model was 70 kV. The dielectric constant of silicone rubber of sample #1 after moisture absorption was 7.20, and the simulation results of the electric field at the end of the insulator are shown in Figure 7. The dielectric constants of silicone rubber of composite insulators were set to be 4.36, 5.30, 6.30, and 7.20, respectively. Figure 8 illustrates the peak values of field strength at the end-sheath surface and the shed root acquired through the electric field simulation.

According to simulation results, when the relative dielectric constant increased, the peak field strength on the sheath surface was basically unchanged and was kept at about 13.35 kV/cm, while the peak field strength at the shed root increased from 13.72 to 14.28 kV/cm, and the peak electric field increased by about 4.1%. As water molecules feature high polarity, they will turn in the direction of an electric field under the action of an alternating electric field. The stronger the electric field, the more complete the turning orientation of polar molecules and the stronger the turning polarization. When field strength increased by 4.1%, the increased turning polarization aggravated the polarization loss.

5.2. Comparison of Temperature Field at High-Voltage End of Composite Insulators after Moisture Absorption

The heat source of the composite insulators after moisture absorption is mainly the dielectric loss of the end sheath. As the electric field on the composite insulator surface is non-uniform, the power expression is selected as follows:

$$P=E^2\omega \epsilon \mathrm{tg}\delta V$$

(2)

where P is the heating power of dielectric loss, E is the electric field strength, ω = 100π rad/s, ε is the relative dielectric constant of silicone rubber, tan δ is the dielectric loss tangent of silicone rubber, and V is the dielectric volume.

Heat dissipation of composite insulators includes heat dissipation in the forms of radiation, solid conduction, and air convection, and the expression is as follows [21]:

$$q_{\mathrm{f}}={\epsilon }_1\sigma (T^4-T_0{}^4)$$

(3)

$$q_{\mathrm{c}}=-\lambda \mathrm{grad}T$$

(4)

$$q_{\mathrm{d}}=h(T-T_0)$$

(5)

where q_f, q_c, and q_d are heat flux densities for heat dissipation in the forms of radiation, solid conduction, and air convection, respectively; ε₁ is the radiation coefficient; σ is the Boltzmann constant; T and T₀ are composite insulator temperature and ambient temperature, respectively; λ is solid thermal conductivity; and h is the heat dissipation coefficient of air convection. These values for each parameter are provided in

Table 5. Electrical and thermal characteristic values of the used parameters.

Parameters	Values	Parameters	Values
ω	100 π rad/s	h	5 W/(m2·°C)
ε	7.2	ε1silicone rubber	0.95
λmetal	49.8 W/(m·°C)	σ	5.67 × 10−8 W/(m2·°C4)
λsilicone rubber	2 W/(m·°C)	T0	20 °C
tan δHV end sheath	12.90	tan δshed and sheath excluding HV end shed and sheath	6.10

, and in Figure 9, a comparison between the simulation results and the infrared images taken on site is illustrated.

According to simulation results of the temperature field at the high-voltage end of composite insulators after moisture absorption, the maximum temperature rise in the composite insulator reached 3.0 K, which is slightly higher than the temperature rise of 2.6 K measured during the infrared temperature rise test of sample #1. Analyzed causes of this error may include the emissivity of the infrared thermal imaging camera, slight disturbance of ambient humidity, infrared measurement distance, etc.

According to the comparison between the simulation results and the infrared image of sample #1, simulation results are basically consistent with the temperature rise and heating position in the infrared image, and the peak temperature rise observed with both methods appeared at the root of the first shed at the high-voltage end, thus further verifying abnormal heating at the end of the insulator is caused by dielectric loss under the combined action of AC high field strength and water molecules in the sheath.

In conclusion, the dielectric constant of the high-voltage end sheath of composite insulators increased after aging and moisture absorption, which led to an increase in the peak value of the electric field at the root of the first shed at the high-voltage end. Meanwhile, the dielectric loss factor increased significantly after water molecules intruded into the aging layer of the sheath as polar molecules. Therefore, the dielectric loss (leakage conductance loss and lossy polarization loss) caused by aging and moisture absorption of the sheath surface under partial high field strength in high-humidity environments was the leading cause of abnormal heating at the high-voltage end of composite insulators.

In view of the abnormal heating phenomenon caused by the aging and moisture absorption of the end sheath, controlling the increase in the electric field becomes the most important preventive measure, and the peak value of the electric field on the surface of the insulator can be reduced by adjusting the shielding depth of the grading ring.

6. Conclusions

In this paper, an infrared temperature rise test was conducted for 110 kV composite insulators decommissioned from the site, and the dielectric properties of silicone rubber before and after moisture absorption were measured. Relevant conclusions of abnormal heating of composite insulators were drawn as follows with the finite element analysis method:

(1)

The abnormal heating of two typical 110 kV composite insulators detected in this study occurred at the high-voltage end, and temperature rise increased with ambient humidity. In low-humidity environments, the peak temperature rise in composite insulators was less than 0.5 K, without obvious heating; the highest temperature of the sample reached about 1.0 K in an environment with 75% RH; the peak temperature rise in the two samples reached 2.6 and 2.0 K, respectively, in an environment with 95% RH;

(2)

It was concluded that surface contamination was not the main cause of abnormal heating of decommissioned composite insulator samples based on a comparison of the changes in the surface temperature rise in composite insulator samples before and after surface contamination cleaning and before and after recoating with grade IV contamination;

(3)

The aging of the high-voltage end sheath was more obvious than that of the medium- and low-voltage ends, and its dielectric constant and dielectric loss factors were 1.14 and 2.11 times those of the medium-voltage end-sheath material, respectively. The dielectric constant and dielectric loss factors of the high-voltage end-sheath surface after aging and moisture absorption in high-humidity environments further increased to 1.65 and 9.92 times those in a dry environment. Combined with finite element simulation analysis, it was concluded that the partial electric field on the high-voltage end-sheath surface was distorted after aging and moisture absorption in high-humidity environments, and this caused the maximum value of the partial electric field to increase by about 4.1%. Therefore, the dielectric loss (leakage conductance loss and lossy polarization loss) caused by aging and moisture absorption of the sheath surface under partial high field strength was the leading cause of abnormal heating at the high-voltage end of composite insulators;

(4)

According to the comparison of the temperature rise and abnormal heating positions of composite insulators in an infrared temperature rise test and electrothermal coupling simulation, the test results and simulation results were basically consistent, further verifying the dielectric loss after aging and moisture absorption of the sheath surface was the leading cause of abnormal heating at the high-voltage end of composite insulators.