1. Introduction
The transition to an industrial structure requiring high power is accelerating due to the Fourth Industrial Revolution and the COVID-19 pandemic. Therefore, the scale of electric facilities is gradually expanding, and accidents and blackouts at power facilities can cause enormous economic losses in a highly power-dependent society. As the number of long-term power facilities has increased, so has interest in evaluating the deterioration state of insulation in major power plants.
Underground power transmission cables are essential for supplying electricity generated from thermal power, hydropower, nuclear power, and renewable energy to industrial sites and households. Depending on the insulating material used, these cables are typically classified into oil-filled (OF) cables made from paper and oil and cross-linked polyethylene (XLPE) cables made from XLPE. Compared to OF cables, XLPE cables pose a lower risk of environmental pollution and fire due to leakage. Therefore, XLPE cables have become increasingly popular as underground power transmission cables in recent years. Moreover, XLPE cables are replacing OF cables due to improved efficacy and longevity.
XLPE is polyethylene (PE) with a cross-linked structure that is produced by reacting PE with organic peroxide under high pressure to generate free radicals. These radicals create cross-links and a network structure in the polymer [
1]. XLPE is used as an electrical insulating material due to its hydrolysis resistance, high insulation properties, abrasion resistance, and excellent mechanical properties [
2,
3,
4,
5,
6]. In addition, it is used in piping systems for chemical storage, in heating and cooling systems, etc., due to its high extrusion speed and low cost.
When XLPE is exposed to the power cable operating environment, deterioration progresses and initial characteristics degrade. The main causes of XLPE cable deterioration are electrical [
7], thermal [
8], mechanical [
9], and environmental factors [
10], while the cable itself deteriorates due to aging and abnormal phenomena. Age deterioration is caused by heat generated by conductor resistance, and the typical abnormality is partial discharge (PD), which also generates heat. Therefore, determining the deterioration characteristics of XLPE when exposed to heat is necessary for cable operation and management.
Since XLPE is a polymer material, various analytical methods, such as physical, chemical, thermal, and electrical analyses, are used for characterization. Physico-mechanical methods, such as tensile strength and elongation, are generally used as standards for determining the longevity of XLPE insulating materials [
11]. Chemical analysis methods include Fourier-transform infrared (FTIR) spectroscopy, X-ray diffraction (XRD), etc., thermal analysis techniques are thermogravimetric analysis (TGA), differential scanning calorimetry (DSC), etc., and electrical analysis methods include breakdown voltage [
12,
13]. Many studies have been conducted on the characterization of XLPE cable insulation using these analysis methods. Typically, the characteristics of XLPE insulation were confirmed according to the operating period of a 110 kV power cable [
14]. As a result, it was confirmed that the XLPE insulation of cables operating for more than 10 to 15 years was noticeably deteriorated, and mechanical, chemical, and electrical properties were reduced. In addition, research was conducted to accurately determine the degree of cable deterioration through machine learning, such as through principal component analysis (PCA) and artificial neural network (ANN), using characteristic analysis data on deterioration [
15]. In order to use artificial intelligence to diagnose the deterioration of power cables in the future, accurate and multifaceted deterioration characterization of XLPE, the main insulating material of XLPE power cables, is required.
Accelerated degradation is generally performed to confirm the properties of polymers, including XLPE. Accelerated degradation tests are useful when a correlation exists with the behavior of materials exposed to natural conditions. However, in general, discrepancies are observed between natural and accelerated conditions. In the case of accelerated degradation, the results are often amplified. Additionally, natural exposures produce different results because degradation conditions are diverse and inconsistent. Nevertheless, artificially accelerated degradation has been adopted in various studies because it is useful for the qualitative assessment of degradation factors in materials [
16]. In addition, under overload or short circuit conditions, it can increase up to 25 °C. Therefore, there have been studies that have performed accelerated degradation at temperatures above 150 °C and investigated the effect of antioxidants added to XLPE degradation inhibition [
17].
In this study, accelerated thermal degradation was artificially performed to confirm the characteristic changes due to heat exposure of XLPE (the main insulation property of XLPE cables). In addition, the change in properties of XLPE exposed to heat was measured through tensile strength and chemical analysis. The results obtained through these analyses were used to derive correlations between the characteristic changes. Thus, we hope to contribute to the accurate and efficient management and operation of XLPE cables for power transmission.
3. Results and Discussions
Figure 2 illustrates the color change of an XLPE sample when heated. Without exposure to heat, XLPE is typically colorless or white. However, it was confirmed that XLPE undergoes a color change when exposed to heat. The sample exposed to 100 °C acquired a bright yellow color, while that exposed to 120 °C showed a darker yellow color. When exposed to 140 °C for an extended period of time, the XLPE changed color from orange and red to brown. The sample exposed to 160 °C turned briefly red before gradually changing to a dark brown color.
The differences in the morphologies of unaged and deteriorated XLPE were investigated using SEM, as shown in
Figure 3. In
Figure 3a, the unaged XLPE was confirmed to have a smooth surface. However, when exposed to heat, cracks were observed to form on the surface. In addition, it was confirmed that the XLPE surface changed to fine particles when aged at high temperatures for a long period of time, and microcracks also increased.
The compositions of XLPE were analyzed using SEM EDX, and the results are shown in
Table 1. The majority of XLPE is composed of carbon. However, it has been found that the oxygen content in XLPE increases as it ages because of oxygen and heat. And in XLPE after 60 days of degradation at 160 °C, the atomic percentage was confirmed to increase to about 11.3%.
Figure 4 shows the tensile strength measurement results of the deteriorated XLPE. The tensile strength of the unheated XLPE sample was 22.36 MPa, which decreased gradually with exposure to heat, accelerating as the temperature increased. After 60 days of exposure, the tensile strengths at 100, 120, 140, and 160 °C were 19.08, 17.62, 7.00, and 4.24 MPa, respectively. This experiment confirmed that the XLPE insulating material deteriorated when exposed to heat, resulting in a decrease in mechanical strength.
Figure 5 depicts the XRD patterns of the new and deteriorated XLPE. Since XLPE is a crystalline polymer, its crystal peak can be confirmed through XRD. The XLPE peaks were observed at 2θ = 23.13° and 24.45° and assigned to the (110) and (200) planes, respectively [
22]. As deterioration progressed, the intensity of this crystal peak decreased. To quantify this pattern, the degree of crystallinity (%) was calculated via XRD using the following equation [
23]:
where
xXRD is the degree of crystallinity (%);
area 1,
area 2, and
area 3 are the peak areas of the amorphous portion, (110) plane, and (200) plane, respectively. New XLPE was confirmed to have a degree of crystallinity (%) of approximately 37.14%. In addition, samples exposed to heat showed an overall decrease in the degree of crystallinity. The degree of crystallinity decreased and then slightly increased at both 100 and 120 °C (relatively low temperatures). This phenomenon was due to the additional crystallization of uncrystallized XLPE molecules by heat exposure. However, XLPE exposed to temperatures of 140 and 160 °C showed a rapid decrease in crystallization in both cases. The degrees of crystallinity (%) derived from the XRD of the sample that deteriorated for 60 days at each temperature were confirmed to be 34.07%, 33.48%, 14.34%, and 9.87%, respectively.
Figure 6 depicts the degree of swelling as a result of XLPE degradation. It is an important characteristic that indicates the degree of polymer crosslinking. The degree of swelling in the undegraded XLPE sample was 10.38%. However, it increased with exposure to high heat. An increase in the degree of swelling indicates that a large amount of solvent had penetrated into the sample. This phenomenon occurs when the cross-linking of XLPE is broken due to deterioration and the cross-linking density is lowered, thereby increasing solvent penetration.
Figure 7 shows the FTIR spectra of the thermally deteriorated XLPE material. Since XLPE is a PE-based material, a peak for C–H bonding was confirmed. The peaks for symmetric and asymmetric stretching vibrations of the –CH
2 group were identified at 2916 and 2848 cm
−1, respectively. Peaks corresponding to wagging vibration and the rocking methylene group of –CH
2 were identified at 1463 and 719 cm
−1, respectively [
24].
There were no significant changes in the XLPE FTIR results after exposure to heat at 100 and 120 °C for 60 days. However, when XLPE was exposed to 140 and 160 °C, a peak corresponding to the carbonyl functional group (C=O) was identified between 1650 and 1800 cm
−1. The absorbance of the C=O peak gradually increased with exposure to heat. In addition, the C=O peak was observed as being divided into three. The C=O peak at 1712 cm
−1, the lowest wavenumber, corresponds to carboxylic acid and ketones. Secondly, the peak appearing at 1732 cm
−1 corresponds to that present in aldehyde or ester. Finally, the peak observed at 1770 cm
−1 corresponds to carboxylic anhydride [
25]. At the beginning of degradation, the C=O peak absorbance at 1712 cm
−1 was the highest. However, as the period of exposure to heat increased, the C=O peak absorbance at 1731 cm
−1 increased, showing the highest value. Due to the deterioration of XLPE, bonds between C and H in the polymer chain were broken and carbon radicals were generated. Then, it reacted with oxygen to form peroxide, and each peroxide was decomposed into radicals. This alkoxy radical breaks other polymer bonds. Due to the continuous degradation reaction, various carbonyl structures were generated in the polymer chain. At this time, the carboxylic acid identified at 1712 cm
−1 was converted into an ester by reacting with another alkoxy radical, or ROH, formed from the decomposition of peroxide in the polymer chain generated during deterioration. Therefore, as the deterioration progressed, the absorbance detected at 1731 cm
−1 became higher than the carbonyl absorbance detected at 1712 cm
−1.
In the samples exposed to heat at 140 and 160 °C, a peak was observed between 1100 and 1250 cm
−1 due to the vibration of the C–O–C bond [
24]. The peak absorbance also increased with increasing exposure to heat. In addition, it was confirmed that the absorption peaks at 800 to 1128 cm
−1 and 1560 to 1650 cm
−1 increased with deterioration. These peaks correspond to the vinylidene and vinylene groups of the C=C bond, respectively [
24,
26]. Finally, as it was exposed to heat, the peak caused by the –OH functional group was confirmed and increased from 3000 to 3500 cm
−1 [
24,
26]. Alkoxy radicals (RO·) and hydroxyl radicals (·OH) generated during XLPE thermal deterioration react with alkyl radicals (R·) or hydrogen radicals (H·) to produce ROR or ROH [
27]. As a result, C–O–C and –OH peaks appeared in the FT-IR analysis.
It was necessary to convert the results obtained through the above FTIR analysis into quantitative values for deterioration. First, C=O and C–O–C functional groups, whose absorbance peaks increased with eterioration, were selected. These peaks were quantified by the following equations [
28]:
where
CI represents the carbonyl index and was calculated in two ways. The carbonyl peak area was calculated as the ratio of the methylene functional group peak area to the C–H stretching vibration peak area. In general, the carbonyl index is calculated using the ratio of the peak absorbance to the functional group [
29,
30]. However, in this study, as mentioned in the FT-IR results, the peak absorbance inversion occurred due to the deterioration of the C=O separation. This made it difficult to calculate the carbonyl index from the peak absorbance; therefore, it was calculated as an area ratio.
Figure 8 shows the carbonyl index-1 and -2 obtained by calculation, and
CI-1 and
CI-2 showed similar trends. There were almost no changes in
CI at 100 and 120 °C. However, the
CI of XLPE exposed to 140 and 160 °C increased with exposure time, and the value changed more rapidly as the temperature increased.
CI indicates an increase in the oxygen ratio in the XLPE material, which consisted of only carbon and hydrogen. An increase in the oxygen ratio can lead to polarity in the non-polar XLPE insulator (the main cause of a decrease in insulating performance).
The DSC thermograms of XLPE are shown in
Figures S1 (heating) and S2 (cooling), respectively, and the melting point (
Tm) and crystallization temperature (
Tc) trends are shown in
Figure 9. The
Tm and
Tc of the unaged XLPE were 104.71 °C and 94.74 °C, respectively, and there was little change in the
Tm and
Tc when the temperature was 100 and 120 °C. However, when exposed to temperatures of 140 and 160 °C,
Tm and
Tc moved to a low temperature after 18 days. Additionally, the
Tm and
Tc peak intensities decreased. To quantify this DSC result, the degree of crystallinity was calculated using the following equation:
where
xDSC (%) is the degree of crystallinity calculated through DSC, Δ
Hm is the fusion enthalpy, and Δ
Ho is the theoretical melting enthalpy for the complete crystallinity PE, Δ
Ho = 287.3 J g
−1 [
31].
The lamellar thickness was calculated using the Thompson–Gibbs equation [
32,
33]:
where
L is the lamellar thickness (nm),
σe is the free surface energy,
is the equilibrium melting temperature,
= 141.45 °C,
Tm is the melting point, and Δ
Hm is the fusion enthalpy.
The
xDSC and
L values are shown in
Figure 10, and the XLPE properties due to degradation obtained through DSC are presented in
Table 2. In the case of unheated XLPE, the
xDSC was confirmed to be 30.01%. There was only a slight increase compared to the initial value (no significant change) when the temperature was 100 and 120 °C. However, in the case of XLPE exposed to 140 and 160 °C,
xDSC reduction proceeded rapidly. The
xDSC values of XLPE for 60 days at each temperature were 34.42%, 38.59%, 15.81%, and 4.05%, respectively. This result was similar to the crystallinity obtained through XRD. In addition, the lamellar thickness of the unaged XLPE was 7.29 nm. The trend of change over temperature and time is similar to the above analysis. The
L values of aged XLPE for 60 days at each temperature were 7.17, 7.15, 3.91, and 3.06 nm, respectively. These analysis results indicate that the XLPE crystallinity becomes amorphous as aging progresses.
UV–vis spectroscopy is a useful method to analyze the optical properties of materials. Among the UV–vis spectra of the XLPE sample obtained in this study, the results for the 160 °C deteriorated XLPE are shown in
Figure 11. In unaged XLPE, the light was absorbed at 300 nm or less, and red-shifted to a longer wavelength as deterioration progressed. This phenomenon can be explained as follows. As shown in the results of the FTIR analysis, during degradation, C=O and C=C are created in the molecule, and as these bonds are continuously formed, a conjugation pi bond is produced [
34]. Therefore, the highest occupied molecular orbital (HOMO) rises and the lowest unoccupied molecular orbital (LUMO) falls, thereby reducing the gap between energy levels entailing red-shift with a longer wavelength. As the deteriorated XLPE is absorbed by the short wavelength in the visible region, the complementary color, red, is observed by the eye, as shown in
Figure 2. The higher the exposure temperature, the faster the formation of double bonds and of a conjugation pi bond within the molecular structure consisting of single bonds. Therefore, a rapid color change in XLPE occurred. Based on these results, the bandgap energy of XLPE was calculated through Tauc’s equation [
35]:
where
α,
h,
ν,
A, and
Ebg represent the absorption coefficient, Planck’s constant, light frequency, a constant, and bandgap energy, respectively. In XLPE not exposed to heat,
Ebg was confirmed to be 3.75 eV. The
Ebg decreased as deterioration progressed, and this trend became more pronounced as temperature increased. Furthermore, the
Ebg of the sample exposed at 160 °C for 60 days decreased to 2.05 eV. The bandgap energy reduction indicates that electrons are excited from the valence band to the conduction band by a small amount of energy, and electrons can move inside and on the surface of the insulator. This also means that the material’s insulating performance is reduced.
An increase in the number of C=C bonds means shortening the PE chain length because the C=C bond is shorter than the C–C bond. Additionally, this causes the PE density to increase. However, XLPE is a crystalline polymer. The more crystalline the structure, including the lamellar structure analyzed above, the higher the density and the lower the degree of swelling. Although the number of C=C bonds increases at 140 °C and 160 °C, the degree of swelling increases because the XLPE’s overall crystalline structure changes to an amorphous structure with degradation, as shown by the degree of crystallinity obtained from XRD and DSC. In other words, although the bond length is shortened by the C=C bond, the degree of swelling increases due to the lower overall degree of crystallinity.
The color change of the sample due to heat exposure was quantified with a yellowness index (
YI) according to ASTM D1925, with results shown in
Figure 12. The
YI obtained increased with temperature and time, and this trend corroborated the other results analyzed above.
The correlation with tensile strength was confirmed by analyzing changes in properties of XLPE when exposed to heat. Various properties, including tensile strength, were found to depend on exposure temperature and duration. Based on the analysis results, a correlation between tensile strength and various property changes was obtained, shown in
Figure 13, and each parameter is presented in
Table S1. The degree of crystallinity obtained by XRD and Tc showed a linear correlation. However, it was confirmed that the remaining factors had nonlinear correlations. Tensile strength showed a negative linear correlation with the carbonyl and yellowness indexes. On the other hand, tensile strength showed positive linear correlations with crystallinity,
Tm,
Tc, lamellar thickness, and bandgap. Among them, tensile strength showed a high linear correlation with
CI-1 and lamellar thickness. The coefficients of determination (
R2) were confirmed to be 0.9214 and 0.9250 for each temperature, respectively.
4. Conclusions
In this study, XLPE, which is used as the main insulator for underground transmission cables, was artificially exposed to heat and its characteristic changes were analyzed. XLPE changed from colorless to red and subsequently dark brown as deterioration progressed. This phenomenon was due to the red shift of the UV–vis spectra from the ultraviolet region to a longer wavelength. According to FTIR confirmation, the degradation of the C–C or C–H single-bond XLPE structure also resulted in the generation and increase in C=O and C=C bonds. The conjunction also occurred due to the resulting double bond. Thus, the bandgap, which is the interval between HOMO and LUMO, was reduced and could be quantified using Tauc’s equation.
The crystallinity of XLPE was obtained through XRD and DSC analyses, which confirmed that both results decreased similarly as the deterioration progressed. It was also possible to obtain a change in the thickness of the lamellar structure that also decreased with the change in crystallinities.
The tensile strength of XLPE also decreased with exposure to heat. Tensile strength was correlated with various property changing factors, such as crystallinity, carbonyl index, yellowness index, Tm, Tc, lamellar thickness, and bandgap. All showed a high linear and nonlinear correlation, and the coefficient of determination of the lamellar thickness, R2, was 0.9250, showing the highest linear correlation. Negative slope tensile strength factors included the carbonyl and yellowness indexes, while those with positive slopes were crystallinity, Tm, Tc, lamellar thickness, and bandgap.
XLPE has excellent properties as an insulation material, so it is currently used in high voltage power cables and demand is increasing. Additionally, as usage increases, the occurrence of breakdowns is also increasing. The results of this study are expected to be helpful in evaluating the condition of XLPE in power cables dismantled due to failure and to be instrumental in establishing insulation material management standards for determining the stable operation condition of XLPE cables. Additionally, the results of this study are expected to be used as basic data in the development of a diagnostic device for the condition of XLPE insulation in the future.