Evaluation of Long-Term Boiling Water Immersion as Artificial Aging Test of Silicone Rubber Insulators

Our country IRAN have long coastal area with length of more than 1830 km. Utility companies with 30 years of experience in using polymeric insulators in this very heavy polluted area. Operational history in these areas indicates many insulating problems. With the advancement of polymer insulation technology, the utilities want to select the longest life insulators for these areas. Laboratory studies accelerated testing of polymer insulators is the only available method to investigate the trend of insulator degradation during the aging process, because it takes a long time for noticeable changes and signs of aging to appear on insulators in the power grid. Different methods such as salt-fog, UV and boiling water are used to test artificial aging, each method has advantages and disadvantages. In our experience, boiling water is used as a low-cost method in compared to the other methods especially salt fog test. The aim of this research was to determine a shorter method for evaluation of SR insulators. In this paper, the accelerated aging of polymer insulators using salt-fog and boiling water immersion tests are investigated. During and after the aging test, the polymer insulator's electrical quality is analysed using leakage current measurement tools, partial discharge, and hydrophobicity. Surface changes are also analysed by Scanning Electron Microscopy (SEM), Energy Dispersive X-ray spectroscopy (EDX), Thermo Gravimetrical Analysis (TGA), and Fourier Transform Infrared Spectroscopy (FTIR) both qualitatively and quantitatively.


Introduction
One of the severe issues of using polymer insulation in contact with environmental and electrical stresses is the penetration of water and moisture, which change the mechanical and electrical properties of insulation. Therefore, the analysis of water penetration resistance by immersion test provides useful information about the polymer's quality [1].
In IEC test designation "IEC 62217", a boiling water immersion test is provided to evaluate the sealing of interface of core and housing insulators and check the core material with respect to its resistance to hydrolysis. In this test, samples shall be immersed in a tank containing 0.1% NaCl in distilled water and boiled for 42 h, and tested with steep-front impulse voltage [2].
According to the ASTM-D570 standard, the immersion test for measurement water absorption of plastics, was performed for 2 h in boiling water, then the samples were dried, and their weight was measured. Finally, the effect of different polymer compounds and weight loss were analysed [3].
For more Information, a 440-h immersion test was performed in boiling water, and it was found that moisture penetration increases the leakage current [4].
In one study, a layer of silicone rubber material was bonded to the epoxy resin without the use of primer material and tested for 100 h in boiling deionized water; then mechanical tests evaluated the quality of the insulation surface and the degradation in the layer between the epoxy and the silicone rubber after the aging process. It was found that if aging continues for a longer time, degradation increases because of hydrolyzing in boiling water [5]. In another study, 100-and 300-h immersion tests were performed to evaluate the aged insulators' adhesion. The adhesion quality of the insulators was directly related to the 300-h test results, and the longer time diffusion test could help to identify insufficient adhesion property and low hydrolytic resistance [6]. In the Chinese standard, the allowed leakage current for the 100-h test is 100 μA and for the 300-h test is 200 μA [7].
The quality of the insulator depends evidently on the adhesion between filler and resin. Low adhesion accelerates the aging processes caused by moisture penetration, partial discharges, and electro-mechanical stresses. The leakage current measurement and partial discharge analysis provide useful information about the aging state of insulators. Insulator changes during aging were measured several times from the virgin to the aged samples.
Being cheaper, availability, and reduction of repair and maintenance costs are the advantages of aging test in boiling water compared to other aging tests.
Boiling water immersion test time varies from 2 to 440 h according to the standards and research studies. The recommended initial time was 1000 h for comparison with the salt-fog standard test. The improved properties of the new polymer insulator increased, the aging process of the insulators to 3000 h or 18 weeks. This paper presents and compares the results of electrical analysis including leakage current, partial discharge, and hydrophobicity. Furthermore, the material investigations include SEM, EDX, TGA, and FTIR on virgin and aged samples during accelerated aging in salt-fog and immersion in boiling water test up to 3000 h.

Experimental Arrangement
20 kV Silicon Rubber (SiR) insulators with fillers of alumina trihydrate (ATH) from a three particular manufacturers with similar electrical specifications and different material specifications and manufacturers were selected A1, A2 & A3 for salt-fog and B1, B2 & B3 for immersion in boiling water test. To validate the test methods, two samples from manufacturer number 3 were used in each test. It was observed that the results for samples number 3 are approximately similar to each other (less than 10% difference in results).
The most popular test of artificial aging is the salt fog test. The salt fog test is performed according to IEC 61109 standard [8]. Figures 1 and 2 show the salt fog test configuration.
All insulators were mounted vertically in a 2 m × 2 m × 2 m chamber in salt fog test. The surface of insulators according to IEC 60507 was contaminated by a solution consists of 40 g kaolin and 10 g NaCl in 1000 g water, the contaminated solution was sprayed on the surface of the samples. Based on IEC 62217, the voltage stress was set to 20 mm/kV, which is equivalent to Unified Specific Creepage Distance (USCD) of 34.6 mm/kV. Therefore, for an insulator with a creepage distance of 710 mm, the test voltage was set to 20.5 kV. To obtain this voltage, a high voltage transformer with a ratio 380 V to 100 kV and 20kVA nominal power was used and a voltage of 78 V was applied to the input of the power transformer by a low voltage autotransformer.
In the boiling water immersion test, samples were placed in a stainless-steel chamber with a 50-L volume. The temperature of the boiling water chamber was controlled by a thermostat from 97 °C to mostly boiling point. Figure 3 shows the immersion in the boiling water test configuration.
One insulator from each sample was considered as a reference. Leakage current and hydrophobicity tests were performed biweekly, partial discharge tests for indication of penetration of moister to the interface of housing and core of insulator were performed every three weeks, and polymer chemical properties were tested at the beginning (for reference) and the end of the aging process.
Aging was stopped in the middle (after week 9) and at the end of tests (after week 18) for three weeks to evaluate the recovery properties of insulators. Therefore, the test time lasted for 24 weeks.

Aging Results
During and after the aging, the electrical properties and surface changes of the polymer insulators were investigated by the following analysis techniques.

Leakage Current Analysis
Leakage current analysis is used as appropriate information about insulator's assessment in recent studies. Reciprocal correlation between destruction factors and leakage current is only one reason for its many applications [10]. One of the acceptance criteria in aging tests is leakage current according to the IEC 62217 standard.
The performance of all samples were evaluated in the boiling water and salt fog tests in a high voltage laboratory by measuring the leakage current in the 20.5 kV voltage test and in salt fog chamber. After turning off the salt fog and boiling water testing devices, all the samples are washed, cleaned and dried in less than 1 h. Water conductivity was set to 4 μS/m at 25 °C during leakage current test in light fog condition for salt fog and boiling water samples. Figure 4 shows the typical circuit diagram for leakage current measurement.
The leakage current was measured after first nine weeks (1500 h) in the middle of the stressing time plus three weeks resting and second nine weeks (total of 3000 h of aging) at the end of the stressing time and by stopping the aging process for three weeks to evaluate the recovery properties of the polymer insulator. Figure 5 shows the leakage current results.
The leakage current of the sample number 1 increased faster than the other samples, due to the lower quality of materials and manufacturing technology. The maximum current for this sample was respectively 770 μA and 296 μA in the salt fog and immersion test after 18 weeks of stressing. According to the recorded leakage current for all samples, it was observed that in the aging test of boiling water, the maximum leakage current is approximately 40% lower than the salt fog test, due to further destruction in the salt fog test.
After three weeks rest in the middle and end of the stressing time, a reduction in leakage current in all samples was evident in the recovery properties of the polymer insulators. The recovery rate is affected by the composition of methyl groups and the presence of low molecular weight molecules that tend to expand on the surface. The main reason for the recovery of silicon rubber insulation in salt-fog test is a displacement of low molecular weight and in boiling water immersion test is reorientation of methyl-groups [11].
It was observed that the leakage current after three weeks of rest decreased approximately 80% in salt fog and 65% in boiling water samples, due to uniform surface aging in boiling water, the recovery properties of the insulators are less. Due to increasing the leakage current by increasing the boiling water test time, the electrical properties were further investigated to evaluate the mechanical properties, including the adhesion of the rod to the insulator body.

Leakage Current Harmonics
Harmonic analysis of leakage current can be considered as a criterion for assessing the state of polymer insulation [12].  Studies show that the low-frequency components of leakage current (fundamental, 3 rd and 5rd harmonic) can be better related to aging than the peak values. Harmonic analysis using Fourier Transform is illustrated in Fig. 6.
When the third harmonic component of the leakage current is smaller than the fifth, the maximum leakage current is limited, and no erosion was observed on the insulator surface. With increasing aging, it's third harmonic significantly increases with respect to the fifth harmonic [13].
As can be seen, the 3rd/5rd harmonic ratio in the sample 1 increased faster than the other samples. In sample A1, the third harmonic surpassed the fifth harmonic after the fifth and fifteenth weeks, and in sample B1 after the ninth and nineteenth weeks. For other samples, the third harmonic amplitude is less than the fifth. In the boiling water test, the 3rd/5rd harmonic ratio is approximately 25% lower than the salt fog test.
Harmonic and maximum leakage current changes were similar in both aging tests. When the third harmonic component of the leakage current is smaller than the fifth component, the maximum leakage current is limited and no erosion was observed on the insulator surface. By increasing the third harmonic component of the leakage current, an increase in the electrical discharge of the dry band and also corrosion was observed on the surface. It was observed that in the weeks when the leakage current of the samples is more than 200 μA, the third harmonic of the leakage current is more than the fifth.
Therefore, harmonic analysis of leakage current and the use of third to fifth harmonic ratio can be considered as a criterion for evaluating the condition of polymer insulation to determine the end of life of insulators and provide maintenance services. The maximum leakage current and harmonic analysis can be used to confirm the onset of aging.

Hydrophobicity Test
According to the STRI classification guide, the degree of the insulation surface's water repellence can be divided into seven hydrophobicity classes (HCs) [14]. HC1 is the highest water-repellent class, while completely hydrophilic surfaces are in the HC7 category.
The hydrophobicity test was performed weekly on all samples before leakage current measurement by water spraying after cleaning and drying the samples. Summary of the hydrophobicity analysis after salt fog and boiling water tests at 18 weeks of stressing and two rest times are presented in Figs. 7 and 8 to evaluate the recovery properties and compare the changes with the reference samples.
Test results show that silicone rubber initially had high hydrophobicity. This property decreased with increasing stressing time. As expected, silicon rubber can regain its hydrophobic properties after a period of rest; the polymeric insulators' hydrophobicity improved up to 2 classes after each rest period. In a situation where environmental factors are constantly changing and polymer insulators are sometimes subjected to severe stresses that cause aging; the recovery feature of polymer insulators enhances the performance of the power grid.
By comparing the hydrophobicity class of insulators in different weeks of aging with leakage current results, it was observed that the hydrophobicity class is directly correlated to the maximum leakage current.

Partial Discharge
The Partial discharge may be considered as the main cause of degradation of polymer insulators. A partial discharge test is performed to detect the cavities and surface erosion of insulator and mechanical defects during  Aging B3 the aging process [15]. In IEC 62,217 standard, defects in connections and moisture penetration are checked after 42 h immersion in boiling water test, by using steep-front impulse voltage test. In this article, to evaluate the electrical and mechanical properties of insulator after the stressing, a partial discharge test method was proposed.
The wideband method was used to record partial discharge pulses. Partial discharge measurements were performed regularly every three weeks at 20 kV voltage. The partial discharge results are presented in Table 1.
At the beginning of the aging process, partial discharge of all samples was below 1 pC and oscillating (due to ambient noise). Therefore, it can be concluded that all samples are free of defects that produce partial discharge, including surface destruction of the sample and internal mechanical damage to the insulator rod, same as reference samples.
At the end of the stressing time, the partial discharge of sample number 1 (A1 & B1) increased in salt fog and boiling water tests, which could be due to surface discharge, degradation, and loss of water hydrophobicity or moisture penetration into the insulator rod. After rest time, the partial discharge returned below 1 pC, which could be due to the improvement of surface properties and the recovery of silicon rubber on the insulator or dry-off of the moisture that has penetrated the fittings' connection, rod, and polymer insulator.
For sample B2, partial discharges reached 10 pC at the end of the stressing time. Also, the mechanical properties of this sample were weakened, which was bent by applying a vertical force to the rod insulator. Therefore, it can be concluded that the boiling water aging test affects simultaneously the electrical and mechanical properties of polymer insulators.

SEM Analysis
Analysis of surface changes was performed by Scanning Electron Microscopy (SEM). SEM analysis provides highresolution imaging, useful for a qualitative estimation of the type and extent of surface degradation or erosion [16].
Specimens with a 10 × 10 mm 2 were cut from the surface of reference and aged insulators and coated with a thin layer of gold under a vacuum situation. These specimens were analyzed with an electron microscope using a voltage of 20.0 kV. The results of SEM test in 5, 50, and 200 μm scales are presented in Fig. 9.
In salt fog tests, due to the dry band discharge, deep degradation was observed in the form of grooves, cavities, and scatters. For sample A1, degradation was observed in the entire surface, but for samples A2 and A3, only a part of the surface is destroyed and cracked. In the boiling water tests, surface degradation was observed uniformly.

EDAX Analysis
Energy Dispersive X-ray Analysis (EDAX) is an x-ray technique used to identify the housing material elemental composition during aging processes [17]. The data generated by EDAX analysis consist of spectra showing peaks corresponding to the elements making up the actual composition of the samples. The pieces of 20 × 20 mm 2 were cut from the reference and aged samples.
Silicone rubber is an elastomer composed of containing duplicate silicone oxygen (Si-O) backbone and two methyl groups (CH3) for each silicon atom and contains fillers to improve its properties. The methyl groups are responsible for the highly hydrophobic surface properties (Fig. 10).
ATH (Alumina Trihydrate) is used as a filler to obtain a combination with enhanced physical and chemical properties and improved long-term resistance [18].
When the polymer insulators aged due to surface degradation for thermal activity, methyl groups (CH3) oxidized and converted to O-H groups with high hydrophilic properties. After stressing, the carbon percentage had decreased, and the oxygen percentage increased. The EDAX analysis results are shown in Table 2, which represents the atomic percentage of elements on the surface of reference and aged samples.

Fig. 10 Silicone insulation composition
By decomposing ATH, water and white powder of aluminum oxide are produced. Therefore, the reduction of the Al element due to ATH decomposition can be introduced as a sign of aging. Due to the higher heat in the electrical discharge, Al decrease is more evident in the salt fog test.
In addition, after stressing due to the decomposition of PDMS bonds, a decrease in C element due to the production

TGA Analysis
Thermo Gravimetrical Analysis (TGA) is a technique that measures the weight changes of a sample during a thermal test. In this method, the temperature is increased at a constant rate, and the weight loss of volatile components is measured. TGA test is an effective method for assessing the thermal stability of polymer content and filler based on weight loss at different temperatures [19].
About 20 mg of samples were cut and heated from 25 °C to 800 °C at the rate of 10 °C per minute. In TGA analysis, material volume sampling was performed, while aging affects the surface of the samples. Thus, weight changes in different aging tests are not noticeable. Figure 11 shows the TGA and derivative thermogravimetry (DTG) diagrams. Two step weight losses are observed. The two-step weight loss percentage and residual weight are presented in Table 3.
The first step of weight loss from 230 °C -370 °C is associated with the release of water from the ATH filler and the production of Al2O3. The second step of weight loss from 370 °C -600 °C is related to the SiR decomposition.
DTG data show the rate of material degradation with temperature change. The minimum value of the DTG curve is at 340° C when water is released from ATH.
The volume of ATH filler in samples 1 and 2 are approximately equal, and in sample 3 is about 25% larger than the other two samples.
In the boiling water test has water uptake and possible leaching of filler particles during the water immersion, hence immersion boiling water test affects more on the filler. The lowest leakage current was recorded for sample 3 with the highest ATH volume. In the salt fog test with higher temperature during electrical discharge, the lowest leakage current was recorded for sample 2 with higher quality SiR.

FTIR Analysis
The polymeric insulators were tested by Fourier Transform Infrared Spectroscopy (FTIR). FTIR spectroscopy is an analytical technique to identify the initial structural bonding of the polymeric housing materials [20].
The peak corresponding to wavenumber 1270-1255 cm −1 belongs to the Si-CH3 group. The peaks of 1100-1000 cm −1 and 840-790 cm-1 belong to the Si-O-Si group and the Si-(CH3)2 group. The peaks of the wave numbers 3700-3200 cm −1 and 1640 cm −1 correspond to the hydroxyl groups [21]. FTIR analysis diagram of virgin samples and after the end of test by considered recovery effects is shown in Fig. 12.
If the samples aged, the polymeric groups corresponding to wavenumber 1270-1255, 1100-1000, and 840-790 cm −1 reduced, and the peaks corresponding to wavenumber 3700-3200 cm −1 increased. The ratio of FTIR peaks associated with Si-O-Si and O-H bonds can be introduced as the oxidation coefficient. Table 4 shows the changes in absorbance bonds of reference and aged samples. According to FTIR results on aged samples, methyl compounds (C-H) significantly decreased compared to reference samples. It was observed that after stressing of the polymer insulator, the bonds related to the spectra of 1255 to 1270, 1000 to 1100, and 780 to 840 decreased, and the bonds related to the spectra of 3200 to 3700 increased. The degree of oxidation of polymeric insulator compounds after stressing can be investigated by the maximum ratio of the Si-O-Si bond spectrum (1000 to 1100) related to the central PDMS polymer insulation bond to the O-H bond (3200 to 3700) of ATH. The change in 1260 wavelength absorption related to the Si-CH3 compound is a good indicator for measuring aging because this chain breaks easily.
The ratio of OH bond spectrum (3200 to 3700) to Si-O-Si bond spectrum (1000 to 1100) for reference samples 2 and 3 is less than sample number 1, which indicates the higher  however the changes in both tests are similar and indicate the overall quality of the insulation. The results of the harmonic analysis of leakage current and hydrophobicity class were consistent with the results of maximum leakage current. It can be said that when the leakage current exceeded 200 μA, the third to fifth harmonic ratio was more than 1, and the hydrophobicity class was above 5.

Summery
Defects in the rod's adhesion to the insulator and the penetration of moisture in sample B2 caused increasing leakage current compared to samples 2 and 3. Also, increase in the partial discharge of sample B2, was simultaneously observed with weakness in the joints and rod visually.
According to the analysis of partial discharge test results, weakness, and defects in the joints due to moisture penetration and severe aging on the polymer insulator's surface were observed.
EDAX test results show that the ratio of element C to O decreased with increasing stressing time.
FTIR analysis results show that the ratio of O-H to Si-O spectrum decreased by the aging of the polymer insulator. When this ratio dropped to more than 20%, the leakage current exceeded 200 μA.
Increase of maximum leakage current to more than 200 μA, third harmonic to fifth harmonic ration, and hydrophobicity to class 5 and decrease of O-H to Si-O bond ratio to more than 20% in FTIR analysis can be measured as indicators of aging.
The provided results indicate that the current experimental study presents a better understanding of the immersion boiling water aging phenomenon on the silicone rubber composite insulators. By increasing the time of boiling water test from 42 h, according to standard for checking the quality of insulator connections, to 3000 h, other characteristics such as the quality of the silicone rubber and filler can also be reviewed. In addition, the boiling water test requires less maintenance, in contrast to the high cost of salt fog test and the need for annual repairs due to salt corrosion.

Authors' Contributions Mohammad Akbari and Amir Abbas
Shayegani-Akmal conceived of the presented idea, developed the theory, performed the computations, discussed the results and contributed to the final manuscript. We confirm that the manuscript has been read and approved by all named authors and that there are no other persons who satisfied the criteria for authorship but are not listed. We further confirm that the order of authors listed in the manuscript has been approved by all of us.

Data availability
We understand that the Corresponding Author is the sole contact for the Editorial process (including Editorial Manager and direct communications with the office). We confirm that we have provided a current, correct email address which is accessible by the Corresponding Author.

Declarations
Ethical Approval We wish to confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome.

Consent to Participate
Informed consent was obtained from all individual participants included in the study.

Consent to Publication
We, the undersigned, give my consent for the publication of identifiable details, which can include photographs and diagrams and details within the text to be published in the Silicon Journal.

Competing Interests
We confirm that we have given due consideration to the protection of intellectual property associated with this work and that there are no impediments to publication, including the timing of publication, with respect to intellectual property. In so doing we confirm that we have followed the regulations of our institutions concerning intellectual property.