Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Detailed procedures of retrofilling transformers with FR3 natural ester and residual mineral oil content testing

Detailed procedures of retrofilling transformers with FR3 natural ester and residual mineral oil... INTRODUCTIONThe retrofilling of mineral‐oil‐immersed transformers with natural esters can extend the life of cellulose insulations [1, 2]. In China, fire safety and environmental friendliness of the liquid in transformers, especially in populous areas, are the main concerns of utilities. Although dozens of retrofillings have been performed by different utilities, the understanding of retrofilling procedures is still insufficient to minimise residual mineral oil and to understand its effect on the temperature rise of transformers [3].Soybean‐based natural esters have good biodegradability and flashpoints higher than 300°C, which is significantly higher than that of mineral oils [4–6]. In addition, the fire point of a natural ester is normally higher than 350°C, and it fulfils the K‐class of the less‐flammable requirements of the IEC 61039 standard [7, 8]. Globally, the highest voltage retrofilled transformer is a 400 kV transformer in Mexico [9]. In China, a 220 kV/16 MVA mineral oil transformer has been retrofilled with an FR3 natural ester [10]. The insulating liquid in the transformer is a mixed liquid of mineral oil and a natural ester. Annex B of the IEEE C57.147 standard is concerned with the effect of natural ester fluid mixed with mineral oil on the fire and flashpoints [11]. The fire point rapidly decreases to lower than 300°C when the mass ratio of mineral oil reaches approximately 7%–8% [12,13], which neither meets K‐class less‐flammable requirements nor utilises the advantages of natural esters. Thus, it is vital to control the mass ratio of the residual mineral oil in retrofilling transformers.Not only natural ester, the transformers also can be retrofilled with other insulation liquids. Fofana expound on the short‐term and long‐term effects of the content of residual liquid after retrofilling transformers filled with a perchloroethylene‐based liquid, with mineral oil and synthetic ester liquid. Although the filling and retrofilling liquid are different from natural ester, an obvious influence of viscosity and density was observed with increased residual liquid content from 1% to 8% [14]. Further research found that there were no significant influences on the dielectric properties at low mineral oil content of 3.95% after retrofilling mineral oil transformer with synthetic ester [15].C.P. McShane proved that the aging rate of insulation Kraft paper is significantly slower in natural ester than in conventional transformer oil through an experimental comparison of the thermal aging of mineral oil and natural ester [16] and performed further experiments [17] for clarification. Provided that the temperature rise of the natural ester transformer does not exceed the temperature rise of mineral oil by more than 21 K, cellulose insulation will last longer in natural ester than in mineral oil transformers [18]. In addition, Moore et al. [19] and Pillai et al. [20] found that the reduction and suppression of moisture buildup in cellulose reduce the aging rate. The Reedy Creek Improvement District reported a case of retrofilling a 67 kV transformer to extend its life and performance [21].Although retrofillings have been performed in many parts of the world, it is still unclear whether retrofilling mineral oil with natural esters can cause an elevated temperature rise. In the case of retrofilling a 66/11 kV 12.5 MVA power transformer with a natural ester, Rajaram Shinde found that the temperature rise is 10–12 K lower than when filling with mineral oil [22]. In addition, Santisteban conducted a simulation and found that when the insulating liquid velocity is lower than 0.78 kg/s, the hot spot temperature of the natural ester transformer becomes 9–11 K lower than that of the mineral oil transformer. In addition, when the insulating liquid velocity is 0.9 kg/s, the hot spot temperatures between the two transformers converge to the same value. When the insulating liquid velocity is higher than 1.0 kg/s, the hot spot temperature of the natural ester transformer is 2–3 K higher than that of the mineral oil transformer [23]. Lee et al. [24], Smith and Barry [25], and Yang et al. [26] retrofilled different transformers and concluded that the temperature rises are 4–6, 5–8, and 4.6 K higher than those of mineral oil transformers after retrofilling with natural ester. Although natural ester transformers have a higher temperature rise than mineral oil transformers in most cases, the higher temperature rise limits of natural ester transformers are considered [27].After retrofilling a transformer with a natural ester, the allowable duration of operation under 1.5, 1.8, and 2.0 of the rated current was increased by 48%, 27%, and 8%, respectively [28], indicating that it is beneficial to study the overload capability of transformers after retrofilling with natural ester. In Zhao et al. [29], two similar transformers filled with mineral oil and natural ester were inoperative. The natural ester has a higher breakdown voltage, and the insulating paper has a higher degree of polymerisation compared with the mineral oil transformer, and the transformer filled with natural ester had a 1.57 times overload event in its service history. In Scatiggio et al. [30], it was found that high temperature insulating materials (natural ester and thermally upgraded Kraft) can provide continuous overloading up to 150% of the nominal rating.Detailed procedures for retrofilling transformers with natural esters were implemented after discussion with the technical staff of Cargill, State Grid Shaanxi Electric Power, and Xian Xibian Zhong Te Electric Co., Ltd. An out‐of‐service S11‐M‐315/10 and a new S13‐M‐400/10 distribution transformers were selected to retrofill their mineral oil with FR3 to study the temperature rise of transformers after retrofilling as well as the characteristics of mixed oil. Before and after the retrofilling process, oil, temperature rise, and overload tests were performed to study the method of controlling the mass ratio of the residual mineral oil. In addition, tests were performed on different ratios of mineral oil in the natural ester to obtain a convenient method for confirming the ratio of residual mineral oil in the retrofilled transformer.MATERIALS AND METHODSMaterialsAn out‐of‐service S11‐M‐315/10 distribution transformer with a 2‐year service history and a new S13‐M‐400/10 distribution transformer were selected as sample transformers, and the main parameters are listed in Table 1. Both units were filled with the conventional mineral oil. The purpose of choosing two similar capacity and voltage transformers was to increase comparability. Considering that transformers with service history may have differences after retrofilling compared with the brand‐new transformers, a transformer in service and a brand new one was selected. The transformer with the service history had been in service for 2 years, which is long enough for oil to impregnate the insulating paper. Based on these considerations, the State Grid Corporation of China (SGCC) provided an out‐of‐service S11‐M‐315/10 distribution transformer and a new S13‐M‐400/10 distribution transformer.1TABLEMain parameters of the target transformerMain parametersS11‐M‐315/10S13‐M‐400/10Production data2016.022018.12Voltage/capacity10 kV/315 kVA10 kV/400 kVAVectorDyn11Dyn11Seal typeSealed/corrugated tankSealed/corrugated tankCooling methodONANONANInsulating resistance (HV to LV and earth)5000 MΩ5000 MΩInsulating resistance (LV to HV and earth)5000 MΩ5000 MΩInsulating resistance (HV to LV)5000 MΩ5000 MΩHV winding resistance (before/after retrofilling)3.045 Ω/2.941 Ω2.283 Ω/2.298 ΩLV winding resistance (before/after retrofilling)2.396 mΩ/2.317 mΩ1.489 mΩ/1.505 mΩNo‐load loss (before/after retrofilling)426 W/427 W383 W/384 WLoad loss (before/after retrofilling)3776 W/3753 W4436 W/4332 WAbbreviations: HV, high voltage; LV, low voltage.A #25 naphthenic mineral oil (Karamay #25 mineral oil, Karamay, China) and a natural ester FR3 (Cargill Bioindustrial, Shanghai, China) were used in this study.Retrofilling procedures in ChinaRetrofilling procedures are under discussion and are being published. The retrofilling procedures that we used in this experiment are presented in the flow chart in Figure A1. The details of the flowchart are discussed in the following section.Transformer check‐up before retrofillingThe transformer nameplate information should be recorded with a special focus on the weight of the insulating liquid on the nameplate.The compatibility of materials with FR3 natural esters and the lowest allowed ambient temperature should be considered. The pour point of FR3 is −21°C, which is higher than that of mineral oil. This issue may lead to the problem of starting up at extremely low ambient temperatures.The requirements of the inspection on the retrofilled transformer are listed in Table A1.Draining oilThe procedures for draining the mineral oil required that the ambient temperature be approximately 20°C and the relative humidity be less than 60% to avoid water entering the oil tank as well as to drain the water that may be condensed from the vapour in the oil conservator. After draining the oil, the oil‐level gauge hole should be closed to prevent the transformer body from being exposed to air.To reduce the mineral oil residue on the active part, 1–2 h of waiting is necessary. Sufficient waiting period can reduce residual mineral oil after retrofilling.FlushingThe transformer should be flushed from the oil inlet. When the transformer has an oil conservator, the highest point of the oil conservator should be used as the oil inlet. When the transformer has no oil conservator, the oil‐level gauge port should be used as the oil inlet. The hole in the oil inlet should not be fully closed because the gas inside the transformer is vented outwards.A new catheter was used to flush the active part and components with the natural ester at the two flanges of the removed bushings. The natural ester used in this procedure is usually approximately 10% of the total oil volume.Draining oil again and replacing the gasketsA large amount of mineral oil can remain in the bottom of the tank during the flushing procedure. The residual oil was drained again using the method described in Section 2.2.2. The gaskets used should meet the requirements for compatibility with natural esters. Silicone rubber, fluororubber, and fluorosilicone rubber seals can be used in transformers with higher thermal classes. Thereafter, the high‐voltage bushing should be reinstalled to ensure a good sealing condition.Retrofilling with natural ester and check‐upIn the retrofilling procedure, the pipe must be extended to the bottom of the transformer to minimise the generation of bubbles.The transformer should be left to stand in the warehouse for no less than 4 h after retrofilling to wait for the air bubbles in the ester liquid to be fully dissipated. After retrofilling, the insulating liquid inside the transformer becomes a mixture of natural ester and a small amount of mineral oil.Finally, the retrofilled transformer should be inspected again according to Table A1 to guarantee a successful retrofilling.Temperature rise tests and the overload testsTemperature rise tests were performed on both retrofilled transformers before and after retrofilling. The special overload tests were only performed on the S13‐M‐400/10 transformer.The temperature rise tests followed the short‐circuit method according to the IEC 60076‐2 standard [31]. The total test continued for 8.5 h. In the first stage, the total loss was applied until the change in the top oil temperature was less than 1 K/h for 3 h. In the second stage, the rated current was applied for 1 h. When the power was disconnected, the temperatures of the top oil and windings were measured.The top and bottom oil temperature was measured at the outlet and inlet of the pipeline in the temperature rise tests. One PT 100 type thermocouple was placed firmly on the surface of the inlet and outlet pipe. The average winding temperature was calculated by the winding resistance.The overload tests followed the technical guide of SGCC, and the test curve is shown in Figure 1. Before the overload tests, the temperature rise of the tested transformer was expected to reach a steady temperature rise under the rated load. First, the transformer was in operation for 3 h under 1.5 times of rated load. Whether the temperature rise was steady, the transformer test proceeded to the next 1.75 times of rated load, as shown in Figure 1. The total overload tests had five steps and then the load is cut off after finishing the total 10 h overload test.1FIGUREDistribution transformer overload test curve of the State Grid Corporation of China (SGCC)The temperature rise of the average windings, average oil, and top oil were measured at the end of the 1.5 times rated load. The temperature rise of the top oil was continuously measured during the test.Oil tests and prediction of mineral oil mass percentageGenerally, retrofilling procedures cannot remove all residual mineral oil, as absorbed oil exists in the solid insulation. Thus, the insulating liquid in the transformer is a mixture of natural esters and mineral oils. To meet the K‐class less‐flammable requirements, the fire point of the mixed oil should be higher than 300°C [7, 32]. Therefore, it is crucial to control the mass ratio of the mineral oil in the mixed oil. The properties of the insulating oil in the two transformers were measured before retrofilling, after retrofilling, and after 1 year.To confirm the exact mass percentage of mineral oil in the mixed oil after retrofilling with natural ester, an experiment with different mass percentages of mineral oil was designed. The mineral oil and natural ester were mixed and allowed to stand idle for 2 days to be mixed well. Specifically, 0%–10% mass percentage of mineral oil in 1% steps was blended, as the residual mineral oil is generally less than 10%.The tested properties included water content (IEC 60814), 90°C dielectric dissipation factor (IEC 60247), 2.5 mm breakdown voltage (IEC 60156), DGA (IEC 60599 for mineral oil, and IEEE C57.155 for natural ester), kinematic viscosity (ISO 2909), open‐cup flash point (ISO 2592), and closed cup flash point (ISO 2719).EXPERIMENT RESULTS AND ANALYSISTemperature rise tests and overload testsTemperature rise testsThe results of the temperature rise tests are shown in Tables 2 and 3, indicating that the two FR3 fluid retrofilled transformers have higher temperature rises than the mineral oil transformers. The top oil temperature rises were higher by 4.1 K (S11‐M‐315/10) and 5.0 K (S13‐M‐400/10). The average oil temperature rises were higher by 3.3 K (S11‐M‐315/10) and 4.0 K (S13‐M‐400/10).2TABLEResults of the S11‐M‐315/10 temperature rise testsTemperature rise forMineral oil transformerRetrofilled transformerLimits (IEC 60076‐14)Top oil (K)44.248.3≤60Average oil (K)35.438.7–High voltage winding (K)51.663.1≤65Low voltage winding (K)58.566.3a≤65aExceed the temperature rise limits for the thermal class of 105°C, but do not exceed the limits for the thermal class of 120°C.3TABLEResults of the S13‐M‐400/10 temperature rise testsTemperature rise forMineral oil transformerRetrofilled transformerLimits (IEC 60076‐14)Top oil (K)41.746.7≤60Average oil (K)33.337.3–High voltage winding (K)58.362.6≤65Low voltage winding (K)58.467.2≤65aExceed the temperature rise limits for the thermal class of 105°C, but do not exceed the limits for the thermal class of 120°C.The thermal class of cellulose paper immersed in mineral oil is 105°C. Thus, the limits in Tables 2 and 3 are the temperature rise limits for the thermal class of 105°C. However, according to IEC 60076‐14 standard Annex C [33], as shown in Table 4, the thermal class of the electrical insulation system can be increased by changing the liquid from mineral oil to natural ester, resulting in temperature rise limits of 90 and 75 K for the top oil and average windings, respectively. Therefore, the temperature rise values of the distribution transformers after retrofilling are up to the temperature rise limits in the IEC 60076‐14 standard Annex C. Considering that the thermal class of natural ester‐immersed cellulose paper is 120°C, although the temperature rise of the low‐voltage windings exceeded the temperature rise limit of mineral oil, the temperature rise of the low‐voltage windings remained within the limits for the ester filled transformers.4TABLEThermal class of mineral oil and natural ester according to IEC 60076‐14Insulating systemsKraft paper with mineral oilKraft paper with natural esterCellulose‐based thermal class(°C)105120Temperature rise limits for top oil (K)6090Temperature rise limits for average winding (K)6575Overload testsThe results of the temperature rise test for the average windings, average oil, and top oil at the end of the 1.5 times rated load (cooling phase) are shown in Table 5. The temperature of the top oil during the overload test is shown in Figure 2. The ambient temperature during the tests before and after retrofilling was 12.3 and 15.1°C, respectively. The ambient temperature was not measured in real time, but only once at the beginning of the test.5TABLETemperature rise at the end of the 1.5 times rated load of S13‐M‐400/10 transformerTemperature rise forMineral oil transformerRetrofilled transformerLimitsTop oil (K)84.091.7≤60Average oil (K)67.273.4–High voltage winding (K)95.8101.4≤65Low voltage winding (K)92.5103.6≤652FIGURETop oil temperature variations for the mineral oil and natural ester transformers under the given over‐load test curveThe retrofilled 400 kVA transformer had increased temperature rises for top oil of 7.7 K, average oil of 6.2 K, high voltage windings of 5.6 K, and low voltage windings of 11.1 K, when compared with the mineral oil transformer. These values are beyond the limits as discussed in Section 3.1.1.As shown in Figure 2, the mineral oil and retrofilled transformers exhibited the same trend as the top oil temperature change in the overload test. The highest temperature rise value of the top oil occurred at the end of the 2.0 times rated load. Furthermore, the highest temperature rise values of the mineral oil and retrofilled transformers were 109.1 and 121.9 K, respectively. Thus, the temperature values are beyond the limits as discussed in Section 3.1.1. Accordingly, a conventionally designed mineral oil transformer cannot fulfil the severe overloading test pattern of the SGCC. Thus, the high overload capacity transformer requires a more specific design, such as thermally upgraded paper or aramid paper for windings or a lower current density, as per the nameplate power rating.Oil tests and mixed oil testsOil testsThe results of the insulating oil tests of S11‐M‐315/10 and S13‐M‐400/10 are shown in Table 6. As previously explained, before retrofilling, the two transformers were filled with mineral oil, and mixed oil after retrofilling. The mixed oil was composed of a majority of natural ester and residual mineral oil. After 1 year, the retrofilled S11‐M‐315/10 operated for another 1 year, while the retrofilled S13‐M‐400/10 unit stood idle for 1 year. At this time, the residual mineral oil and natural ester liquids were expected to be mixed well, and the insulation materials were impregnated with natural esters.6TABLEOil tests results of the S11‐M‐315/10 and S13‐M‐400/10 transformersS11‐M‐315/10S13‐M‐400/10PropertiesBefore retrofillingAfter retrofillingAfter 1 yearBefore retrofillingAfter retrofillingAfter 1 yearFire point (°C)/352335/353338Dielectric dissipation factor @ 90°C0.00560.01270.08850.00220.03530.0793Breakdown voltage (2.5 mm) (kV)43.4461.3378.0063.5768.0665.31Kinematic viscosity @ 40°C (mm2/s)//32.97//32.67Open cup flash point (°C)//272//278Closed cup flash point (°C)//220//209Before retrofilling, the dielectric dissipation factor at 90°C and the breakdown voltage of the mineral oil were acceptable according to the IEC 60422 standard [34]. After retrofilling, the properties of the natural ester were well within the requirements of the IEC 62770 standard [35] for unused natural esters. Although dissipation factors at 90°C increased, the values were within the “good” limit as per IEC 62975. Robert discovered a similar phenomenon through continuous overload for 1 year [36]. This can be explained by the gradual homogenisation of the mineral oil in the cellulose paper with the natural ester. All characteristics meet the requirements in the IEC standard, which means successful retrofillings.In addition to the basic properties of mineral oil, dissolved gas was also included in the study. It is necessary to mention that the DGA results of the two transformers before retrofilling were acceptable as per IEC 60599 standard [37]. The DGA results of the S13‐M‐400/10 transformer are shown in Figures A2 and A3.It can be observed that except for H2 and C2H2, all gases increased after the overload test for the mineral oil transformer. However, all gas contents were still within the limits of IEC 60599. For the transformer after retrofilling, all gases except C2H2 increased after the overload test. The CO, H2, CH4, C2H4, and total hydrocarbon contents exceeded the threshold values for soybean‐based natural esters in IEEE C57.155. From the DGA perspective only, the overload test may not be suitable for the retrofilled transformer of traditional mineral oil design.After 1 year under sealed conditions, the C2H6 and total hydrocarbon contents of the retrofilled transformer (natural ester transformer) were even higher. Therefore, C2H6 and total hydrocarbons will increase naturally even when the transformer filled with the natural ester is offline. Thus, C2H6 and total hydrocarbons cannot be used as the key gases used for qualitative determination of fault types in transformers filled with natural esters.A comparison of the DGA of the S13‐M‐400/10 transformer shows that the overload test after retrofilling causes the gas content increased continuously. This increase could not be resolved by de‐energising the transformer. Therefore, it is prudent to consider whether to conduct an overload test after retrofilling. If necessary, the rated load multiple of the maximum load for the overload test must be considered.Prediction of mineral oil mass percentageUnused mineral oil and unused natural ester were used for samples to obtain the data in these experiments. Based on practical engineering considerations, the performance of the mixed insulating oil with mineral oil content in a certain linear range was tested in these experiments. The dielectric dissipation factor has a large discrete range [1], and the breakdown voltage is highly influenced by the moisture [38,39]. Thus, the kinematic viscosity, fire point, open cup flash point, and closed cup flash point were tested.The kinematic viscosity of the oil blend was measured every 10°C over a temperature range of 40–100°C, as shown in Figure 3. For the viscosity of the mass ratio of mineral oil from 0% to 6%, a clear linear change was observed. When it reaches 7%, there is a sudden drop in viscosity, followed by an unexpected increase at a ratio of 9%. These unusual phenomena can be seen in McShane's research [1] and leads to poor linearity of viscosity. Hence, it is better to use 0%–6% mineral oil ratios to describe the variation in viscosity with mineral oil ratios. This allows the prediction of only mineral oil ratios below 6%. In addition, the viscosities of the different mineral oil ratios were similar at the same temperature. Therefore, the viscosity at 60°C has the ideal fitting equation in the following form, with the highest coefficient of determination (R2) of 0.989:1y1=−0.234x+18.6\begin{equation}{y_1} = - 0.234x + 18.6\end{equation}3FIGUREKinematic viscosity of low mass ratio of mineral oil mixed with natural esterHere, x$x$ represents the mass ratio of mineral oil and y1${y_1}$ represents the kinematic viscosity at 60°C values.The results of the fire points are shown in Figure 4. Owing to the rapid decrease in the fire point of the mixed oil in relation to the percentage of mineral oil at the 5%–10% stage, the fitted curve should be divided into two parts. The first part is a slowly decreasing part with a fire point above 330°C. The second part is the rapidly decreasing part with a fire point below 330°C. The second part is the main discussion point, as the fire point will be lower than 300°C, and does not meet the K‐class less‐flammable requirements. Figure 4 compares the different fitted curves in the second part. Although the fitted curve using 6%–10% ratios of the mineral oil has the highest R2, the fire point of this fitted curve at a 5% ratio is 315°C. The fire point at the 5% ratio of the first part is 340°C, which is significantly higher than the fitted curve using 6%–10% ratios. The fire points at 5% ratio of the fitted curves using 5%–10% ratios and 5%–9% are 325°C and 328°C, respectively. Two fitted curves have similar fire points at a 5% ratio, but the latter has a higher R2. Thus, the ideal fitting equation for the fire point (y2) below 340°C is of the following form:2y2=−21.8x+437.6\begin{equation}{y_2} = - 21.8x + 437.6\end{equation}4FIGUREFire point of low mass ratio of mineral oil mixed with natural esterThe fitting equation for the fire point (y2) above 340°C is as follows:3y2=−2.57x+353.1\begin{equation}{y_2} = \ - 2.57x + 353.1\end{equation}where x$x$ represents the mass ratio of mineral oil and y2${y_2}$ represents the fire point values.The flash point has good linearity at a 0%–10% mass ratio of mineral oil in the mixed oil, as shown in Figure 5. The closed cup (y3) and open cup (y4) flash points are of the following form:4y3=−9.34x+255\begin{equation}{y_3} = - 9.34x + 255\end{equation}5y4=−13.8x+329\begin{equation}{y_4} = - 13.8x + 329\end{equation}5FIGUREFlash point of low mass ratio of mineral oil mixed with natural esterHere, x$x$ represents the mass ratio of mineral oil, and y3${y_3}$ and y4${y_4}$ represent the closed cup and open cup values.The kinematic viscosity at 60°C has an extremely high linearity with an R2 of 0.989. However, the dispersion of the fire and flash points is considerable. Thus, only kinematic viscosity and its fitted curve at 60°C were used to predict the residual mineral oil content in this study. Note that the change in the natural ester and mineral oil category may impact the parameters of the fitted curve and should refit the curve if necessary. In addition, the distribution of the fire point with the mass ratio of mineral oil is not exactly the same as the widely used curves according to McShane's research [1]. This is caused by the use of different mineral oils.DISCUSSIONOverload testFor the 400 kVA transformer, the hot spot temperature can be calculated according to Figure 2 and IEC 60076‐2 as the following formula:6Δθh=Δθo+Hg\begin{equation}\Delta {\theta _h} = \ \Delta {\theta _o} + Hg\end{equation}where Δθh$\Delta {\theta _h}$ represents the hot‐spot temperature rise, Δθo$\Delta {\theta _o}$ represents the top oil temperature rise, H$H$ represents the hot spot factor, and g$g$ represents the average winding gradient.The hot spot factor H is assumed to be 1.1. For the ONAN transformer up to 2500 kVA, the average liquid temperature rise above the ambient temperature may be considered as 80% of the top liquid temperature rise. Thus, the average winding gradient (g) was calculated using the following formula:7g=Δθw−0.8Δθo\begin{equation}g = \Delta {\theta _w} - 0.8\Delta {\theta _o}\end{equation}where Δθw$\Delta {\theta _w}$ represents the average winding temperature rise and Δθo$\Delta {\theta _o}$ represents the top oil temperature rise.The calculated hot‐spot temperatures are shown in Figure 6. The highest hot spot temperatures of mineral oil and natural ester are 149.0°C and 167.8°C, respectively. The closed‐cup flash point of the natural ester after retrofilling is 209°C, which is significantly higher than the hot spot temperature of 167.8°C. However, the closed cup flash point of mineral oil is 140°C, and the hot spot temperature is 9.0°C higher than the closed cup flash point, which can cause danger during the overload test. This indicates that retrofilling with natural ester can significantly improve the fire resistance of the transformer. In addition, a specially designed mineral oil transformer is required for an overload test. An overly rigorous overload test required by SGCC was performed.6FIGUREHot spot temperature variations for the mineral oil and natural ester transformers under the given over‐load test curveRetrofilling with a natural ester liquid can improve the overload capacity of a conventionally designed mineral oil transformer. The durations in which the temperatures of the top oil and the hot spot reached the required limits in IEC 60076–7 and IEC 60076‐14, respectively, with a rated load of 1.5 were calculated. The duration was chosen as the closest value, and the results are listed in Table 7.7TABLEOverload capacity of the 1.5 times rated load before and after retrofillingTop oil temperature (°C)Hot spot temperature (°C)400 kVA transformer filled withAmbient temperature (°C)Measured valueLimitsCalculated valueLimitsaTime to reach limit (min)Mineral oil12.391.6105119.2120100Natural ester15.197.9105130.8130130aLimits of mineral oil required in IEC 60076‐7 and limits of retrofilled transformer required in IEC 60076‐14.The hot spot temperatures of both the mineral oil and retrofilled transformers reached their limits prior to those of the top oil temperatures. After retrofilling with natural ester, it took 130 min to reach the temperature limit under 1.5 times rated load, which is 30% higher than when the same transformer is filled with mineral oil. Therefore, the overload capacity of the 1.5 times load was improved by retrofilling. In addition, the overloading test pattern of SGCC is significantly difficult for conventionally designed transformers, thus requiring a special design.Prediction of mineral oil mass percentageBecause the kinematic viscosity at 60°C has a good linearity with respect to the mass ratio of mineral oil, it is possible to predict the mass ratio of mineral oil in the retrofilled transformers after retrofilling with natural ester.The prediction results of the residual mineral oil content using the kinematic viscosity are shown in Figure 7. The prediction uses prediction bands at the 95% confidence level. The red bands are the prediction bands, thus the provided kinematic viscosity at 60°C has a 95% confidence in the prediction bands.7FIGUREPrediction bands at 95% confidence level of kinematic viscosity at 60°C and the ratio of residual mineral oil predictionFor the S11‐M‐315/10 transformer, the kinematic viscosity at 60°C is 18.31 cSt. The prediction result of the mass ratio of residual mineral oil using kinematic viscosity at 60°C is 0.53% to 1.96% at a 95% confidence level, with an average level of 1.24%. For the S13‐M‐400/10 transformer, the kinematic viscosity at 60°C is 18.12 cSt. The prediction result of the mass ratio of residual mineral oil using kinematic viscosity at 60°C is 1.39%–2.75% at a 95% confidence level, with an average level of 2.05%.The fire points of the S11‐M‐315/10 and S13‐M‐400/10 transformers were 335°C and 338°C, respectively. Thus, according to Equation (2), the mass ratios of the two transformers were 4.7% and 4.5%, respectively. A similar result can be calculated by the flash point according to Equations (4) and (5) for the two transformers by 3.9% and 4.3%, respectively.Owing to the low linearity and piecewise equation of fire point and flash point, it is not recommended to predict the mass ratio of residual mineral oil using fire points and flash points, but to judge whether the insulating oil meets the K‐class less‐flammable requirements or not. The insulating oil of two retrofilled transformers remain a K‐class less‐flammable liquid.Predictions using the prediction bands at a 95% confidence level for kinematic viscosity provided the most favourable results. The prediction results of the residual mineral oil mass ratio after retrofilling were 1.24% for the S11‐M‐315/10 transformer and 2.05% for the S13‐M‐400/10 transformer. This mass ratio of residual mineral oil is significantly lower than the 7% expected residual mineral oil from [1], which should be attributed to the detailed retrofilling guidance presented in this study.CONCLUSIONDetailed procedures for retrofilling 10 kV distribution transformers with natural esters were described in this study. A method to predict the residual mineral oil content after retrofilling mineral oil transformers with natural esters was proposed using kinematic viscosity at 60°C. This method addresses the issue of testing residual mineral oil content. The two retrofilled transformers had mass ratios of residual mineral oil of 1.24% for the S11‐M‐315/10 transformer and 2.05% for the S13‐M‐400/10 transformer. The mass ratio of mineral oil was believed to be approximately 7% from previous studies.For the 315 and 400 kVA retrofilled transformers, the temperature rises at rated load of top oil increased by 4.1 and 5.0 K, average oil by 3.3 and 4.0 K, and low voltage windings by 7.8 and 8.8 K, respectively, compared with mineral oil transformers. According to the IEC 60076‐14 standard, the thermal class of the oil‐paper insulation system can be improved by changing the oil from mineral oil to natural ester, thus resulting in the temperature rise limits of the top oil and the average windings for the natural ester‐Kraft paper system to reach 90 and 75 K, respectively. Thus, the obtained temperatures using the traditional temperature rise test can meet the limits of retrofilling.Retrofilling with natural ester can significantly improve the fire resistance of a mineral oil transformer. The results of the special overload test of the 400 kVA transformer show that the highest hot spot temperature of mineral oil was 149.0°C, which is 9.0°C higher than that of the closed cup flash point, causing danger during the overload test. The closed‐cup flash point of the natural ester after retrofilling is 209°C, which is significantly higher than the hot spot temperature. This indicates that a specially designed mineral oil transformer cannot fulfil the severe overloading pattern of SGCC, even with retrofilling. Thus, transformers with a high overload capacity need more specific designs.After retrofilling with natural ester, it took 130 min to reach the temperature limits under 1.5 times rated load, and the duration was 30% longer than when the same transformer was filled with mineral oil. The overload capacity of the 1.5 times load can therefore be improved by retrofilling.ACKNOWLEDGEMENTSThe authors would like to express their gratitude to Dong Han of Xian Xibian Zhong Te Electric CO, Ltd. and Bin Wei of State Grid Shaanxi Electric Power Company for their technical support in this study.CONFLICT OF INTERESTThe authors declare no conflict of interest.DATA AVAILABILITY STATEMENTThe data that support the findings of this study are available from the corresponding author upon reasonable request.REFERENCESMcShane, C.P., Luksich, J. & Rapp, K.J.: Retrofilling aging transformers with natural ester based dielectric coolant for safety and life extension. In: Cement Industry Technical Conference, TX, US, 4–9 May 2003Pillai R., Havaldar F., Chitnis C.: Natural esters for life and capacity enhancement of distribution transformers. CIRED ‐ Open Access Proceedings Journal. 2017(1), 355–358 (2017)Zheng, Z.Y., Zhang, J.Q., et al.: Performance change analysis of replacing natural ester insulating oil with mineral insulating oil distribution transformer. Transformer 56(6), 27–30 (2019)Liao, R.J., et al.: A comparative study of thermal aging of transformer insulation paper impregnated in natural ester and in mineral oil. Eur. Trans. Electr. Power 20(4), 518–533 (2010)Hosier, I.L., et al.: Aging of biodegradable oils and assessment of their suitability for high voltage applications. IEEE Trans. Dielectr. Electr. Insul. 18(3), 728–738 (2011)Yu, H., et al.: Dielectric and physicochemical properties of mineral and vegetable oils mixtures. In: IEEE International Conference on Dielectric Liquids, Manchester, UK, 25–29 June 2017IEC 61039 : Classification of insulating liquids. IEC (2008)Stockton, D.P., Bland, J.R., et al.: Seed‐oil‐based coolants for transformers. IEEE Ind. Appl. Mag. 15(1), 68–74 (2008)Edmund, L., et al.: Cargill‐sustainable electrical energy using natural ester. In: 2015 Transformer Life Management (TLM) International Conference Shanghai, Shanghai, China, November, 2015Huang, Z.Q., Wang, R.F., et al.: Test comparison study on a natural ester retro‐filled 220kV power transformer. In: IEEE International Conference on High Voltage Engineering and Application, Beijing, China, 6–10 September 2020IEEE C57:147: IEEE guide for acceptance and maintenance of natural ester fluids in transformers. IEEE (2008)Asano, R., Page, S.A.: Reducing environmental impact and improving safety and performance of power transformers with natural ester dielectric insulating fluids. IEEE Trans. Ind. Appl. 50(1), 134–141 (2013)Josken, J., Wareham, D.: Seed based oil as an alternative to mineral oil. In: Rural Electric Power Conference, Scottsdale, AZ, 25–25 May, 2004Fofana, I., Wasserberg, V., Borsi, H., Gockenbach, E.: Preliminary investigations for the retrofilling of perchlorethylene based fluid filled transformers. IEEE Trans. Dielectr. Electr. Insul. 9(1), 97–103 (2002)Fofana, I., Borsi, H., Gockenbach, E.: Oil filled transformer retrofilled with ester liquid—Facts and arguments. In: 15th International Symposium on High‐Voltage Engineering. ISH, Ljubljana, Slovenia, August 2007McShane, C.P., et al.: Aging of paper insulation retrofilled with natural ester dielectric fluid. In: 2003 Annual Report Conference Electrical Insulation Dielectric Phenomena CEIDP, Albuquerque, NM, 19–22 October 2003McShane, C.P., et al.: Natural ester dielectric fluid development. In: 2005/2006 IEEE/PES Transmission Distribution Conference Exhibition, Dallas, TX, 21–24 May 2006McShane, C.P., et al.: Some considerations for new and retrofil applications of natural ester dielectric fluids in medium and large power transformers. In: 2005/2006 IEEE/PES Transmission and Distribution Conference and Exhibition, Dallas, May 2006Moore, S., Kevin, R., Ramona, B.: Transformer insulation dry out as a result of retrofilling with natural ester fluid. In: PES Transmission Distribution Conference Exhibition 2012, Orlando, 21–24 May, 2012Pillai, R., et al.: DGA of natural ester filled transformers: Experience with retrofilled and new transformers. In: IEEE International Conference Dielectric Liquids, Manchester, UK, 25–29 June 2017Murphy, J.R., Graham, J.: Distribution utility experience with natural ester dielectric coolants. In: IEEE Power Energy Society General Meeting, Calgary, Canada, 26–30 July 2009Shinde, R.: Condition monitoring of a retro‐filled power transformer by natural ester Envirotemp FR3 fluid. In: International Conference Cond. Assessment Techniques Electrical Systems, Rupnaggar, Punjab, India, 16–18 November 2017Santisteban, A., et al.: Thermal analysis of natural esters in a low‐voltage disc‐type winding of a power transformer. In: IEEE International Conference Dielectric Liquids, Manchester, UK, 25–29 June 2017Lee, B., Jeong, J., Song, I.: The characteristics of winding temperature for the natural ester filled transformer up to and including 500 kVA. In: Annual Report Conference Electric Insulation Dielectric Phenomena, Quebec, Canada, 26–29 October 2008Smith, S.D., Barry, L.B.: Design and test experience with natural ester fluid for power transformers update. In: 2009 IEEE Power Energy Society General Meeting, AB, Canada, 26–30 July 2009Yang, T., et al.: Feasibility of directly changing natural ester insulating oil for conventional mineral insulating oil‐filled distribution transformer. Insul. Mater. 51(2), 39–43 (2008)Frimpong, G.K., Oommen, T.V., Asano, R.: A survey of aging characteristics of cellulose insulation in natural ester and mineral oil. IEEE Electr. Insul. Mag. 27(5), 36–48 (2011)Wang, R.F., et al.: Study on load performance of distribution transformer before/after filling vegetable insulating oil instead of mineral oil. Transformer 56(4), 78–80 (2019)Zhao, Y.H., Qian, Y.H., et al.: In‐service ageing comparison study of natural ester and mineral oil filled distribution transformers. In: IEEE International Conference Dielectric Liquids, Roma, Italy, 23–27 June 2019Scatiggio, F., Pepe, F., et al.: Increased loadability of transformers immersed in natural ester. In: IEEE International Conference Dielectric Liquids, Roma, Italy, 23–27 June 2019IEC 60076‐2: Power transformers—Part 2: Temperature rise for liquid‐immersed transformers. IEC (2011)Mazzaro, M., et al.: Fire simulation tests of mineral oil and natural esters transformers. In: IEEE International Conference Dielectric Liquids, Roma, Italy, 23–27 June 2019IEC 60076‐14: Power transformers—Part 14: Liquid‐immersed power transformer using high‐temperature insulation materials. IEC (2013)IEC 60422: Mineral insulating oils in electrical equipment—Supervision and maintenance guidance. IEC (2013)IEC 62770: Fluids for electrotechnical applications—Unused natural esters for transformers and similar electrical equipment. IEC (2013)Robert, C.B., et al.: Evaluation of natural ester retrofilled transformers after one year of continuous overload. In: IEEE Electrical Insulation Conference, Calgary, Canada, 16–19 June 2019IEC 60599 : Mineral oil‐filled electrical equipment in service—Guidance on the interpretation of dissolved and free gases analysis. IEC (2015)Hiramatsu, Y., Kamidani, K., Muramoto, Y.: Effect of water on AC breakdown properties of vegetable‐oil‐based insulating fluid mixed with mineral oil. In: Proceeding International Symposium Electrical Insulation Materials, Toyohashi, Japan, 11–15 September 2017Boris, H., et al.: Dielectric behaviour of silicone and ester fluids for use in distribution transformers. IEEE Trans. Dielectr. Electr. Insul. 26(4), 755–762 (1991)AAPPENDIXTable A1 presents the requirements of retrofilled transformer inspection at the beginning and ending of the retrofill. Figure A1 shows the flow chart of retrofilling transformer with FR3 natural ester. Figures A2 and A3 shows dissolved gas content of S13‐M‐400/10 transformer before and after retrofilling.There are four main parts in the retrofilling procedures: (1) Preparation before retrofilling; (2) draining oil procedures. (3) flush procedures. (4) retrofill procedures.In the first procedure ‘preparation before retrofilling’, inspections and test items of transformer in Table A1 should be inspected and tested and the results should meet the requirements of indicated standard. The ambient temperature of approximately 20°C and relative humidity of not higher than 60% are required. The gas should be released for transformers with gas relay. Oil should be drained for transformers with an oil conservator. Valve and drain plug should be opened before and after draining oil for transformers with butterfly valve. These additional preparations are necessary for different transformer constructions to ensure a successful retrofill.In the second procedure ‘draining oil procedures’, oil should be drained from the oil vent valve for transformers with oil vent valve, or from the oil level gauge for transformers without oil vent valve. Note that 1–2 h waiting is valuable. Sufficient waiting can reduce residual mineral oil after retrofilling. After waiting, the residual oil under the tank should be drained through the drain valve at the bottom of oil tank, otherwise a suitable angle is necessary to reduce the residual mineral oil.In the third procedure ‘flush procedures’, if the transformer cannot be flushed and retrofilled at once after the draining oil procedures, the transformer should be filled with dry nitrogen gas to cut off with moisture and oxygen in the air. Flush the windings, core, and oil tank internal surface with natural ester. Then, another hour's waiting can reduce the oil generated by flushing.In the fourth procedures ‘retrofill procedures’, the transformer should be retrofilled with the natural ester to the bottom. Again, the inspections and test items in Table A1 should meet the requirements due to the change of insulating system. Qualified test results mean a successful retrofill.A1FIGUREFlow chart of retrofilling transformer with FR3 natural esterA2FIGUREDissolved gas content of the S13‐M‐400/10 transformer filled with mineral oil (before retrofilling)A3FIGUREDissolved gas content of the S13‐M‐400/10 transformer filled with the natural ester (after retrofilling)A1TABLERequirements of retrofilled transformer inspectionApplicable transformer for retrofillingOnline transformerOffline transformerNew transformerInspection and test items (Standard)Before retrofillingAfter retrofillingBefore retrofillingAfter retrofillingBefore retrofillingAfter retrofillingTransformer discharge (IEC 60076‐3)RequiredVisual inspection and cleaningRequiredRequiredRequiredRequiredRequiredRequiredInsulation resistance measurement (IEC 60076‐1)RequiredRequiredRequiredRequiredRequiredRequiredWinding DC resistance measurement (IEC 60076‐1)RequiredRequiredRequiredRequiredVoltage ratio measurement and connection group label verification (IEC 60076‐1)RequiredRequiredRequiredRequiredRequiredRequiredNo‐load current and No‐load loss (IEC 60076‐1)OptionalOptionalShort‐circuit impedance and load loss (IEC 60076‐1)OptionalOptionalApplied voltage withstand test (IEC 60076‐3)OptionalaRequiredaOptionalaRequiredaOptionalaRequiredbInduced voltage test (IEC 60076‐3)RequiredLightning impulse voltage test (IEC 60076‐3)OptionalInsulating oil test:Breakdown voltage (IEC 60156)Dielectric dissipation factor (IEC 60247)Water content (IEC 60814)RequiredRequiredRequiredRequiredRequiredRequiredLeak test (IEC 60076‐1)RequiredRequiredRequiredDissolved gas analysis (IEC 60599 and IEEE C57.155)OptionalOptionalOptionalOptionalOptionalOptionalaThe applied voltage during the high voltage test is 80% of the standard voltage.bThe applied voltage during the high voltage test is 80%‐100% of the standard voltage. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png "IET Generation, Transmission & Distribution" Wiley

Detailed procedures of retrofilling transformers with FR3 natural ester and residual mineral oil content testing

Loading next page...
 
/lp/wiley/detailed-procedures-of-retrofilling-transformers-with-fr3-natural-mUH8DwHPFw

References (7)

Publisher
Wiley
Copyright
© 2022 The Institution of Engineering and Technology
eISSN
1751-8695
DOI
10.1049/gtd2.12402
Publisher site
See Article on Publisher Site

Abstract

INTRODUCTIONThe retrofilling of mineral‐oil‐immersed transformers with natural esters can extend the life of cellulose insulations [1, 2]. In China, fire safety and environmental friendliness of the liquid in transformers, especially in populous areas, are the main concerns of utilities. Although dozens of retrofillings have been performed by different utilities, the understanding of retrofilling procedures is still insufficient to minimise residual mineral oil and to understand its effect on the temperature rise of transformers [3].Soybean‐based natural esters have good biodegradability and flashpoints higher than 300°C, which is significantly higher than that of mineral oils [4–6]. In addition, the fire point of a natural ester is normally higher than 350°C, and it fulfils the K‐class of the less‐flammable requirements of the IEC 61039 standard [7, 8]. Globally, the highest voltage retrofilled transformer is a 400 kV transformer in Mexico [9]. In China, a 220 kV/16 MVA mineral oil transformer has been retrofilled with an FR3 natural ester [10]. The insulating liquid in the transformer is a mixed liquid of mineral oil and a natural ester. Annex B of the IEEE C57.147 standard is concerned with the effect of natural ester fluid mixed with mineral oil on the fire and flashpoints [11]. The fire point rapidly decreases to lower than 300°C when the mass ratio of mineral oil reaches approximately 7%–8% [12,13], which neither meets K‐class less‐flammable requirements nor utilises the advantages of natural esters. Thus, it is vital to control the mass ratio of the residual mineral oil in retrofilling transformers.Not only natural ester, the transformers also can be retrofilled with other insulation liquids. Fofana expound on the short‐term and long‐term effects of the content of residual liquid after retrofilling transformers filled with a perchloroethylene‐based liquid, with mineral oil and synthetic ester liquid. Although the filling and retrofilling liquid are different from natural ester, an obvious influence of viscosity and density was observed with increased residual liquid content from 1% to 8% [14]. Further research found that there were no significant influences on the dielectric properties at low mineral oil content of 3.95% after retrofilling mineral oil transformer with synthetic ester [15].C.P. McShane proved that the aging rate of insulation Kraft paper is significantly slower in natural ester than in conventional transformer oil through an experimental comparison of the thermal aging of mineral oil and natural ester [16] and performed further experiments [17] for clarification. Provided that the temperature rise of the natural ester transformer does not exceed the temperature rise of mineral oil by more than 21 K, cellulose insulation will last longer in natural ester than in mineral oil transformers [18]. In addition, Moore et al. [19] and Pillai et al. [20] found that the reduction and suppression of moisture buildup in cellulose reduce the aging rate. The Reedy Creek Improvement District reported a case of retrofilling a 67 kV transformer to extend its life and performance [21].Although retrofillings have been performed in many parts of the world, it is still unclear whether retrofilling mineral oil with natural esters can cause an elevated temperature rise. In the case of retrofilling a 66/11 kV 12.5 MVA power transformer with a natural ester, Rajaram Shinde found that the temperature rise is 10–12 K lower than when filling with mineral oil [22]. In addition, Santisteban conducted a simulation and found that when the insulating liquid velocity is lower than 0.78 kg/s, the hot spot temperature of the natural ester transformer becomes 9–11 K lower than that of the mineral oil transformer. In addition, when the insulating liquid velocity is 0.9 kg/s, the hot spot temperatures between the two transformers converge to the same value. When the insulating liquid velocity is higher than 1.0 kg/s, the hot spot temperature of the natural ester transformer is 2–3 K higher than that of the mineral oil transformer [23]. Lee et al. [24], Smith and Barry [25], and Yang et al. [26] retrofilled different transformers and concluded that the temperature rises are 4–6, 5–8, and 4.6 K higher than those of mineral oil transformers after retrofilling with natural ester. Although natural ester transformers have a higher temperature rise than mineral oil transformers in most cases, the higher temperature rise limits of natural ester transformers are considered [27].After retrofilling a transformer with a natural ester, the allowable duration of operation under 1.5, 1.8, and 2.0 of the rated current was increased by 48%, 27%, and 8%, respectively [28], indicating that it is beneficial to study the overload capability of transformers after retrofilling with natural ester. In Zhao et al. [29], two similar transformers filled with mineral oil and natural ester were inoperative. The natural ester has a higher breakdown voltage, and the insulating paper has a higher degree of polymerisation compared with the mineral oil transformer, and the transformer filled with natural ester had a 1.57 times overload event in its service history. In Scatiggio et al. [30], it was found that high temperature insulating materials (natural ester and thermally upgraded Kraft) can provide continuous overloading up to 150% of the nominal rating.Detailed procedures for retrofilling transformers with natural esters were implemented after discussion with the technical staff of Cargill, State Grid Shaanxi Electric Power, and Xian Xibian Zhong Te Electric Co., Ltd. An out‐of‐service S11‐M‐315/10 and a new S13‐M‐400/10 distribution transformers were selected to retrofill their mineral oil with FR3 to study the temperature rise of transformers after retrofilling as well as the characteristics of mixed oil. Before and after the retrofilling process, oil, temperature rise, and overload tests were performed to study the method of controlling the mass ratio of the residual mineral oil. In addition, tests were performed on different ratios of mineral oil in the natural ester to obtain a convenient method for confirming the ratio of residual mineral oil in the retrofilled transformer.MATERIALS AND METHODSMaterialsAn out‐of‐service S11‐M‐315/10 distribution transformer with a 2‐year service history and a new S13‐M‐400/10 distribution transformer were selected as sample transformers, and the main parameters are listed in Table 1. Both units were filled with the conventional mineral oil. The purpose of choosing two similar capacity and voltage transformers was to increase comparability. Considering that transformers with service history may have differences after retrofilling compared with the brand‐new transformers, a transformer in service and a brand new one was selected. The transformer with the service history had been in service for 2 years, which is long enough for oil to impregnate the insulating paper. Based on these considerations, the State Grid Corporation of China (SGCC) provided an out‐of‐service S11‐M‐315/10 distribution transformer and a new S13‐M‐400/10 distribution transformer.1TABLEMain parameters of the target transformerMain parametersS11‐M‐315/10S13‐M‐400/10Production data2016.022018.12Voltage/capacity10 kV/315 kVA10 kV/400 kVAVectorDyn11Dyn11Seal typeSealed/corrugated tankSealed/corrugated tankCooling methodONANONANInsulating resistance (HV to LV and earth)5000 MΩ5000 MΩInsulating resistance (LV to HV and earth)5000 MΩ5000 MΩInsulating resistance (HV to LV)5000 MΩ5000 MΩHV winding resistance (before/after retrofilling)3.045 Ω/2.941 Ω2.283 Ω/2.298 ΩLV winding resistance (before/after retrofilling)2.396 mΩ/2.317 mΩ1.489 mΩ/1.505 mΩNo‐load loss (before/after retrofilling)426 W/427 W383 W/384 WLoad loss (before/after retrofilling)3776 W/3753 W4436 W/4332 WAbbreviations: HV, high voltage; LV, low voltage.A #25 naphthenic mineral oil (Karamay #25 mineral oil, Karamay, China) and a natural ester FR3 (Cargill Bioindustrial, Shanghai, China) were used in this study.Retrofilling procedures in ChinaRetrofilling procedures are under discussion and are being published. The retrofilling procedures that we used in this experiment are presented in the flow chart in Figure A1. The details of the flowchart are discussed in the following section.Transformer check‐up before retrofillingThe transformer nameplate information should be recorded with a special focus on the weight of the insulating liquid on the nameplate.The compatibility of materials with FR3 natural esters and the lowest allowed ambient temperature should be considered. The pour point of FR3 is −21°C, which is higher than that of mineral oil. This issue may lead to the problem of starting up at extremely low ambient temperatures.The requirements of the inspection on the retrofilled transformer are listed in Table A1.Draining oilThe procedures for draining the mineral oil required that the ambient temperature be approximately 20°C and the relative humidity be less than 60% to avoid water entering the oil tank as well as to drain the water that may be condensed from the vapour in the oil conservator. After draining the oil, the oil‐level gauge hole should be closed to prevent the transformer body from being exposed to air.To reduce the mineral oil residue on the active part, 1–2 h of waiting is necessary. Sufficient waiting period can reduce residual mineral oil after retrofilling.FlushingThe transformer should be flushed from the oil inlet. When the transformer has an oil conservator, the highest point of the oil conservator should be used as the oil inlet. When the transformer has no oil conservator, the oil‐level gauge port should be used as the oil inlet. The hole in the oil inlet should not be fully closed because the gas inside the transformer is vented outwards.A new catheter was used to flush the active part and components with the natural ester at the two flanges of the removed bushings. The natural ester used in this procedure is usually approximately 10% of the total oil volume.Draining oil again and replacing the gasketsA large amount of mineral oil can remain in the bottom of the tank during the flushing procedure. The residual oil was drained again using the method described in Section 2.2.2. The gaskets used should meet the requirements for compatibility with natural esters. Silicone rubber, fluororubber, and fluorosilicone rubber seals can be used in transformers with higher thermal classes. Thereafter, the high‐voltage bushing should be reinstalled to ensure a good sealing condition.Retrofilling with natural ester and check‐upIn the retrofilling procedure, the pipe must be extended to the bottom of the transformer to minimise the generation of bubbles.The transformer should be left to stand in the warehouse for no less than 4 h after retrofilling to wait for the air bubbles in the ester liquid to be fully dissipated. After retrofilling, the insulating liquid inside the transformer becomes a mixture of natural ester and a small amount of mineral oil.Finally, the retrofilled transformer should be inspected again according to Table A1 to guarantee a successful retrofilling.Temperature rise tests and the overload testsTemperature rise tests were performed on both retrofilled transformers before and after retrofilling. The special overload tests were only performed on the S13‐M‐400/10 transformer.The temperature rise tests followed the short‐circuit method according to the IEC 60076‐2 standard [31]. The total test continued for 8.5 h. In the first stage, the total loss was applied until the change in the top oil temperature was less than 1 K/h for 3 h. In the second stage, the rated current was applied for 1 h. When the power was disconnected, the temperatures of the top oil and windings were measured.The top and bottom oil temperature was measured at the outlet and inlet of the pipeline in the temperature rise tests. One PT 100 type thermocouple was placed firmly on the surface of the inlet and outlet pipe. The average winding temperature was calculated by the winding resistance.The overload tests followed the technical guide of SGCC, and the test curve is shown in Figure 1. Before the overload tests, the temperature rise of the tested transformer was expected to reach a steady temperature rise under the rated load. First, the transformer was in operation for 3 h under 1.5 times of rated load. Whether the temperature rise was steady, the transformer test proceeded to the next 1.75 times of rated load, as shown in Figure 1. The total overload tests had five steps and then the load is cut off after finishing the total 10 h overload test.1FIGUREDistribution transformer overload test curve of the State Grid Corporation of China (SGCC)The temperature rise of the average windings, average oil, and top oil were measured at the end of the 1.5 times rated load. The temperature rise of the top oil was continuously measured during the test.Oil tests and prediction of mineral oil mass percentageGenerally, retrofilling procedures cannot remove all residual mineral oil, as absorbed oil exists in the solid insulation. Thus, the insulating liquid in the transformer is a mixture of natural esters and mineral oils. To meet the K‐class less‐flammable requirements, the fire point of the mixed oil should be higher than 300°C [7, 32]. Therefore, it is crucial to control the mass ratio of the mineral oil in the mixed oil. The properties of the insulating oil in the two transformers were measured before retrofilling, after retrofilling, and after 1 year.To confirm the exact mass percentage of mineral oil in the mixed oil after retrofilling with natural ester, an experiment with different mass percentages of mineral oil was designed. The mineral oil and natural ester were mixed and allowed to stand idle for 2 days to be mixed well. Specifically, 0%–10% mass percentage of mineral oil in 1% steps was blended, as the residual mineral oil is generally less than 10%.The tested properties included water content (IEC 60814), 90°C dielectric dissipation factor (IEC 60247), 2.5 mm breakdown voltage (IEC 60156), DGA (IEC 60599 for mineral oil, and IEEE C57.155 for natural ester), kinematic viscosity (ISO 2909), open‐cup flash point (ISO 2592), and closed cup flash point (ISO 2719).EXPERIMENT RESULTS AND ANALYSISTemperature rise tests and overload testsTemperature rise testsThe results of the temperature rise tests are shown in Tables 2 and 3, indicating that the two FR3 fluid retrofilled transformers have higher temperature rises than the mineral oil transformers. The top oil temperature rises were higher by 4.1 K (S11‐M‐315/10) and 5.0 K (S13‐M‐400/10). The average oil temperature rises were higher by 3.3 K (S11‐M‐315/10) and 4.0 K (S13‐M‐400/10).2TABLEResults of the S11‐M‐315/10 temperature rise testsTemperature rise forMineral oil transformerRetrofilled transformerLimits (IEC 60076‐14)Top oil (K)44.248.3≤60Average oil (K)35.438.7–High voltage winding (K)51.663.1≤65Low voltage winding (K)58.566.3a≤65aExceed the temperature rise limits for the thermal class of 105°C, but do not exceed the limits for the thermal class of 120°C.3TABLEResults of the S13‐M‐400/10 temperature rise testsTemperature rise forMineral oil transformerRetrofilled transformerLimits (IEC 60076‐14)Top oil (K)41.746.7≤60Average oil (K)33.337.3–High voltage winding (K)58.362.6≤65Low voltage winding (K)58.467.2≤65aExceed the temperature rise limits for the thermal class of 105°C, but do not exceed the limits for the thermal class of 120°C.The thermal class of cellulose paper immersed in mineral oil is 105°C. Thus, the limits in Tables 2 and 3 are the temperature rise limits for the thermal class of 105°C. However, according to IEC 60076‐14 standard Annex C [33], as shown in Table 4, the thermal class of the electrical insulation system can be increased by changing the liquid from mineral oil to natural ester, resulting in temperature rise limits of 90 and 75 K for the top oil and average windings, respectively. Therefore, the temperature rise values of the distribution transformers after retrofilling are up to the temperature rise limits in the IEC 60076‐14 standard Annex C. Considering that the thermal class of natural ester‐immersed cellulose paper is 120°C, although the temperature rise of the low‐voltage windings exceeded the temperature rise limit of mineral oil, the temperature rise of the low‐voltage windings remained within the limits for the ester filled transformers.4TABLEThermal class of mineral oil and natural ester according to IEC 60076‐14Insulating systemsKraft paper with mineral oilKraft paper with natural esterCellulose‐based thermal class(°C)105120Temperature rise limits for top oil (K)6090Temperature rise limits for average winding (K)6575Overload testsThe results of the temperature rise test for the average windings, average oil, and top oil at the end of the 1.5 times rated load (cooling phase) are shown in Table 5. The temperature of the top oil during the overload test is shown in Figure 2. The ambient temperature during the tests before and after retrofilling was 12.3 and 15.1°C, respectively. The ambient temperature was not measured in real time, but only once at the beginning of the test.5TABLETemperature rise at the end of the 1.5 times rated load of S13‐M‐400/10 transformerTemperature rise forMineral oil transformerRetrofilled transformerLimitsTop oil (K)84.091.7≤60Average oil (K)67.273.4–High voltage winding (K)95.8101.4≤65Low voltage winding (K)92.5103.6≤652FIGURETop oil temperature variations for the mineral oil and natural ester transformers under the given over‐load test curveThe retrofilled 400 kVA transformer had increased temperature rises for top oil of 7.7 K, average oil of 6.2 K, high voltage windings of 5.6 K, and low voltage windings of 11.1 K, when compared with the mineral oil transformer. These values are beyond the limits as discussed in Section 3.1.1.As shown in Figure 2, the mineral oil and retrofilled transformers exhibited the same trend as the top oil temperature change in the overload test. The highest temperature rise value of the top oil occurred at the end of the 2.0 times rated load. Furthermore, the highest temperature rise values of the mineral oil and retrofilled transformers were 109.1 and 121.9 K, respectively. Thus, the temperature values are beyond the limits as discussed in Section 3.1.1. Accordingly, a conventionally designed mineral oil transformer cannot fulfil the severe overloading test pattern of the SGCC. Thus, the high overload capacity transformer requires a more specific design, such as thermally upgraded paper or aramid paper for windings or a lower current density, as per the nameplate power rating.Oil tests and mixed oil testsOil testsThe results of the insulating oil tests of S11‐M‐315/10 and S13‐M‐400/10 are shown in Table 6. As previously explained, before retrofilling, the two transformers were filled with mineral oil, and mixed oil after retrofilling. The mixed oil was composed of a majority of natural ester and residual mineral oil. After 1 year, the retrofilled S11‐M‐315/10 operated for another 1 year, while the retrofilled S13‐M‐400/10 unit stood idle for 1 year. At this time, the residual mineral oil and natural ester liquids were expected to be mixed well, and the insulation materials were impregnated with natural esters.6TABLEOil tests results of the S11‐M‐315/10 and S13‐M‐400/10 transformersS11‐M‐315/10S13‐M‐400/10PropertiesBefore retrofillingAfter retrofillingAfter 1 yearBefore retrofillingAfter retrofillingAfter 1 yearFire point (°C)/352335/353338Dielectric dissipation factor @ 90°C0.00560.01270.08850.00220.03530.0793Breakdown voltage (2.5 mm) (kV)43.4461.3378.0063.5768.0665.31Kinematic viscosity @ 40°C (mm2/s)//32.97//32.67Open cup flash point (°C)//272//278Closed cup flash point (°C)//220//209Before retrofilling, the dielectric dissipation factor at 90°C and the breakdown voltage of the mineral oil were acceptable according to the IEC 60422 standard [34]. After retrofilling, the properties of the natural ester were well within the requirements of the IEC 62770 standard [35] for unused natural esters. Although dissipation factors at 90°C increased, the values were within the “good” limit as per IEC 62975. Robert discovered a similar phenomenon through continuous overload for 1 year [36]. This can be explained by the gradual homogenisation of the mineral oil in the cellulose paper with the natural ester. All characteristics meet the requirements in the IEC standard, which means successful retrofillings.In addition to the basic properties of mineral oil, dissolved gas was also included in the study. It is necessary to mention that the DGA results of the two transformers before retrofilling were acceptable as per IEC 60599 standard [37]. The DGA results of the S13‐M‐400/10 transformer are shown in Figures A2 and A3.It can be observed that except for H2 and C2H2, all gases increased after the overload test for the mineral oil transformer. However, all gas contents were still within the limits of IEC 60599. For the transformer after retrofilling, all gases except C2H2 increased after the overload test. The CO, H2, CH4, C2H4, and total hydrocarbon contents exceeded the threshold values for soybean‐based natural esters in IEEE C57.155. From the DGA perspective only, the overload test may not be suitable for the retrofilled transformer of traditional mineral oil design.After 1 year under sealed conditions, the C2H6 and total hydrocarbon contents of the retrofilled transformer (natural ester transformer) were even higher. Therefore, C2H6 and total hydrocarbons will increase naturally even when the transformer filled with the natural ester is offline. Thus, C2H6 and total hydrocarbons cannot be used as the key gases used for qualitative determination of fault types in transformers filled with natural esters.A comparison of the DGA of the S13‐M‐400/10 transformer shows that the overload test after retrofilling causes the gas content increased continuously. This increase could not be resolved by de‐energising the transformer. Therefore, it is prudent to consider whether to conduct an overload test after retrofilling. If necessary, the rated load multiple of the maximum load for the overload test must be considered.Prediction of mineral oil mass percentageUnused mineral oil and unused natural ester were used for samples to obtain the data in these experiments. Based on practical engineering considerations, the performance of the mixed insulating oil with mineral oil content in a certain linear range was tested in these experiments. The dielectric dissipation factor has a large discrete range [1], and the breakdown voltage is highly influenced by the moisture [38,39]. Thus, the kinematic viscosity, fire point, open cup flash point, and closed cup flash point were tested.The kinematic viscosity of the oil blend was measured every 10°C over a temperature range of 40–100°C, as shown in Figure 3. For the viscosity of the mass ratio of mineral oil from 0% to 6%, a clear linear change was observed. When it reaches 7%, there is a sudden drop in viscosity, followed by an unexpected increase at a ratio of 9%. These unusual phenomena can be seen in McShane's research [1] and leads to poor linearity of viscosity. Hence, it is better to use 0%–6% mineral oil ratios to describe the variation in viscosity with mineral oil ratios. This allows the prediction of only mineral oil ratios below 6%. In addition, the viscosities of the different mineral oil ratios were similar at the same temperature. Therefore, the viscosity at 60°C has the ideal fitting equation in the following form, with the highest coefficient of determination (R2) of 0.989:1y1=−0.234x+18.6\begin{equation}{y_1} = - 0.234x + 18.6\end{equation}3FIGUREKinematic viscosity of low mass ratio of mineral oil mixed with natural esterHere, x$x$ represents the mass ratio of mineral oil and y1${y_1}$ represents the kinematic viscosity at 60°C values.The results of the fire points are shown in Figure 4. Owing to the rapid decrease in the fire point of the mixed oil in relation to the percentage of mineral oil at the 5%–10% stage, the fitted curve should be divided into two parts. The first part is a slowly decreasing part with a fire point above 330°C. The second part is the rapidly decreasing part with a fire point below 330°C. The second part is the main discussion point, as the fire point will be lower than 300°C, and does not meet the K‐class less‐flammable requirements. Figure 4 compares the different fitted curves in the second part. Although the fitted curve using 6%–10% ratios of the mineral oil has the highest R2, the fire point of this fitted curve at a 5% ratio is 315°C. The fire point at the 5% ratio of the first part is 340°C, which is significantly higher than the fitted curve using 6%–10% ratios. The fire points at 5% ratio of the fitted curves using 5%–10% ratios and 5%–9% are 325°C and 328°C, respectively. Two fitted curves have similar fire points at a 5% ratio, but the latter has a higher R2. Thus, the ideal fitting equation for the fire point (y2) below 340°C is of the following form:2y2=−21.8x+437.6\begin{equation}{y_2} = - 21.8x + 437.6\end{equation}4FIGUREFire point of low mass ratio of mineral oil mixed with natural esterThe fitting equation for the fire point (y2) above 340°C is as follows:3y2=−2.57x+353.1\begin{equation}{y_2} = \ - 2.57x + 353.1\end{equation}where x$x$ represents the mass ratio of mineral oil and y2${y_2}$ represents the fire point values.The flash point has good linearity at a 0%–10% mass ratio of mineral oil in the mixed oil, as shown in Figure 5. The closed cup (y3) and open cup (y4) flash points are of the following form:4y3=−9.34x+255\begin{equation}{y_3} = - 9.34x + 255\end{equation}5y4=−13.8x+329\begin{equation}{y_4} = - 13.8x + 329\end{equation}5FIGUREFlash point of low mass ratio of mineral oil mixed with natural esterHere, x$x$ represents the mass ratio of mineral oil, and y3${y_3}$ and y4${y_4}$ represent the closed cup and open cup values.The kinematic viscosity at 60°C has an extremely high linearity with an R2 of 0.989. However, the dispersion of the fire and flash points is considerable. Thus, only kinematic viscosity and its fitted curve at 60°C were used to predict the residual mineral oil content in this study. Note that the change in the natural ester and mineral oil category may impact the parameters of the fitted curve and should refit the curve if necessary. In addition, the distribution of the fire point with the mass ratio of mineral oil is not exactly the same as the widely used curves according to McShane's research [1]. This is caused by the use of different mineral oils.DISCUSSIONOverload testFor the 400 kVA transformer, the hot spot temperature can be calculated according to Figure 2 and IEC 60076‐2 as the following formula:6Δθh=Δθo+Hg\begin{equation}\Delta {\theta _h} = \ \Delta {\theta _o} + Hg\end{equation}where Δθh$\Delta {\theta _h}$ represents the hot‐spot temperature rise, Δθo$\Delta {\theta _o}$ represents the top oil temperature rise, H$H$ represents the hot spot factor, and g$g$ represents the average winding gradient.The hot spot factor H is assumed to be 1.1. For the ONAN transformer up to 2500 kVA, the average liquid temperature rise above the ambient temperature may be considered as 80% of the top liquid temperature rise. Thus, the average winding gradient (g) was calculated using the following formula:7g=Δθw−0.8Δθo\begin{equation}g = \Delta {\theta _w} - 0.8\Delta {\theta _o}\end{equation}where Δθw$\Delta {\theta _w}$ represents the average winding temperature rise and Δθo$\Delta {\theta _o}$ represents the top oil temperature rise.The calculated hot‐spot temperatures are shown in Figure 6. The highest hot spot temperatures of mineral oil and natural ester are 149.0°C and 167.8°C, respectively. The closed‐cup flash point of the natural ester after retrofilling is 209°C, which is significantly higher than the hot spot temperature of 167.8°C. However, the closed cup flash point of mineral oil is 140°C, and the hot spot temperature is 9.0°C higher than the closed cup flash point, which can cause danger during the overload test. This indicates that retrofilling with natural ester can significantly improve the fire resistance of the transformer. In addition, a specially designed mineral oil transformer is required for an overload test. An overly rigorous overload test required by SGCC was performed.6FIGUREHot spot temperature variations for the mineral oil and natural ester transformers under the given over‐load test curveRetrofilling with a natural ester liquid can improve the overload capacity of a conventionally designed mineral oil transformer. The durations in which the temperatures of the top oil and the hot spot reached the required limits in IEC 60076–7 and IEC 60076‐14, respectively, with a rated load of 1.5 were calculated. The duration was chosen as the closest value, and the results are listed in Table 7.7TABLEOverload capacity of the 1.5 times rated load before and after retrofillingTop oil temperature (°C)Hot spot temperature (°C)400 kVA transformer filled withAmbient temperature (°C)Measured valueLimitsCalculated valueLimitsaTime to reach limit (min)Mineral oil12.391.6105119.2120100Natural ester15.197.9105130.8130130aLimits of mineral oil required in IEC 60076‐7 and limits of retrofilled transformer required in IEC 60076‐14.The hot spot temperatures of both the mineral oil and retrofilled transformers reached their limits prior to those of the top oil temperatures. After retrofilling with natural ester, it took 130 min to reach the temperature limit under 1.5 times rated load, which is 30% higher than when the same transformer is filled with mineral oil. Therefore, the overload capacity of the 1.5 times load was improved by retrofilling. In addition, the overloading test pattern of SGCC is significantly difficult for conventionally designed transformers, thus requiring a special design.Prediction of mineral oil mass percentageBecause the kinematic viscosity at 60°C has a good linearity with respect to the mass ratio of mineral oil, it is possible to predict the mass ratio of mineral oil in the retrofilled transformers after retrofilling with natural ester.The prediction results of the residual mineral oil content using the kinematic viscosity are shown in Figure 7. The prediction uses prediction bands at the 95% confidence level. The red bands are the prediction bands, thus the provided kinematic viscosity at 60°C has a 95% confidence in the prediction bands.7FIGUREPrediction bands at 95% confidence level of kinematic viscosity at 60°C and the ratio of residual mineral oil predictionFor the S11‐M‐315/10 transformer, the kinematic viscosity at 60°C is 18.31 cSt. The prediction result of the mass ratio of residual mineral oil using kinematic viscosity at 60°C is 0.53% to 1.96% at a 95% confidence level, with an average level of 1.24%. For the S13‐M‐400/10 transformer, the kinematic viscosity at 60°C is 18.12 cSt. The prediction result of the mass ratio of residual mineral oil using kinematic viscosity at 60°C is 1.39%–2.75% at a 95% confidence level, with an average level of 2.05%.The fire points of the S11‐M‐315/10 and S13‐M‐400/10 transformers were 335°C and 338°C, respectively. Thus, according to Equation (2), the mass ratios of the two transformers were 4.7% and 4.5%, respectively. A similar result can be calculated by the flash point according to Equations (4) and (5) for the two transformers by 3.9% and 4.3%, respectively.Owing to the low linearity and piecewise equation of fire point and flash point, it is not recommended to predict the mass ratio of residual mineral oil using fire points and flash points, but to judge whether the insulating oil meets the K‐class less‐flammable requirements or not. The insulating oil of two retrofilled transformers remain a K‐class less‐flammable liquid.Predictions using the prediction bands at a 95% confidence level for kinematic viscosity provided the most favourable results. The prediction results of the residual mineral oil mass ratio after retrofilling were 1.24% for the S11‐M‐315/10 transformer and 2.05% for the S13‐M‐400/10 transformer. This mass ratio of residual mineral oil is significantly lower than the 7% expected residual mineral oil from [1], which should be attributed to the detailed retrofilling guidance presented in this study.CONCLUSIONDetailed procedures for retrofilling 10 kV distribution transformers with natural esters were described in this study. A method to predict the residual mineral oil content after retrofilling mineral oil transformers with natural esters was proposed using kinematic viscosity at 60°C. This method addresses the issue of testing residual mineral oil content. The two retrofilled transformers had mass ratios of residual mineral oil of 1.24% for the S11‐M‐315/10 transformer and 2.05% for the S13‐M‐400/10 transformer. The mass ratio of mineral oil was believed to be approximately 7% from previous studies.For the 315 and 400 kVA retrofilled transformers, the temperature rises at rated load of top oil increased by 4.1 and 5.0 K, average oil by 3.3 and 4.0 K, and low voltage windings by 7.8 and 8.8 K, respectively, compared with mineral oil transformers. According to the IEC 60076‐14 standard, the thermal class of the oil‐paper insulation system can be improved by changing the oil from mineral oil to natural ester, thus resulting in the temperature rise limits of the top oil and the average windings for the natural ester‐Kraft paper system to reach 90 and 75 K, respectively. Thus, the obtained temperatures using the traditional temperature rise test can meet the limits of retrofilling.Retrofilling with natural ester can significantly improve the fire resistance of a mineral oil transformer. The results of the special overload test of the 400 kVA transformer show that the highest hot spot temperature of mineral oil was 149.0°C, which is 9.0°C higher than that of the closed cup flash point, causing danger during the overload test. The closed‐cup flash point of the natural ester after retrofilling is 209°C, which is significantly higher than the hot spot temperature. This indicates that a specially designed mineral oil transformer cannot fulfil the severe overloading pattern of SGCC, even with retrofilling. Thus, transformers with a high overload capacity need more specific designs.After retrofilling with natural ester, it took 130 min to reach the temperature limits under 1.5 times rated load, and the duration was 30% longer than when the same transformer was filled with mineral oil. The overload capacity of the 1.5 times load can therefore be improved by retrofilling.ACKNOWLEDGEMENTSThe authors would like to express their gratitude to Dong Han of Xian Xibian Zhong Te Electric CO, Ltd. and Bin Wei of State Grid Shaanxi Electric Power Company for their technical support in this study.CONFLICT OF INTERESTThe authors declare no conflict of interest.DATA AVAILABILITY STATEMENTThe data that support the findings of this study are available from the corresponding author upon reasonable request.REFERENCESMcShane, C.P., Luksich, J. & Rapp, K.J.: Retrofilling aging transformers with natural ester based dielectric coolant for safety and life extension. In: Cement Industry Technical Conference, TX, US, 4–9 May 2003Pillai R., Havaldar F., Chitnis C.: Natural esters for life and capacity enhancement of distribution transformers. CIRED ‐ Open Access Proceedings Journal. 2017(1), 355–358 (2017)Zheng, Z.Y., Zhang, J.Q., et al.: Performance change analysis of replacing natural ester insulating oil with mineral insulating oil distribution transformer. Transformer 56(6), 27–30 (2019)Liao, R.J., et al.: A comparative study of thermal aging of transformer insulation paper impregnated in natural ester and in mineral oil. Eur. Trans. Electr. Power 20(4), 518–533 (2010)Hosier, I.L., et al.: Aging of biodegradable oils and assessment of their suitability for high voltage applications. IEEE Trans. Dielectr. Electr. Insul. 18(3), 728–738 (2011)Yu, H., et al.: Dielectric and physicochemical properties of mineral and vegetable oils mixtures. In: IEEE International Conference on Dielectric Liquids, Manchester, UK, 25–29 June 2017IEC 61039 : Classification of insulating liquids. IEC (2008)Stockton, D.P., Bland, J.R., et al.: Seed‐oil‐based coolants for transformers. IEEE Ind. Appl. Mag. 15(1), 68–74 (2008)Edmund, L., et al.: Cargill‐sustainable electrical energy using natural ester. In: 2015 Transformer Life Management (TLM) International Conference Shanghai, Shanghai, China, November, 2015Huang, Z.Q., Wang, R.F., et al.: Test comparison study on a natural ester retro‐filled 220kV power transformer. In: IEEE International Conference on High Voltage Engineering and Application, Beijing, China, 6–10 September 2020IEEE C57:147: IEEE guide for acceptance and maintenance of natural ester fluids in transformers. IEEE (2008)Asano, R., Page, S.A.: Reducing environmental impact and improving safety and performance of power transformers with natural ester dielectric insulating fluids. IEEE Trans. Ind. Appl. 50(1), 134–141 (2013)Josken, J., Wareham, D.: Seed based oil as an alternative to mineral oil. In: Rural Electric Power Conference, Scottsdale, AZ, 25–25 May, 2004Fofana, I., Wasserberg, V., Borsi, H., Gockenbach, E.: Preliminary investigations for the retrofilling of perchlorethylene based fluid filled transformers. IEEE Trans. Dielectr. Electr. Insul. 9(1), 97–103 (2002)Fofana, I., Borsi, H., Gockenbach, E.: Oil filled transformer retrofilled with ester liquid—Facts and arguments. In: 15th International Symposium on High‐Voltage Engineering. ISH, Ljubljana, Slovenia, August 2007McShane, C.P., et al.: Aging of paper insulation retrofilled with natural ester dielectric fluid. In: 2003 Annual Report Conference Electrical Insulation Dielectric Phenomena CEIDP, Albuquerque, NM, 19–22 October 2003McShane, C.P., et al.: Natural ester dielectric fluid development. In: 2005/2006 IEEE/PES Transmission Distribution Conference Exhibition, Dallas, TX, 21–24 May 2006McShane, C.P., et al.: Some considerations for new and retrofil applications of natural ester dielectric fluids in medium and large power transformers. In: 2005/2006 IEEE/PES Transmission and Distribution Conference and Exhibition, Dallas, May 2006Moore, S., Kevin, R., Ramona, B.: Transformer insulation dry out as a result of retrofilling with natural ester fluid. In: PES Transmission Distribution Conference Exhibition 2012, Orlando, 21–24 May, 2012Pillai, R., et al.: DGA of natural ester filled transformers: Experience with retrofilled and new transformers. In: IEEE International Conference Dielectric Liquids, Manchester, UK, 25–29 June 2017Murphy, J.R., Graham, J.: Distribution utility experience with natural ester dielectric coolants. In: IEEE Power Energy Society General Meeting, Calgary, Canada, 26–30 July 2009Shinde, R.: Condition monitoring of a retro‐filled power transformer by natural ester Envirotemp FR3 fluid. In: International Conference Cond. Assessment Techniques Electrical Systems, Rupnaggar, Punjab, India, 16–18 November 2017Santisteban, A., et al.: Thermal analysis of natural esters in a low‐voltage disc‐type winding of a power transformer. In: IEEE International Conference Dielectric Liquids, Manchester, UK, 25–29 June 2017Lee, B., Jeong, J., Song, I.: The characteristics of winding temperature for the natural ester filled transformer up to and including 500 kVA. In: Annual Report Conference Electric Insulation Dielectric Phenomena, Quebec, Canada, 26–29 October 2008Smith, S.D., Barry, L.B.: Design and test experience with natural ester fluid for power transformers update. In: 2009 IEEE Power Energy Society General Meeting, AB, Canada, 26–30 July 2009Yang, T., et al.: Feasibility of directly changing natural ester insulating oil for conventional mineral insulating oil‐filled distribution transformer. Insul. Mater. 51(2), 39–43 (2008)Frimpong, G.K., Oommen, T.V., Asano, R.: A survey of aging characteristics of cellulose insulation in natural ester and mineral oil. IEEE Electr. Insul. Mag. 27(5), 36–48 (2011)Wang, R.F., et al.: Study on load performance of distribution transformer before/after filling vegetable insulating oil instead of mineral oil. Transformer 56(4), 78–80 (2019)Zhao, Y.H., Qian, Y.H., et al.: In‐service ageing comparison study of natural ester and mineral oil filled distribution transformers. In: IEEE International Conference Dielectric Liquids, Roma, Italy, 23–27 June 2019Scatiggio, F., Pepe, F., et al.: Increased loadability of transformers immersed in natural ester. In: IEEE International Conference Dielectric Liquids, Roma, Italy, 23–27 June 2019IEC 60076‐2: Power transformers—Part 2: Temperature rise for liquid‐immersed transformers. IEC (2011)Mazzaro, M., et al.: Fire simulation tests of mineral oil and natural esters transformers. In: IEEE International Conference Dielectric Liquids, Roma, Italy, 23–27 June 2019IEC 60076‐14: Power transformers—Part 14: Liquid‐immersed power transformer using high‐temperature insulation materials. IEC (2013)IEC 60422: Mineral insulating oils in electrical equipment—Supervision and maintenance guidance. IEC (2013)IEC 62770: Fluids for electrotechnical applications—Unused natural esters for transformers and similar electrical equipment. IEC (2013)Robert, C.B., et al.: Evaluation of natural ester retrofilled transformers after one year of continuous overload. In: IEEE Electrical Insulation Conference, Calgary, Canada, 16–19 June 2019IEC 60599 : Mineral oil‐filled electrical equipment in service—Guidance on the interpretation of dissolved and free gases analysis. IEC (2015)Hiramatsu, Y., Kamidani, K., Muramoto, Y.: Effect of water on AC breakdown properties of vegetable‐oil‐based insulating fluid mixed with mineral oil. In: Proceeding International Symposium Electrical Insulation Materials, Toyohashi, Japan, 11–15 September 2017Boris, H., et al.: Dielectric behaviour of silicone and ester fluids for use in distribution transformers. IEEE Trans. Dielectr. Electr. Insul. 26(4), 755–762 (1991)AAPPENDIXTable A1 presents the requirements of retrofilled transformer inspection at the beginning and ending of the retrofill. Figure A1 shows the flow chart of retrofilling transformer with FR3 natural ester. Figures A2 and A3 shows dissolved gas content of S13‐M‐400/10 transformer before and after retrofilling.There are four main parts in the retrofilling procedures: (1) Preparation before retrofilling; (2) draining oil procedures. (3) flush procedures. (4) retrofill procedures.In the first procedure ‘preparation before retrofilling’, inspections and test items of transformer in Table A1 should be inspected and tested and the results should meet the requirements of indicated standard. The ambient temperature of approximately 20°C and relative humidity of not higher than 60% are required. The gas should be released for transformers with gas relay. Oil should be drained for transformers with an oil conservator. Valve and drain plug should be opened before and after draining oil for transformers with butterfly valve. These additional preparations are necessary for different transformer constructions to ensure a successful retrofill.In the second procedure ‘draining oil procedures’, oil should be drained from the oil vent valve for transformers with oil vent valve, or from the oil level gauge for transformers without oil vent valve. Note that 1–2 h waiting is valuable. Sufficient waiting can reduce residual mineral oil after retrofilling. After waiting, the residual oil under the tank should be drained through the drain valve at the bottom of oil tank, otherwise a suitable angle is necessary to reduce the residual mineral oil.In the third procedure ‘flush procedures’, if the transformer cannot be flushed and retrofilled at once after the draining oil procedures, the transformer should be filled with dry nitrogen gas to cut off with moisture and oxygen in the air. Flush the windings, core, and oil tank internal surface with natural ester. Then, another hour's waiting can reduce the oil generated by flushing.In the fourth procedures ‘retrofill procedures’, the transformer should be retrofilled with the natural ester to the bottom. Again, the inspections and test items in Table A1 should meet the requirements due to the change of insulating system. Qualified test results mean a successful retrofill.A1FIGUREFlow chart of retrofilling transformer with FR3 natural esterA2FIGUREDissolved gas content of the S13‐M‐400/10 transformer filled with mineral oil (before retrofilling)A3FIGUREDissolved gas content of the S13‐M‐400/10 transformer filled with the natural ester (after retrofilling)A1TABLERequirements of retrofilled transformer inspectionApplicable transformer for retrofillingOnline transformerOffline transformerNew transformerInspection and test items (Standard)Before retrofillingAfter retrofillingBefore retrofillingAfter retrofillingBefore retrofillingAfter retrofillingTransformer discharge (IEC 60076‐3)RequiredVisual inspection and cleaningRequiredRequiredRequiredRequiredRequiredRequiredInsulation resistance measurement (IEC 60076‐1)RequiredRequiredRequiredRequiredRequiredRequiredWinding DC resistance measurement (IEC 60076‐1)RequiredRequiredRequiredRequiredVoltage ratio measurement and connection group label verification (IEC 60076‐1)RequiredRequiredRequiredRequiredRequiredRequiredNo‐load current and No‐load loss (IEC 60076‐1)OptionalOptionalShort‐circuit impedance and load loss (IEC 60076‐1)OptionalOptionalApplied voltage withstand test (IEC 60076‐3)OptionalaRequiredaOptionalaRequiredaOptionalaRequiredbInduced voltage test (IEC 60076‐3)RequiredLightning impulse voltage test (IEC 60076‐3)OptionalInsulating oil test:Breakdown voltage (IEC 60156)Dielectric dissipation factor (IEC 60247)Water content (IEC 60814)RequiredRequiredRequiredRequiredRequiredRequiredLeak test (IEC 60076‐1)RequiredRequiredRequiredDissolved gas analysis (IEC 60599 and IEEE C57.155)OptionalOptionalOptionalOptionalOptionalOptionalaThe applied voltage during the high voltage test is 80% of the standard voltage.bThe applied voltage during the high voltage test is 80%‐100% of the standard voltage.

Journal

"IET Generation, Transmission & Distribution"Wiley

Published: May 1, 2022

There are no references for this article.