Advanced
Experimence Study of Trace Water and Oxygen Impact on SF<sub>6</sub> Decomposition Characteristics Under Partial Discharge
Experimence Study of Trace Water and Oxygen Impact on SF6 Decomposition Characteristics Under Partial Discharge
Journal of Electrical Engineering and Technology. 2015. Jul, 10(4): 1786-1795
Copyright © 2015, The Korean Institute of Electrical Engineers
This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
  • Received : October 05, 2014
  • Accepted : March 31, 2015
  • Published : July 01, 2015
Download
PDF
e-PUB
PubReader
PPT
Export by style
Article
Author
Metrics
Cited by
TagCloud
About the Authors
Fuping Zeng
School of Electrical Engineering, Wuhan University, China. (Fuping.Zeng@whu.edu.cn)
Ju Tang
Corresponding Author: School of Electrical Engineering, Wuhan University, China. (whtangju@whu.edu.cn)
Yanbin Xie
State Grid Chongqing Electric Power Company, Shiqu Power Supply Company, China. (yanbinse@126.com)
Qian Zhou
State Grid Chongqing Electric Power Company, China. (cquzhou@163.com)
Chaohai Zhang
Wuhan NARI Limited Company of State Grid Electric Power Research Institute, China. (zhangchaohai@sgepri.sgcc.com.cn)

Abstract
It is common practice to identify the insulation faults of GIS through monitor the contents of SF 6 decomposed components. Partial discharges (PD) could lead to the decomposition of SF 6 dielectric, so new reactions usually occur in the mixture of the newly decomposed components including traces of H 2 O and O 2 . The new reactions also cause the decomposed components to differ due to the different amounts of H 2 O and O 2 even under the same strength of PD. Thus, the accuracy of assessing the insulation faults is definitely influenced when using the concentration and corresponding change of decomposed components. In the present research, a needle-plate electrode was employed to simulate the PD event of a metal protrusion insulation fault for two main characteristic components SO 2 F 2 and SOF 2 , and to carry out influence analysis of trace H 2 O and O 2 on the characteristic components. The research shows that trace H 2 O has the capability of catching an F atom, which inhibits low-sulfide SF x from recombining into high-sulfide SF 6 . Thus, the amount of SOF 2 strongly correlates to the amount of trace H 2 O, whereas the amount of SO 2 F 2 is weakly related to trace H 2 O. Furthermore, the dilution effect of trace O 2 on SOF 2 obviously exceeds that of SO 2 F 2 .
Keywords
1. Introduction
Different patterns and strengths of partial discharge (PD) always occur when SF 6 electrical equipment have some earlier insulation faults. High local electromagnetic energy caused by PD would cause SF 6 to decompose into several kinds of low-fluoride sulfide SF x [1 - 4] . If trace levels of H 2 O and O 2 exist in the equipment, the decomposed components would have further reactions with them and produce new characteristic components, such as SO 2 F 2 , SOF 2 , SO 2 , and so on [5 - 11] .The concentration and variation regularity of these characteristic components have close relationship with the patterns of insulation faults, as well as the trace levels of H 2 O and O 2 in gaseous SF 6 , making it more difficult to recognize the internal insulation deficiency when using them. Although the new gas SF 6 contains few impurities, trace levels of H 2 O and O 2 would enter the gas chamber as they are released from internal material or by penetration from the outside air over time [12] . There would be extra-trace levels of H 2 O and O 2 inevitably existing in the SF 6 electrical equipment. Hence, when extra-trace levels of H 2 O and O 2 exist in SF 6 gas, learning about the decomposition mechanism from both theory and experiment under PD are necessary; obtaining the influence regularity and influence mechanism of trace H 2 O and O 2 on decomposed components is urgent. Furthermore, it is imperative to offer a amendment method considering the impacts of trace H 2 O and O 2 so that all of the aforementioned methods would lay a solid theoretical foundation for the correct identification and evaluation of the internal insulation faults of SF 6 electrical equipment when making use of the decomposed components.
R. J. Van Brunt from the U.S. National Bureau of Standards conducted a systematic research about the SF 6 decomposition mechanism under PD. He studied the main source [13] of the O atom in SO 2 F 2 , SOF 2 , and SOF 4 using the isotopic tracer technique under the condition of needle-plate electrode of corona discharge. His study pointed out that the O atom of SO 2 F 2 mainly comes from O 2 , the O atom of SOF 2 mainly comes from H 2 O, and the O atom of SOF 4 comes from both O 2 and H 2 O. However, the paper also claimed that SO 2 F 2 obtains the O atom from H 2 O and SOF 2 obtains the O atom from O 2 . Nevertheless, the Van Brunt research used the same fixed concentration of H 2 O and O 2 without considering their levels of variation. According to Arrhenius’ law of chemical reaction kinetics and mass action law [14] , the chemical reaction rate depends on the reaction temperature, reactant concentration, and catalyst. Although Derdouri studied the impact of diverse concentrations of H 2 O on SF 6 gas under PD, there is a lack of explanation of the process [15] .
In this paper, the authors take advantage of the PD decomposition platform in the laboratory and study the concentration of decomposed SF 6 components and their variation trends under PD when different trace levels of H 2 O and O 2 are mixed with SF 6 . Moreover, the mechanism of how the various concentrations of trace water and oxygen act on the characteristic decomposed components from the angle of related chemical reaction rate is explained. Considering that random factors may lead to unfavorable results during the experiment, statistical inference using ANOVA is used to investigate the degree of impact of trace H 2 O and O 2 on decomposed characteristic components of SF 6 .
2. Decomposition Experiment and Quantitative Measurement
- 2.1 Experiment
This work studies the degree of influence of H 2 O and O 2 on decomposed characteristic components of SF 6 under PD from the statistical perspective. Hence, repeating the experiment independently n (here n =4) times under the same trace levels of H 2 O and O 2 (the level is A i ) and making sure that each experiment group has only one variable. The procedure suggests that the concentration of O 2 is controlled below 100ppm (the rate of oxygen analyzer is 100ppm) in the experiment gas sample when the experiment on the influence of different trace levels of H 2 O was conducted. Likewise, the concentration of H 2 O is controlled below 150 ppm when the experiment on the influence of different trace levels of O 2 was conducted. Experimental factors A (H 2 O) and B (O 2 ) were subjected into seven experimental levels, as shown in Table 1 .
Factors affecting the by-product yields
PPT Slide
Lager Image
Factors affecting the by-product yields
Experiment material: SF 6 (purity: 99.99%, H 2 O ≤ 100 ppm, O 2 ≤100ppm), H 2 O, and O 2 were used as experiment materials. The experiment was conducted in the multi-function electrical decomposition of SF 6 equipment designed by our group, which is shown in Fig. 1 [16] . The main body of gas chamber is cylinder and both ends are oval structure to guarantee its air tightness. The volume of the chamber is approximately 10L and the maximum tolerance of air pressure can reach 0.5Mpa. The material of gas chamber is made of stainless steel for its corrosion resistance since the corrosive decomposed compositions of SF 6 may be produced during the experiment. Lead the HV conductor in the gas chamber through HV bushing and the model of insulation faults is positioned in the middle of the gas chamber so that it can connect with bottle of the HV conductor. Gas inlet and gas outlet is equipped to fill in SF 6 required in the experiment and gather the mixed gases sample after PD experiment.
PPT Slide
Lager Image
SF6 decomposition gas chamber
The gas chamber was filled with 0.2 MPa of SF 6 . The experimental system diagram is shown in Fig. 2 . The needle-plate electrode is needed to simulate the common insulation fault (metal protrusion insulation fault) in the equipment. Moreover, the experiment made use of non-inductive detected impedance to send the pulse current signal to the WavePro 7100XL oscilloscope (Analog band: 1 GHz; sampling rate: 20 GHz; memory depth: 48 MB), which can monitor whether the PD is stable.
PPT Slide
Lager Image
Experimental system diagram
- 2.2 Experiment methods
This experiment uses needle-plate electrode model: spacing d is 10mm, curvature radius of needle tip is 0.3mm, diameter of ground electrode is 120mm and its thickness is 10mm. All the experiments are conducted at the same condition: the laboratory temperature is controlled at 15℃ and relative humility at 50%, to avoid the impacts of different temperatures and humidity and ensure the experimental results are comparable. The specific experimental requisition and steps are listed as follows:
  • (1) Measurement of the initial voltage Usof the intrinsic PD of the equipment (without putting insulation faults model) and the initial voltage U0of the PD of the equipment (after putting the needle-plate electrodes). The respective measurements are Us=45 kV and U0= 15 kV.
  • (2) The gas chamber is vacuumized and then filled with new gas, SF6, and vacuumized again. This process is repeated two or three times for purification.
  • (3) For the experimental procedure on the influence of H2O on the decomposed characteristic components of SF6, step (4) is used. Otherwise, for the experiment on the influence of O2on the decomposed characteristic components of SF6, step (5) is used.
  • (4) The gas chamber is filled with the required amount of H2O by gas-syringe when the chamber is in vacuum condition, and subsequently heated in the equipment for 15 minutes. Another 15 minutes is spent to permit the H2O to undergo gasification and uniform distribution in the gas chamber. Step (6) follows.
  • (5) The gas chamber is filled with the required amount of O2when the chamber is in vacuum condition. Another 15 minutes is spent to permit the full volume of O2to be uniformly distributed in the chamber.
  • (6) The gas chamber is filled with SF6equivalent to a pressure of 0.25 MPa and put aside for 24 hours, so that H2O (or O2) and SF6are fully mixed.
  • (7) The concentration of H2O and O2in the mixed gas is measured. If the concentration fails to meet the experimental standards, the procedure goes back to step (2). When the measured concentrations have satisfied the standards, the gas sample is collected and its intrinsic components are analyzed. The respective concentrations of the constituent gases are also measured. Afterward, the mixed gas pressure is adjusted to 0.2 MPa.
  • (8) The electrical wiring is connected as shown inFig. 2. The experimental voltage is then gradually raised to 1.5U0(22.5 kV) and the PD decomposition experiment is conducted for 10 hours under this voltage. This part of the procedure ensures that the contents of the characteristic components are stable. The oscilloscope is used to monitor the electrical discharge of the needle-plate electrode.
  • (9) After 10 hours, the concentration of the different decomposed components in the collected gas sample is analyzed using gas chromatograph (GC).
  • (10) After measuring all the experimental parameters, the gas chamber is vacuumized and put aside for 1 hour to enable absorption of the decomposed components by the surface of the electrode. The time allowance is also aimed to fully extricate the decomposed components attached on the chamber wall and inhibit the impact of the remaining components of the on-going experiment to the next experiment. The procedure is then repeated from step (2) for the next experiment.
- 2.3 Quantitative measurement of decomposed components
In the aforementioned experiments, the gas chromatograph (Varian CP-3800) was used to quantitatively measure the sample gas components produced by the discharge. The GC used the packed column Porapak QS and special capillary column CP-Sil 5 CB in parallel to separate the components in the mixture. Moreover, the chromatograph used PDHID double detectors (detection precision can reach up to 0.01ppm) to quantitatively detect each separated component. The chromatographic column was operated in the He (purity: 99.999%) carrier gas and the working conditions were flow rate, 2 mL/min; constant column temperature, 40 °C; sample size, 1 mL; and split ratio, 10:1. Under these conditions, the packed column could separate air, CF 4 , and CO 2 effectively, and the special capillary column could separate air, SF 6 , SO 2 F 2 , SOF 2 , H 2 S, and SO 2 effectively. Fig. 3 shows the standard chromatograph.
PPT Slide
Lager Image
Standard chromatogram
This study used the external standard method combined with the standard chromatogram to qualitatively and quantitatively detect the decomposed components of SF 6 . Since the SO 2 F 2 and SOF 2 are the most important characteristic decomposed components of SF 6 [1 , 3 , 6 - 8 , 13 , 17 , 18] , the present study conducted intensive research on both. The raw data of each experiment as show in Table 2 , Figs. 4 and 5 are the results of production amounts of the SO 2 F 2 and SOF 2 yields under PD at different levels of H 2 O and O 2 in 10 hours (Each result is the production average value of four times repeated experiments under the same level of trace H 2 O or O 2 ).
The raw data of each experiment (ppm)
PPT Slide
Lager Image
The raw data of each experiment (ppm)
PPT Slide
Lager Image
Yield of by-products influenced by H2O
PPT Slide
Lager Image
Yield of by-products influenced by O2
3. Influence of H2O and O2on Characteristic Components
Fig. 4 and 5 show that different levels of H 2 O and O 2 contribute to different concentrations of SO 2 F 2 and SOF 2 produced by SF 6 even under the same strength and time length of PD. Besides, the inevitable random factors which exert impacts on the experiment results should be considered in the experiment. Hence, the authors used ANOVA to study the impact of the various levels of trace H 2 O and O 2 on characteristic components SO 2 F 2 and SOF 2 and indentify the main influence factors on the production of SO 2 F 2 and SOF 2 .
- 3.1 Analysis of variance
ANOVA was introduced by the American statistician Fisher in an agricultural experiment [19] . Subsequently, the method has been widely used in other areas, especially in data analysis of industrial experiments where the method ANOVA shows that the total variance
PPT Slide
Lager Image
in the sample data can be divided into two parts: variance between groups
PPT Slide
Lager Image
and variance within groups
PPT Slide
Lager Image
.
PPT Slide
Lager Image
is caused by controllable influential factors of different levels and
PPT Slide
Lager Image
is caused by all random errors, that is
PPT Slide
Lager Image
. The size of the difference between groups and the size of the difference within groups are compared to identify the degree of impact of each level on the experimental results, where
PPT Slide
Lager Image
,
PPT Slide
Lager Image
, and
PPT Slide
Lager Image
can be achieved from Equ. (1) to (3):
PPT Slide
Lager Image
PPT Slide
Lager Image
PPT Slide
Lager Image
In the above equations, x ij is the j-th independent experimental result under the i-th concentration level of trace H 2 O and O 2 ( A i ), which means it is the result of the concentration of SOF 2 or SOF 2 when conducting the j-th experiment independently under the i-th concentration level of trace H 2 O and O 2 ( A i );
PPT Slide
Lager Image
represents the group mean of the product of SO 2 F 2 or SOF 2 under the i-th concentration level of trace H 2 O and O 2 ( A i ) when conducting experiment n times independently;
PPT Slide
Lager Image
represents the mean of all the products of SO 2 F 2 or SOF 2 under the same influence factor (trace H 2 O or O 2 ) in r experimental levels; and r is the number of influence factor (trace H 2 O and O 2 ) concentration level. There are seven concentration levels in the present study, hence, r =7. n is the number of times the experiment is repeated under the same condition, and in the current study, the experiment is repeated 4 times under the same concentration of trace H 2 O and O 2 , thus, n =4.
If all the experimental factors (trace H 2 O and O 2 ) have no significant influence in the experimental results,
PPT Slide
Lager Image
is almost equal to
PPT Slide
Lager Image
and statistics can prove that
PPT Slide
Lager Image
In the equation, ( r −1), r ( n −1) are the degrees of freedom of
PPT Slide
Lager Image
and
PPT Slide
Lager Image
, respectively. Furthermore, let
PPT Slide
Lager Image
PPT Slide
Lager Image
and call them mean variance, so that equation (4) can be simplified as
PPT Slide
Lager Image
ANOVA merely makes use of equation (5) to identify the degree of impact of each level (trace H 2 O or O 2 ) on the experimental results by comparing the differences between groups and the differences within groups. Given the significance level α , when the calculated value of F is above the critical value F 1−α ( r − 1, rn − r ), the influence factor (trace H 2 O or O 2 ) has significant influence on the experimental index (generation amount of SO 2 F 2 and/or SOF 2 ). Furthermore, the bigger the F value of the sample, the more significant influence the factor has on the experimental index. Hence, specific attention has to be accorded on such influence factor, and additionally such influence factors should be controlled during practical production.
- 3.2 Analysis of the significance level of influence factors
Significance level α is a critical probable value that represents the possibility to commit the fallacy of refusing the ‘assumption’ in a ‘statistical hypothesis test’ when using the sample information to draw conclusion. The smaller the value of α , the lesser the possibility of making the mistake of refusing the ‘assumption’. When analyzing data in the field of general industry, α = 0.05 ; in the field of biology and medicine, α = 0.01 . In the present study which examines the degree of influence of trace H 2 O and O 2 on the main characteristic components of SF 6 (SO 2 F 2 and SOF 2 ) under PD, high precision is required. Hence, a significance level of α = 0.01 is adopted. The table of F -distribution critical values shows that F 0.99 (6, 21) = 3.81. The result is presented in Table 3 .
Analysis of variance result
PPT Slide
Lager Image
Analysis of variance result
By collating and analyzing the results from Fig. 4 , Fig. 5 , and Table 3 , the findings indicate that both H 2 O and O 2 exert an influence on the main characteristic components SO 2 F 2 and SOF 2 . However, the products and the degree of influence of H 2 O and O 2 are different. The differences include the fact that H 2 O has an obvious influence on SOF 2 as the F value reaches 87.94, which is much larger than its influence on SO 2 F 2 . The formation of SOF 2 is linearly proportional to the concentration of H 2 O, but the formation of SO 2 F 2 has almost no relationship with the concentration of H 2 O. O 2 has a significant effect on both SO 2 F 2 and SOF 2 , but the impact on SOF 2 is significantly higher than the impact on SO 2 F 2 .
The existence of H 2 O has an effect on the decomposition, as shown by Van Brunt. However, the effect caused by O 2 I is different, which may be due to the fact that Van Brunt did his experiment under the same concentration of O 2 and H 2 O without taking into account the concentration of the reactants on the relevant reaction when exploring the sources of O in SOF 2 and SO 2 F 2 .
4. The influence Mechanism of H2O and O2on the Characteristic Decomposed Components
Under PD, a series of characteristic components are produced by the reaction between the low-fluoride sulfide caused by the decomposition of SF 6 and the trace levels of H 2 O and O 2 mixed in the gas. Van Brunt carried out a more detailed study of the SF 6 decomposition mechanism under PD with a needle-plate electrode mode. He proposed using the Plasma Chemical Model to explain the SF 6 decomposition mechanism under PD [8] . He pointed out that under the effect of the high-energy electrons generated by the PD, the following reaction will occur in SF 6 :
PPT Slide
Lager Image
The high-energy electrons lead to the decomposition of SF 6 to produce low-fluoride sulfide SF x (x = 1 to 5). When no other impurities exist in SF 6 , SF x will recover quickly with the following reaction:
PPT Slide
Lager Image
Here, k is the rate constants of the reaction. However, during the long-term operation of the SF 6 gas-insulated equipment, it is inevitable that different amounts of ultra-impurity gases, such as H 2 O and O 2 , will appear in the chamber, released by the internal material of the device and the penetration of external H 2 O and O 2 into the equipment. The impurities will lead to a series of more complex chemical reactions with SF x and generate SO 2 F 2 , SOF 2 , HF, SO 2 , and other compounds. Therefore, trace amounts of H 2 O and O 2 play a key role in the production of SO 2 F 2 and SOF 2 .
- 4.1 Analysis of the characteristic decomposed component with the impact of H2O
H 2 O will undergo the following reaction under PD when H 2 O exists:
PPT Slide
Lager Image
PPT Slide
Lager Image
Meanwhile, the following reactions will occur among H 2 O and SF 6 decompositions:
PPT Slide
Lager Image
PPT Slide
Lager Image
The reaction rate constants k of the reaction (10), (11) and reaction (7), are in the same order of magnitude. On the other hand, the mass action law [14] tells us that the chemical reaction rate r depends on the reactant concentration, C i , the stoichiometry number, b i , rate and the constant k, and the relationship is as follows:
PPT Slide
Lager Image
Fortunately, under PD, for all of the chemical reactions where SF 6 and H 2 O are involved, the stoichiometry number b i is one. This finding suggests that the reaction rate r is proportional to the concentration of the reactants. Therefore, when traces of H 2 O exist, H 2 O has a capture function of F equivalent to the inhibition of the recovery reaction SF x + (6−x) F → SF 6 . H 2 O inhibits the low-fluoride sulfide SF x (x = 1, 2, 3, 4, 5) composite to SF 6 , so that the concentration of SF 4 , SF 5 , and other components are increased. Additionally, under PD, the trace amount of H 2 O has always been small compared with a variety of low-fluoride sulfide SF x . Thus, the rates of reaction above are mainly determined by the concentration of H 2 O. The higher the concentration of H 2 O, the more severe the reaction and the more obvious the inhibition, as explained by the following reactions:
PPT Slide
Lager Image
PPT Slide
Lager Image
PPT Slide
Lager Image
Formulas (8) to (11) and (13) to (14) show that when SF 6 is mixed with H 2 O, H 2 O plays a role in providing OH and O. Hence, the formation of SOF 4 is promoted. Meanwhile, reactions (6) to (11) and (13) to (15) constitute a comprehensive reaction, which is the means by which SO 2 F 2 is generated. The generation capacity for SO 2 F 2 is determined by trot reaction (15), and with the increasing concentration of H 2 O, the amount of SO 2 F 2 will slightly increase. However, with nearly 10 orders of magnitude of reaction rate in (15) than the rates of reaction in (10), (11), (13), and (14), and with increased concentration of H 2 O, the increase of SO 2 F 2 is not obvious, as shown in Fig. 4(a) .
On the other hand, the reaction between SF 4 and H 2 O will occur as follows [8] :
PPT Slide
Lager Image
The reactions in (16) and (12) show that when the concentration of H 2 O increases in SF 6 , it will promote the production of SOF 2 , as shown in Fig. 4 (b) . However, the reaction rate constant k of reaction (16) is 2 orders of magnitude higher than reaction (15). Therefore, with the increased concentration of H 2 O in SF 6 , the rate of increase of the SOF 2 produced is significantly higher than that of SO 2 F 2 , as shown in Fig. 4(b) .
The effect of H 2 O on SF 6 decomposition characteristics under PD can be summarized as follows:
  • (1) H2O has a capture function of F which inhibits the low-fluoride sulfide SFx(x = 1, 2, 3, 4, 5) composite to recombine with SF6, leading to the increase in the main low-fluoride sulfide SF5, SF4, and other components. The higher the concentration of H2O, the more severe the reaction and the more obvious the inhibition.
  • (2) H2O provides OH and O for the generation of oxygen-containing-sulfur-fluoride compounds, and promotes the generation of the intermediate product SOF4.
  • (3) H2O plays a role in promoting the generation of the final and stable oxygen-containing-sulfur- fluoride compounds, such as SO2F2and SOF2.However, because the hydrolysis reaction rate of SF4is nearly two orders of magnitude higher than SOF4, the chemical reaction raterof SF6and H2O under PD is proportional to the concentration of H2O. Thus, with the growth of the concentration of H2O in SF6, the growth rate of SOF2is significantly higher than that of SO2F2. From the foregoing generalizations, the impact of H2O on SOF2is significantly higher than the impact of SO2F2. Thus, the formation of SOF2has a positive linear association with the concentration of H2O.
- 4.2 Analysis of the impact of O2on the characteristic decomposed component
In the case of SF 6 mixed with O 2 , under the impact of high-energy electrons produced by PD, in addition to reactions (6) and (7), the following reactions will occur:
PPT Slide
Lager Image
Besides the fact that the free state O generated by reaction (17) will react with SF 5 generated by PD and generate SOF 4 , the action below will happen and generate SOF 4 :
PPT Slide
Lager Image
Then, both SOF 4 and SF 4 react with the H 2 O released by the electrodes and the internal wall of the decomposition equipment and generate SO 2 F 2 , and SOF 2 . While O 2 exists, SF 2 will be involved in the following reaction:
PPT Slide
Lager Image
At present, the reaction rate constant of reaction (19) has not been found yet. Reference [8] has given the maximum rate constant k=5.0×10 −16 cm 3 /s. Similarly, under PD, the stoichiometric number b i of reaction (14) and the chemical reactions SF 6 and O 2 are involved in are also equal to one, and O 2 is always a small amount compared with a variety of low-fluoride sulfide SF x . Thus, the rate of reaction in (19) is proportional to the concentration of O 2 . The higher the concentration of O 2 , the more severe the reaction and the more SO 2 F 2 is generated.
For SO 2 F 2 , it can be seen from Fig. 5 (a) that when the concentration of O 2 mixed in SF 6 is less than 460ppm, the formation of SO 2 F 2 decreases with the increase of O 2 . When the concentration of O 2 is higher than 460ppm, the concentration of SO 2 F 2 is positively correlated with the concentration of O 2 because an increase in the concentration of O 2 in SF 6 is equivalent to the dilution of SF 2 , SF 4 , SF 5 , and other low-sulfur and fluorine F. Thus, O 2 plays an inhibitory effect on the reaction:
  • SFx(6 − x)F → SF6, x = 1 ~ 5
Although the concentrations of SF 2 , SF 4 , SF 5 , and other low-sulfur components increase with the discharge and promote the reaction in (18) ~ (19), with the increase in the concentration of O 2 , the concentration of H 2 O released by the electrodes and the internal wall of the decomposition equipment is diluted, making the rate of reactions in (8) to (11) and (15) decrease. Reaction (19) is at lower status when competing with reaction (14), (15) and (18), thus leading to the reduction in the amount of SO 2 F 2 generated within 10 hours.
However, when the concentration of O 2 is higher than 460ppm, with a further increase of O 2 , the rate of reaction in (18) undergoes a significant increase, and the rates of reaction in (8) to (11) and (15) are no longer significantly reduced. This time, since the concentration of O 2 is high, reaction (19) plays a dominant role in the generation of SO 2 F 2 when competing with reaction (14), (15), (18). Thus, when the concentration of O 2 is above 460ppm, the yield of SO 2 F 2 increases with the increase of O 2 . Hence, with a low concentration of O 2 (the concentration of O 2 < 460 ppm), the dilution of the inherent moisture in the device is the most important factor that affects the formation of SO 2 F 2 and the stability of the decomposition under PD. Furthermore, the reactions in (8) to (11) and (15), (18) play leading roles in the formation of SO 2 F 2 . But at high concentration of O 2 (the concentration of O 2 > 460ppm), reaction (19) plays a leading role in the generation of SO 2 F 2 , as shown in Fig. 5(a) .
For SOF 2 , its formation always decreases with the increase of O 2 , but the reduction is not obvious when the concentration of O 2 is higher than 460ppm. The reason for the practically unobservable reduction is that with the increase of O 2 mixed in SF 6 , SF 2 , SF 4 , SF 5 , and other low-sulfur, fluorine F undergoes a dilution process, thus playing an inhibitory effect on the reaction:
  • SFx(6 − x)F → SF6, x = 1 ~ 5
Although the concentrations of SF 2 , SF 4 , SF 5 , and other low-sulfur components increase with the discharge, as the concentration of O 2 increases, the concentration of H 2 O released by the electrodes and the internal wall of the chamber is diluted at the same time. The rate of the reaction which plays a decisive role in the generation of SOF 2 is shown in the following reaction:
  • SF4+ H2O → SOF2+ 2HF, k = 1.5 × 10−19cm3/s
will decrease with the decrease in the concentration of H 2 O. This phenomenon is most prominent when H 2 O is diluted (the concentration of O 2 < 460ppm). With further dilution of H 2 O (the concentration of O 2 > 460ppm), the decrease in the reaction rate is not obvious, resulting in a significant decrease in the formation of SOF 2 with the increase of O 2 when O 2 is at a low concentration. When O 2 is at a high concentration, the decrease in the formation of SOF 2 is not obvious with the increase of O 2 , as shown in Fig. 5(b) .
In summary, the concentration of H 2 O in the reaction chamber decreases because of the dilution effect of O 2 , resulting in the reaction rate of a series of reactions in which H 2 O decreases and the yield of SOF 2 decreases with the increase of O 2 . However, as O 2 promotes the formation of SO 2 F 2 , at the same time, the formation of SO 2 F 2 has a U-shaped relationship curve with the concentration of O 2 .
5. Conclusion
  • (1) Both H2O and O2influence the main characterisitic components SO2F2and SOF2during PD, but their by-products and degrees of influence are different. The influence of H2O on SOF2is the most significant and the formation of SOF2has a positive linear relationship to the concentration of H2O while its influence on SO2F2is not obvious. The concentration of O2influences the formation of both SO2F2and SOF2while the influence is much more obvious on SOF2.
  • (2) H2O has the ability to catch an F atom and to inhibit the low-fluoride sulfide SFxby recombining to SF6, which increases the concentration of SF5and SF4. H2O offers OH and O for the formation of oxygenated-sulfur fluoride, which creates a favorable condition for the ultimate formation of SO2F2and SOF2. However, the hydrolysis rate of SF4is much higher than the hydrolysis rate of SOF4(nearly two orders of magnitude higher), as a result, the increase in the rate of SOF2is much higher than that of SO2F2when the concentration of H2O increases.
  • (3) When the concentration of O2is low, the content of H2O in the equipment is the main factor which influences the fomation of SO2F2. When the concentration of O2is high, the reaction SF2+ O2→ SO2F2contributes mostly to the fomation of SO2F2, Thus, O2is the main factor. As for SOF2, an increase of concentration would diminish the H2O concentration, in which case O2becomes the most important factor in the decrease of SOF2.
  • (4) The trace levels of H2O and O2play key roles on the formation of characteristic decomposed components of SF6during PD and have significant influence on the products, so it is necessary to study the decomposition mechanism of SF6under different concentrations of H2O and O2under the long run PD, and research on different concentrations will help achieve sufficient knowledge on what influences the regularity in the reactions to propose correction methods accordingly. Acquiring sufficient knowledge on the decomposition mechanism and the factors that affect variation in the reactions under PD will lay a solid foundation in using decomposed components of SF6to assess insulation status and will support related repair guidelines for gas insulated electrical equipment.
Acknowledgements
The research work has been funded by National Natural Science Foundation of China (Grant No. 51177181), Key Technology R&D Innovation Program of Hubei Province (Grant No. 2014AAA015) and Foundation of State Key Laboratory of Power Transmission Equipment & System Security (Grant No. 2007DA10512714102). The authors sincerely thank the granting agency.
BIO
Fuping Zeng was born in Chongqing, China in 1984. He obtained his bachelor’s degree in electrical engineering from Dalian Maritime University, and his master’s and doctoral degrees from the State Key Laboratory of Power Transmission Equipment & System Security and New Technology, Chongqing University, China. He is currently a postdoctoral research at the School of Electrical Engineering, Wuhan University, China. He is involved in the online monitoring and fault diagnosis of high-voltage electric equipment insulation. E-mail: Fuping.Zeng@whu.edu.cn
Ju Tang was born in Pengxi, Si Chuan Province, China in 1960. He obtained his bachelor’s degree from Xi’an Jiaotong University, and his master’s and doctoral degrees from Chongqing University. Dr. Tang is a professor at the State Key Laboratory of Power Transmission Equipment and System Security and New Technology, Chongqing University; a professor at the School of Electrical Engineering in Wuhan University, China; and the chief scientist presiding over the National Basic Research Program of China (973 Program) (2009CB724500). He is currently involved in the online monitoring and fault diagnosis of high-voltage electric equipment insulation.
Yanbin Xie was born in Chongqing, China, in 1980. He received the Bachel / Master and doctoral degrees in electrical engineering from Chongqing University, China. He is now working in State Grid Chongqing Electric Power Company, Shiqu Power Supply Company, China, and he is involved in online highvoltage detection equipment and signal processing. E-mail:yanbinse @126.com
Qian Zhou was born in Kaifeng, Henan Province, China, in 1980. She received the Bachel/Master and doctoral degrees in electrical engineering from Chongqing University, China. She is now working in State Grid Chongqing Electric Power Company, China, and she is involved in on-line high-voltage detection equipment and signal processing. E-mail: cquzhou@163.com
Chaohai Zhang was born in Harbin, Heilongjiang Province, China in 1966. He obtained his bachelor’s degree from Harbin Institute of Technology, his master’s from Naval University of Engineering, and his doctoral degrees from Hong Kong Polytechnic University. Dr. Zhang is now working in Wuhan NARI Limited Company of State Grid Electric Power Research Institute, China. He is currently involved in the online monitoring and fault diagnosis of high-voltage electric equipment insulation.
References
1988 “By-Product Formation in Spark Breakdown of SF6/O2 Mixtures,” Plasma Chemistry and Plasma Processing 8 (2) 247 - 262    DOI : 10.1007/BF01016160
Van Brunt R. J. 1989 “Processes Leading to SF6 Decomposition in Glow-Type Corona Discharges,” Physics of Ionized Gases 161 - 172
Van Brunt R. J. , Sieck L. W. , Sauers I. , Siddagangappa M. C. 1988 “Transfer of F− in the Reaction of SF6− with SOF4: Implications for SOF4 Production in Corona Discharges,” Plasma Chemistry and Plasma Processing 8 (2) 225 - 246    DOI : 10.1007/BF01016159
Pradayrol C. , Casanovas A.M. , Aventin C. , Casanovas J. 1997 “Production of SO2F2, SOF4, (SOF2+SF4), S2F10, S2OF10 and SO2F10 in SF6 and (50-50) SF6-CF4 Mixtures Exposed to Negative Coronas,” Journal of Physics D-Applied Physics 30 (9) 1356 - 1369    DOI : 10.1088/0022-3727/30/9/011
Casanovas A. M. , Casanovas J. , Lagarde F. , Belarbi A. “Study of the Decomposition of SF6 under DC Negative Polarity Corona Discharges (Point-to-PlaneGeometry) - Influence of the Metal Constituting the Plane Electrode,” Journal of Applied Physics 72 (8) 3344 - 3354    DOI : 10.1063/1.351456
Chu F. Y. 1986 “SF6 Decomposition in Gas-insulated Equipment,” IEEE Trarnsactions on Electrical Insulatiorn 21 (5) 693 - 725
Van Brunt R. J. , Herron J. T. 1990 “Fundamental Processesof SF6 Decomposition and Oxidation in Glow and Corona Discharges,” IEEE Transactions on Electrical Insulation 25 75 - 94    DOI : 10.1109/14.45235
Van Brunt R. J. , Herron J. T. 1994 “Plasma Chemical-Model for Decomposition of SF6 in a Negative Glow Corona Discharge,” Physica Scripta T53 9 - 29    DOI : 10.1088/0031-8949/1994/T53/002
Tang Ju , Zeng Fuping , Pan Jianyu 2013 “Correlation Analysis Between Formation Process of SF6 Decomposed Components and Partial Discharge Qualities,” IEEE Transactions on Dielectrics and Electrical Insulation 20 864 - 875    DOI : 10.1109/TDEI.2013.6518956
Zeng Fuping , Tang Ju , Fan Qingtao , Pan Jianyu , Zhang Xiaoxing 2014 Decomposition Characteristics of SF6 under Thermal Fault for Temperature below 400℃ IEEE Transactions on Dielectrics andElectrical Insulation 21 (3) 995 - 1004    DOI : 10.1109/TDEI.2014.6832242
Tang Ju , Zeng Fuping , Zhang Xiaoxing , Pan Jianyu 2014 Relationship between Decomposition Gas Ratios and Partial Discharge Energy in GIS, and the Influence of Residual Water and Oxygen IEEE Transactions on Dielectrics and Electrical Insulation 21 (3) 1226 - 1234    DOI : 10.1109/TDEI.2014.6832269
2004 IEC60480-2004, “Guidelines for the checking and treatment of sulfur hexafluoride (SF6) taken from electrical equipment and specification for its re-use,”
Van Brunt R. J. , Siddagangappa M. C. 1988 “Identification of Corona Discharge-Induced SF6 Oxidation Mechanisms Using SF6/(18O2)/(H216O) and SF6/(16O2)/(H218O) Gas-Mixtures” Plasma Chemistry and Plasma Processing 8 (2) 207 - 223    DOI : 10.1007/BF01016158
Espenson James H. 1995 “Chemical Kinetics and Reaction Mechanism,” McGraw-Hill Press, Inc. New York Ch.2 39 - 56
Derdouri A. , Casanovas J. , Hergli R. , Grob R. , Mathieu J. 1989 “Study of the Decomposition of Wet SF6, Subjected to 50-Hz AC Corona Discharges,” Journal of Applied Physicssa 65 (5) 1852 - 1857    DOI : 10.1063/1.342919
Tang Ju , Liu Fan 2012 “Partial Discharge Recognition by Analysis of SF6 Decomposition Products Part 1: Decomposition Characteristic of SF6 under Four Different Partial Discharges,” IEEE Trans. Dielectrics and Electrical Insulation 19 (1) 29 - 36    DOI : 10.1109/TDEI.2012.6148499
Casanovas A. M. , Coll I. , Casanovas J. 1999 “Decomposition Products From Negative and 50 Hz AC Corona Discharges in Compressed SF6 and SF6/N2(10:90) Mixtures. Effect of Water Vapour Added to the Gas,” Journal of Physics D-Applied Physics 32 (14) 1681 - 1692    DOI : 10.1088/0022-3727/32/14/321
Tsang W. , Herron J. T. 1992 “Kinetics and Thermodynamics of the Recation SF6 Reversible SF5 + F,” Journal of Chemical Physics 96 (6) 4272 - 4282    DOI : 10.1063/1.462821
Shao J. 2009 “Mathematical statistics,” World Publishing, Co. Ch.6 81 - 93