International Journal of Emerging Technology and Advanced Engineering Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 7, July 2013)

Assessing Partial Discharge on Composite Insulators Under Desert Pollution Conditions Mohammed El-Shahat, Hussein Anis1 Department of Electrical Power, Cairo University, Egypt Abstract— Transmission lines located in the Sinai desert are subjected to desert climate, one of whose features, especially in the spring, is sand storms. Line insulators are thus exposed to sand contamination, which may produce streamer discharges in the air surrounding insulator surface. Such discharges may lead to flashover. This paper examines the electric field and potential distributions on sand-polluted composite insulators. These insulators are nominated to replace ceramic insulators on transmission lines in Sinai. A case study is presented using an ABBdesigned composite insulator, where the insulator surface electric field is investigated under dry sand-pollution conditions and with different sand grain sizes. The latter were provided by a previous study of the Sinai desert. Finite element techniques are used to simulate the insulator and seek electric field distribution on the polluted insulator model and around it. This study helps in partially assessing –and thus justifying- the use of these insulators in those critical areas.

Under desert dry conditions sand grains deposited on the insulator surface intensifies the electric field on the surface and may thus lead to the formation of localized discharges in the air surrounding the insulator. Discharges in air may lead to flashover. Meanwhile, even without materializing complete flashover those discharges over long periods of time can cause the insulator to age. Deterioration by aging is of great concern in the performance of silicon rubber insulator (B.A.Arafa et al., 2012; M. M.Awad et al., 2002; B. Zegnini et al., 2009; W. Que, et al 2002; Abdel-Salam H.A. Hamza et al., 2002). Artificial tests were performed on composite insulator to check the aging of insulator (M. A. R. M. Fernando et al., 2010; Z. X. Cheng, et al., 2003). An earlier study examined the pollution of power line insulators in the Sinai desert by air-borne sand (A. Mahdy et al., 2001). The distributions of sand grains size in the desert soil was acquired from random samples, where their sizes and salinity were measured. This paper studies the electric field distribution along composite insulator, and subsequently the chances of developing discharges in the surrounding air. A typical two-shed insulator, which is suitable for 220 kV power lines, is used as a case study. The surface location of maximum electric field and its magnitude are determined with different sand grain sizes. The results are compared to those under clean condition. Electric field distribution on clean composite insulator was reported (E. P. Nicolopoulou, et al 2011). Other studies simulated the electric field on polluted composite insulator using a 2-D computer program (WL Vosloo, et al 2002). In another study the electric field distribution along a 765 kV polluted composite insulator used the COULOMB program (W. Que, et al 2002). In the present study insulator simulation was carried out using a 3-D ANSYS software program, which is based on the Finite Elements method. The program needed higher performance computing and gave results with high accuracy.

Keywords— Composite insulators, Desert pollution, Partial discharge, ANSYS software, Sinai.

I. INTRODUCTION Insulators constitute an essential component of high voltage transmission systems, where their performance is critical to the reliability of the power line. To maintain system reliability insulator service life should be increased by keeping electric field below the threshold of discharge in air. Meanwhile, composite insulators are widely used because of their lower weight, higher mechanical strength, higher design flexibility, and their reduced maintenance. They display lower leakage current due to their higher surface resistance. Furthermore, a silicone rubber insulator hinders flashover under pollution because of its hydrophobic recovery or water repellent feature and its low free surface energy (Z. X. Cheng et al., 2003; Yong Zhu et al., 2005). However, because of the good hydrophobicity water droplets will be formed on insulator surface with some conductive contamination causing discharge to occur and insulator would then lose its hydrophobicity allowing dry bands to form leading to surface discharges and the insulator to age. In desert climate in Sinai -especially during and after sand storms- sand grains are carried at high speeds, which have different sizes depending on the type of soil in this region.

II. MATERIALS AND METHODS A. Insulator Computational Model This paper uses a 220 kV ABB silicone rubber insulator as shown in Fig.1; its dimensions are given in Table I

492

International Journal of Emerging Technology and Advanced Engineering Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 7, July 2013) TABLE I COMPOSITE INSULATOR DIMENSIONS Silicone rubber insulator dimensions Inner diameter 1 Inner diameter 2 Length 1 Length 2 Length 3 Maximum length Distance between two sheds Height of long shed Height of small shed

Symbol D1 D2 L1 L2 L3 Lmax S P P1

Linear variation of potential is assumed within the triangular elements, i.e. uniform field is assumed within the element. Consequently:

Value (mm) 250 219 680 855 470 2005 55 55 25

√ √

[ ]

=[

(2) ]

=[

[ ]

[ ]

(3)

]

[ ]

(4)

The next step was to assemble all such elements in the solution region. The energy associated with the assemblage of all elements in the mesh is:

=179.629 kV at the other side.

W=∑

The analysis produced the directional electric field (its X, Y, and Z components) and the resultant total electric field along the insulator with a clean surface. The distribution of the maximum is graphically shown in Fig.3 for only a specific sector of the insulator‘s 13 sheds, where the absolute maximum field was found to exist.

= ε [V]‘[C][V]

(5)

where, [ ]= [ ] , and

[C] is called the overall or global coefficient matrix, which is the assemblage of individual element coefficient matrices. The resulting equation is:

B. Sample Insulator Sector It is both a tedious task and unnecessary to microanalyze the potential and field distributions on the entire insulator. Instead, a sample sector of the insulator was selected, where the absolute maximum field materialized on the insulator surface. The insulator sector has two sheds; one shed is long and the other is short with a total creepage distance of 186.14 mm. The sample sector is shown in Fig.4. To enable the analysis of the selected sector its boundary conditions obtained from the global boundary conditions were imposed on that sector.

W= ε [

][

][ ]

(6)

where, subscripts f and p, respectively, refer to nodes with free and fixed (or prescribed) potentials. Since Vp is constant, differentiation is made only with respect to Vf , yielding [

] [ ]=- [

][ ]

(7)

The potential at the free nodes can be found by: [ ]=[

C. Computational Analysis The sample was subjected to examination using ANSYS based on FEM, which solves irregularly shaped boundaries. First, discretization is made for the model surface by a number of triangular finite elements. For the element "e" the potential polynomial distribution equation is: Ve (x,y,z)=a+bx+cy+dz

Ve

where n is the number of nodes , N is the number of elements. The coefficients a, b, c ,d are obtained from the above equation by:

The Unigraphics program was used to draw the insulator model in 3-D model and export it to the ANSYS program as shown in Fig.2. In ANSYS program the material of this insulator model was defined as silicone rubber. Appropriated mesh was then used for potential and field analysis. The potentials at the ends of the insulator were ground at one end and the peak phase voltage

Ee = -

]-1 .(- [

][ ])

(8)

The analysis was applied to clean insulators as well as insulators polluted by sand grains under different dry conditions of sand grain size. III. RESULTS AND D ISCUSSION A. Clean Insulator Based on the complete insulator simulation the sample section of Fig.4 was subjected to the resulting boundary conditions. The potentials on the two ends of the sample sector –as acquired from the global analysis- were 54.196 kV and 49.828 kV.

(1)

Where, the potential Ve in general is nonzero within element ‗e‖ but zero outside ―e‖; a, b, c, d are coefficients of the polynomial equation.

493

International Journal of Emerging Technology and Advanced Engineering Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 7, July 2013) The resulting electric field distribution along this model is graphically depicted in Fig.5, where the maximum electric field was computed to be 1.271 kV/cm.

C. Statistical Distribution of Electric Field 1) In The Vicinity of 220 kV Power Lines: Since sand grains in the desert are randomly varying in size, and since the maximum electric field, which is likely to cause surface discharges, depends on the grain size such a field is expected to be also random in magnitude. In the earlier study (A. Mahdy et al., 2001), sand samples were collected from regions near power lines, and frequency distributions of sand grain sizes were subsequently produced. A typical frequency distribution of the samples acquired near 220 kV lines is shown in Fig.9. The probability density distribution of these samples distribution is shown in Fig.10. The distribution is best fitted by a lognormal distribution whose function is

B. Sand Pollution Effect An earlier study (A. Mahdy et al., 2001) reported – based on the statistical distributions of sand grain sizes in Sinai- that sand grains with diameters in the range of 1 to 2 mm prevailed. The effects of those sand grains on the potential and electric field on insulator surface were thus sought. In the simulation, the sand grain was placed on the insulator surface in several, randomly selected, locations in the vicinity of where the clean-insulator field showed its maximum value. Following are the results of this situation. With sand grain sizes of 1, 1.5, and 2mm, the field distributions in those critical positions were computed. Computations were made with a 1 mm sand grain placed at the following locations (all values measured in mm): on the upper side of the long shed at (x = 162.15, y = 19.15), on lower side of long shed at (x = 112.1, y = 16.3), on the straight distance between sheds near the long one at (x = 110.2, y = 19.4), on upper side of short shed at (x = 133.1, y = 42.3), and on lower side of short shed at (x = 111.3, y = 42.4). The symbols x is the normal to insulator axis, y is in the direction of insulator axis, and all are measured from the neck point of the longer shed. The electric field distribution in the presence of a 1mm sand grain is shown in Fig.6, at the third location above. The field distribution profile is graphically plotted against the creapage distance in Fig.7. The same procedure was repeated for sand grains of sizes 1.5 and 2 mm. Results are summarized in Table II. Fig. 8 depicts the field distribution of the above cases as compared to the case of clear insulator. The figure shows the effect of grain size on the magnitude and location of the maximum electric field.

p(x, µ, σ) =

1.0 mm size 1.5 mm size 2.0 mm size

Location 2

Location 3

Location 4

Location 5

1.078

0.936

1.638

0.796

1.141

0.996

1.788

0.964

1.282

1.282

1.078

1.829

1.633

1.437

,x

(9)

2) Overall Desert: Sand samples were collected from different places in Sinai desert (A. Mahdy et al., 2001), and frequency distributions of sand grain sizes were subsequently produced. A typical frequency distribution of the samples acquired in all Sinai is shown in Fig.13.The corresponding probability density distribution is shown in Fig.14. The distribution is best fitted by a lognormal distribution, the mean of which is 0.401 mm, and the standard deviation is 0.346 mm with a square error of 0.11. The frequency of occurrence of the maximum electric field was mapped by that of the sand grain size using the above methodology. The result is seen in Fig 15. The corresponding maximum electric field probability density is seen in Fig 16. The best fit to that distribution was found to be a Gama distribution. whose function is p(x, α, β) =

1.160

σ√

Where the mean = µ, which in this case = 0.411 mm and the standard deviation = σ, which in this case = 0.0861 mm with a square error of 0.118. The frequency of occurrence of the maximum electric field was mapped by that of the sand grain size using the above methodology. The result is seen in Fig.11, with the corresponding maximum electric field probability density seen in Fig.12. The best fit to that distribution was found to be also a lognormal distribution, the mean of which is 1.3 kV/cm and the standard deviation is 0.113 kV/cm with a square error of 0.118.

TABLE II Maximum Electric Field (Kv/Cm) In Different Grain Sizes Locations Location 1

µ σ

α

β

(10)

The parameters of this distribution α = 157.196 and β = 158.3 and the mean of this Gama distribution is 1.29 kV/cm, and the standard deviation is 0.45 kV/cm with Square Error = 0.0353.

494

International Journal of Emerging Technology and Advanced Engineering Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 7, July 2013) Published works indicate that –in air- avalanche are likely to form streamers, i.e. create discharges, when they acquire a volume indicated by a multiplication factor K in excess of 5 (V. Chirokov et al., 2001). According to Fig. 20, it is concluded that –for the case study and under dry conditions- sand grains of sizes greater than 1.7 mm are not likely to produce ionization in the air surrounding the insulator. Sand grains of smaller sizes would produce ionization; however, they are not likely to produce discharge streamers on the insulator‘s surface. It is worth mentioning that under wet conditions those same sand grains may present a certain risk of insulator failure as proven in another paper.

D. Grain Size-Based Discharge Placing the sand grains on the insulator surface causes the electric field in the air surrounding the surface to be highly non-uniform, which may lead to discharge in the air. The extent of this possibility with different sand grain sizes is investigated. The grain sizes used in this investigation are between 0.8 mm and 1.5 mm, which were found to be the most prevailing in the previous study (A. Mahdy et al., 2001). 1) Streamers Orientation: Discharges in air are said to occur when field-produced ionization avalanches develop into streamers. To determine the trajectory, along which discharges in air will most likely occur, the electric field profile in different directions off the grain surface is first explored for each case. In this section the above concept is applied to the case of 0.8 mm grain size as an example. The field trajectories in five different directions (angles 300, 150, 00, -150, -300) all starting at the grains‘ outer surface and extending into air are shown in Fig.17. The resulting electric field profiles are shown in Fig.18. Since the profile with the highest electric field is seen at angle 00 (normal to the surface), further computations of avalanche size are confined to this trajectory.

IV. CONCLUSIONS 1. The potential and electric field distributions on the surface of a composite high voltage insulator was computed using a 3-D finite element method assisted by a Unigraphics program to draw the insulator model and export it to the computational routine. 2. To save computational time, that sector of the insulator, where the absolute maximum field occurred on the insulator surface, was identified and then subjected to the field and discharge microanalysis. 3. Electric field distributions were sought both under clean and sand-polluted conditions. While on a clean insulator the maximum field was 1.271 kV/cm it reached a value of 1.829 kV/cm in the presence of 2 mm sand grains. 4. The statistical distribution of sand grains in the desert was obtained, where it was best described by a Log-normal distribution. When the entire desert territory was considered, that distribution was mapped to the maximum electric field, whose distribution was then found to fit a Gamma distribution with a mean of 1.29 kV/cm and a standard deviation 0.45 kV/cm. 5. Possible discharges in the air surrounding the insulator under polluted dry conditions were investigated using physical ionization parameters. It was found that grain sizes in excess of 1.7 mm produced too low electric field to cause ionization of air. Smaller grains would cause ionization but are not likely to lead to self-sustained discharges.

2) Discharge With Different Grain Sizes: Ionization and electron-attachment simultaneously occur in air and may produce an avalanche. Their variation with the field E and gas pressure P is expressed by the following equations:α/p = A

(11)

α is the ionization and the constants A and B for air are A= 4.7786 cm-1. torr1 , B= 221 V.cm-1. torr-1 . while, electron attachment is approximated by the quadratic relation: η/P = A1 + B1 (E/P) + C1 (E/P)2 (12) η is the attachment, and the constants A1, B1 and C1 for air are A1= 0.01298 cm-1.torr-1 , B1= - 0.541 10-3 V.cm-1. torr-1 , C1= 0.87 10-5 V2.cm-2. torr-2 . The multiplication factor (K) that represent discharge is: K=∫

α

η

REFERENCES

(13)

[1]

X0 is the distance from grain size face in air at which the ionization and attachment have equal values; this is graphically shown in Fig. 19 for the case of 0.8 mm sand grain, in which case K equals 4.6724. Fig.20 depicts the variation of the multiplication factor with grain size.

[2]

495

Mahdy, A., Anis, H., Amer, R., El-Morshedy, A. 2001. Insulator pollution assessment in Sinai using Geographic Information Systems. Proceedings of the Middle East Power Conference (Mepcon 2001). Cheng, Z. X., Liang, X. D., Zhou, Y. X., Wang, S. W., Guan, Z. C. 2003. Observation of corona and flashover on the surface of composite insulators. Bologna Power Tech Conference, Bologna, Italy.

International Journal of Emerging Technology and Advanced Engineering Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 7, July 2013)

[4]

[5]

[6]

[7]

[8]

[9]

[10]

Fernando, M. A. R. M., Gubanski, S. M. 2010. Ageing of silicone rubber insulators in coastal and inland tropical environment. IEEE Transactions on Dielectrics and Electrical Insulation Vol. 17, No. 2. Arafa, B., Nosseir, A. 2012. Effect of Severe Sandstorms on the Performance of Polymeric Insulators. PARIS CIGRE, F-75008 PARIS, Session D1-104. Awad, M. M., Said, H. M., Arafa, B. A., Sadeek, A. 2002. Effect of Sandstorms With Charged Particles on The Flashover and Breakdown of Transmission Lines. PARIS CIGRE, F-75008 Paris, Session 15-306. Hamza, H. A., Abdelgawad, N. M. K, Arafa, B. A. 2002. Effect of desert environmental conditions on the flashover voltage of insulators. Energy Conversion and Management Volume 43, Issue 17, Pages 2437–2442. Zhu, Y. , Otsubo, M. , Honda C., Tanaka, S. 2005. Loss and recovery in hydrophobicity of silicone rubber exposed to corona discharge. http://dx.doi.org/10.1016/j.polymdegrads tab.2005. Zegnini, B., Mahi, D., Chaker, A. 2009. Modeling parameters optimizations of 750kV insulators flashover voltage under pollution conditions in high altitude areas using RBF artificial neural networks. Acta Electrotechnica et Informatica, Vol. 9, No. 4. Que, W. , Stephen, A. 2002. Electric field and voltage distribution along non-ceramic insulators. http://www.tcipower.com/list1.asp?id=366. Nicolopoulou, P., Gralista, E. N., Kontargyri, V. T., Gonos, Stathopulos, I. F., I. A. 2011. Electric field and voltage distribution around composite insulators. XVII International Symposium on High Voltage Engineering, Germany.

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

Satheesh, G., Basavaraja, B., Nirgude, P. M. 2012. Electric field along surface of silicone rubber insulator under various contamination conditions using FEM. International Journal of Scientific & Engineering Research Volume 3, No. 5. Vosloo, W.L., Holtzhausen Eskom J.P. 2002. The electric field of polluted insulators. Africon Conference in Africa, IEEE AFRICON, 6th, Volume: 2. Gao, H., Jia, Z., Mao, Y., Guan, Z., Wang L. 2008. Effect of Hydrophobicity on Electric Field Distribution and Discharges along Various Wetted Hydrophobic Surfaces. IEEE Transactions on Dielectrics and Electrical Insulation Vol. 15, No. 2. Akyuz, M., Gao, L., Cooray, V., Gustavsson, T. G., Gubanski, S. M., Larsson, A. 2001. Positive Streamer Discharges along Insulating Surfaces. IEEE Transactions on Dielectrics and Electrical Insulation, Vol. 8, No.6. Du, X., Liu, Y. 2011. Pattern Analysis of Discharge Characteristics for Hydrophobicity Evaluation of Polymer Insulator. IEEE Transactions on Dielectrics and Electrical Insulation Vol. 18, No. 1. Sarma, P., Janischewskyj, W. 1969. D.C. corona on smooth conductors in air Steady-state analysis of the ionization layer. PROC. IEE, Vol. 116, No. I. Katada, K., Takada, Y., Takano, M., Nakanishi, T., Hayashi, Y. , Matsuoka, R. 2000. Corona Discharge Characteristics of Water Droplets on Hydrophobic Polymer Insulator Surface. Proceedings of the 6th International Conference on Properties and Applications of Dielectric Materials. Chirokov, V. 2005. Stability of Atmospheric Pressure Glow Discharges. Ph.D. Thesis. Drexel University, Philadelphia.

1.4

1.2

Electric field (kv/cm)

[3]

1

0.8

0.6

0.4

0.2

0

0.4

0.6

0.8

1

1.2

1.4

1.6

Creapage distance (m)

Fig.3: Electric Field Distribution for 13 Sheds

Fig.1: Insulator Shape with the Shed as In ABB Design Guide

Fig.4: Sample Sector of Composite Insulator

Fig.2: Unigraphics 3-D Model Insulator

496

International Journal of Emerging Technology and Advanced Engineering Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 7, July 2013) 40

35

30

Frequancy

25

20

15

10

5

0

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

Grain size (mm)

Fig.9: Total Sample Curve Near 220 kV T.L Fig.5: Electric Field Distribution on Clean Sample Sector 2.5

40

Fig.6: Maximum Electric Field Distribution of 1 mm Grain Size

2

1.5

20

1

Frequancy

Probability density function (1/mm)

Grain frequancy probability density function

0.5

0

0

0.5

1

1.5

2

2.5

0 3

Grain size (mm)

Fig.10: Probability Density of Grain Sizes near 220 kV

Fig.7: Electric Field Profile Curve of 1 mm Grain Size

Fig.11: Maximum Electric Field Distribution Referring to Grain Sizes

Fig.8: Maximum Electric Field Distribution with Different Grain Sizes

497

International Journal of Emerging Technology and Advanced Engineering Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 7, July 2013)

Fig.15: Maximum Electric Field Distribution Referring to Grain Sizes in All Sinai

Fig.12: Probability Density Function of Maximum Electric Field Near 220 kV 45

40

35

Frequancy

30

25

20

15

10

5

0

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

Grain size (mm)

Fig.13: Total Sample Distribution in All Sinai Fig 16: Probability Density Function of Maximum Electric Field in All Sinai 2.5

50

2

40

1.5

30

1

20

0.5

10

0

0

0.5

1

1.5

2

2.5

Frequancy

Probability density function (1/mm)

probability density function grain frequancy

0 3

Grain size (mm)

Fig.14: Probability Density Function of Grain Sizes in All Sinai

Fig. 17: Insulator With Different Analysis Lines

498

International Journal of Emerging Technology and Advanced Engineering Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 7, July 2013) 38

6 0 angle 15 angle 30 angle -15 angle - 30 angle

36

5

Multiplication factor (K)

Electric field (kv/cm)

34

32

30

28

26

24

22

3

2

1

20

18

4

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Distance (mm)

0 0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

Grain size (mm)

Fig.18: Electric Field Distribution for Different Angles Fig.20: Multiplication factor versus grain size of sand 40

Electric field (kV/cm)

35

30

25

20

15

10

5

0

x0

0.5

1

1.5

2

2.5

3

Distance (mm)

Fig.19: Electric Field in Air Surround Insulator Near 0.8 mm Size

499