‘Ain Shams Engincering Journal (2018) 9, 1349-1356 Ain Shams University Ain Shams Engineering Journal www elsevier com locatease) Sworwesleneedinee com ELECTRICAL ENGINEERING Electric field distribution of optimized composite — @ coi insulator profiles under different pollution conditions El-Sayed M. El-Refaie, M.K. Abd Elrahman, M. Kh. Mohamed ~ Faculty of Engineering, Helwan University, Cairo 19177, Egypt Received 20 March 2016; revised 2 Available online 1 October 2016 ine 2016; accepted 17 August 2016 KEYWORDS Abstract This paper presents the simulation results of electric field distribution for optimally etre fed distribution signed medium voltage composite insulators under polluted conditions with and without water ‘Comporit insulators droplets. Four different profs of 11 kV composite insulators were used. Simulation was executed Pallation; for both typical and optimized insulators by using Comsol Multiphysics program. The main objec Water droples: tive ofthis paper was investigating the effect of pollution conditions and water droplets on the cee tric field distribution of the optimally designed insulators. The effect of pollution layer thickness and conductivity on the electric fied distribution was studied. The simulation results showed that the maximum electric field streses of optimally designed insulators were less than those of typically signed insulators in clean, polluted and wet conditions. The results of this study are able to pro: vide theorctical support to design and select the suitable profile af composite insulators in order to ‘obtain better performance under clean and polluted conditions. © 2016 Production and hosting by Elsevier B.V. on bebal of Ain Shams University. Thisisan open assess arti under the CC BY-NC-ND license (tp ereativecommons org leeases/by-ne-nd/4.0), 1. Introduction in comparison with porcelain and glass insulators. Composite insulators have many advantages such as good performance in Different types of insulating materials can be used to insulate contaminated environments duc to their hydrophobic nature, transmission lie conductors from the towers. Nowadays siz weigh, considerably low instalation eos, easy handling Con ubher ulstorsare commonly used dus teitesuperority and [ow maintenance (, The electric field distribution along oe im deren pea ‘composite insulators affects both the long- and short-term per- formances, fundamental understanding of the electric Feld distribution and it effet on the insulator performance during ‘lean, polluted and wet conditions i very important in design ing and applying composite insulators effectively 2) Composite insulators used in distribution, transmission lines, an substations are subjscted (o various environmental conditions, As these conditions are different from a region to another, these inslators should withstand the electrical and * Corresponding author Email addresses: disiayme@ yahoo.com (E-SM. EFRefti), 1m_kamall968@ yahoo.com (MLK. Abd Elrahman), mustala_alid3@ hreng helwan.edu-eg (M. Kh. Mohamed), Peer review under responsibility of Ain Shams Univesity ‘iver | Production and hosting by Elsevier mechanical streses they are subjected to. The electri field hid doi.org/0. 1016.6). 2016.08.012 2000-41459 © 2016 Production and hosting by Ekevie B.V. on behalf of Ain Shams University This isan open acess article under the CC BY-NC-ND license (hp: /ereatvecommonsore boenses/byne- 4.0).1350 ESM. ELRefaie etal distribution of composite insulators depends upon many parameters such as applied voltage, insulator design, corona ring and hardware design, and phase spacing [2,3] Insulator design includes the number of sheds, their radii, the spacing between sheds, shed inclination angle and the radius of the metal end fittings. The optimum values of these parameters increase the reliability of insulators by reducing the electric, field stress within and in the vicinity of insulators. Recently there was an attempt to introduce an eflicient ‘optimization technique to improve the profile of composite insulators [4]. Integration between the particle swarm opti- mization algorithm and Finite Element Method was used for this purpose. Two-dimensional models af four different pro- files of 11 KV composite insulators are established under elean ‘and dry condition, MATLAB and COMSOL Multiphysies are ‘combined to optimize electric field distribution of composite insulators. ‘The results indicate that the electric fiekd stress was reduced significantly by using the proposed technique. The electric field distributions of composite insulators have been studied at different pollution severity levels (5). The pol lution severity was classified aecording to its corresponding, pollution layer conductivity. The sinvulation results showed that electric field stress increases as the pollution severity increases. The highest electric field stress was observed at the jjunetion of the shed and sheath region “The electric field and potential distributions on two straight sheds silicone rubher insulators have been investigated under clean and various contamination conditions [6]. Two different Contaminations of Plywood and cement dust are used in this, simulation. Contaminants and water droplets have no effect, ‘on potential distribution along the insulator surface. However, they caused highly non-uniform electric field distributions especially on the trunk portion, Water droplets play sever roles in the pollution Mashover and aging of insulators. The presence of a water drop on the insulator surface increases, the electrical field intensity along the insulator surface which ‘may lead to an electrical breakdown [7-10 Tn this paper the electric field distribution along the surface ‘of composite insulators has been studied under different condi- tions for typically and optimally designed insulators to check the effectiveness of optimization technique for composite insu- lators during polluted and set conditions as in clean and dry conditions. 2. The proposed optimization technique In a previous research paper [4) the authors have investigated the distribution of electric field along the surface of optimized composite insulators for clean and dry conditions by using Finite Element Method (FEM) with the aid of commercial ile Edt Vew Options Hep oees SOQ d-s& 7 Meteor = epic Rreer eco eam 408 mod enagh ode 1 GeblDions . Pau pen) 4) Model {mod} 1 Dian 6p ty ee Bed Sten Bi viewt ent) A Geet gen Miwa rot © sed ina) © com {mat} © icone ruber ots 4 Shae © Chae Connon °5 Zero Cherget fl} © ta ft © Gendt Blectric Potential (pot) (@ Mesh fest) & suse Gi Resuts 200 Pega oy Ret xece-|QQgeel|t-\etOselm 450° ‘100° '50 0 so 100. "1502002 COMSOL Multiphysics graphical user interface.Electric field distribution of optimized composite insulator profiles under different pollution conditions 1351 (b) Model 1 © ® fa) (b) Model igure 2 optimization and (b) after optimization sofware package COMSOL Multiphysics [11]. The particle swarm optimization (PSO) algorithm has been implemented using MATLAB program. COMSOL Multiphysics and MATLAB are combined to conduct optimization process for the insulator profile structure. The optimized variables are insu lator shed diameter, shed spacing and metallic end fitting diam- eter. Typical dimensions of the composite insulator forthe fist ‘model have been taken from [3]. whereas typical dimensions of the second model have been taken from (12). The typical dimen- sions of the third and the fourth models have been taken from the catalogue of the factory of composite insulators (UMR) in the industrial zone Badr city, Cairo, Egypt ‘The variables are updated by using PSO through MATLAB program until the optimum value of electric leld stress is obtained: each variable is changed within a recommended range according (0 the composite insulator profile design of TEC/TS GOSIS-3 [13]. The maximum value of electric field stresses reduced significantly for all models during clean and dry conditions by using the proposed optimization technique, and the maximum percentage of reduction was about 17% (4) (b) Model2 | i (a) (b) ‘Model 4 Composite insulator profile of four different simulation models during uniform dry pollution conditions (a) before 3. Simulation of composite insulators The composite insulators consist of a metal fitting at both ends usually made of forged stecl and a Fiber Reinforced Plastic (FRP) core as a load bearing structure, where silicon rubber is used for weather sheds and sheath due to its hydrophobic nature. The four insulator models were simulated in Comsol Multiphysics as 2D geometry as shown in Fig. 1. Initially, ‘Table 1_ Properties of material used for insulators modeling Material ‘Relative penalty (Sim) Forged seel 1 Air background 1 Silicone uber 4.3 FRP core 72 Pollution layer 7-1 Water droplets 80)1382 ESM. ELRefaie etal @) o) Figure 3 Simulation of the third model with water droplets on clean insulator surface (a) before optimization and (b) after ‘ptinization. o Oy Figure 4 Simulation of the thi! model with water droplets on polluted insulator surface (a) before optimization and (b) after ‘optimization. the simulation was carried out for a dry and clean insulator and electric field distribution along the insulator surface is calculated and plotted. To simulate the pollution severity for {ypical and optimized insulator models & uniform pollution layer of | mm thickness was considered at the insulator surface 4s shown in Fig. 2. Input parameters for the COMSOL Mull physics program are the geometrical dimensions of insulators, permittivity, and conductivity oF the materials and the bound ary conditions (potential of metal parts). In this study, four different composite insulator profiles of 11 KV were selected, to investigate the ability of the proposed optimization tech- nique in different conditions. The relative permittivity and clectrie conductivity of each domain are listed in Table | ‘The water droplets on the sheds and the shank of model 3 are simulated for typical and optimized insulator dimensions in two cases the first case is water droplets on a clean insulator surface as shown in Fig. 3, and the second ease is water dro- plets on uniformly polluted surface as shown in Pie. Each drop is modeled with 4.0 mm of diameter and a contact angle with the insulator surface of 90°. The technical specifications aand the typical geometrical dimensions forall insulators used, in simulation models are listed in Table 2. 4, Simulation results and discussions 4.1. Electric field stress under dry uniform pollution condition All insulator models are designed to be polluted with uniform pollution layer for a dry condition and the electric field was calculated for both typical and optimized insulator dimension, ‘models. The pollution permittivity is 7.1 and its eleetrieal con- ductivity is 0.6 (4Sim, Fig. 5 shows the maximum elects field stress of the first ‘model when the surface of the insulator is polluted uniformly ‘with | mm thickness of pollution layer. As seen, the maximum, clectric field stress of the optimized model is 109 (kim). whereas the maximum electric field stress of the typical model is 126 (k¥/m). Fig. 6 shows the maximum electri field stress of the second model when the surface of the insulator is polluted, ‘uniformly with 1 mm thickness of pollution layer. As seen, the ‘maximum clectrie field stress of optimized model is 118 (kV/ ‘m), whereas the maximum electric field stress of the typical ‘model is 136 (kV/m). Fig, 7 shows the maximum clectre fetd stress of the third model when the surface of the insulator is polluted uniformly with I mm thickness of pollution layer. [As seen, the maximum electric field stress of optimized model is 133 (kV/m), whereas the maximum cleetre field stress of the typical model is 144 (kV/m). Fig. 8 shows the maximum elec- tric fcld stress of the fourth model when the surface of the insulator is polluted uniformly with 1 mm thickness of pollu tion layer. As soon, the maximum electric fold stress of opt ‘mized model is 132 (kV/m) whereas the maximum electric field stress of the typical model is 146 (kV /m). Its clear that from the results of Figs. 5-8 the maximum electric field stress reduced significantly for optimized! models compared to typical models under pollution condition, 4.2, Electric field stress under water droplets To investigate the optimization results, the third model has ‘been tested under clean and dry conditions and under dry pol- luted conditions and in these two cases the model proves the ability of optimization technique in reducing the electric field, and now the third model will be tested for clean and wet con- ditions and for wet and polluted conditions.Electric field distribution of optimized composite insulator profiles under different pollution conditions 1353 ‘Table 2 Technical specications of insulators used in simulation. Structure parameters Model Model 2 Model 3 Model 4 Shed arrangement ype Uniform shed Uniform shed ‘Alternating shed ‘Altemating shed Creepage distance (mm) 430 225 325 495 ‘Arcing distance (mm) 180 190 8 18 Sectional length (mam) 360) 216 318 ais Core thickness (mm) 18 is 6 6 No, of sheds 4 3 5 5 Sted diameter (mn) 85 50 sajos Iajos Shed spacing (mm) 8 ss 0 it ™ 7 c ‘pete e883 EEectte Feld intensity (kV) seas wa np Dt Figure $ Electric fild stresses of original and optimized inst lator dimensions for the frst model under uniformly. polluted ‘condition ™ a ea Deais nid ses ‘a Figure 6 Electric field stresses of original and optimized insu- lator dimensions for the second model under uniformly polluted ‘conditions, Fig. 9 shows the maximum electric field stress of the third ‘model during water droplets on the clean surface of the insula tor. As seen, the maximum electri field stress of optimized ‘model is 273 (kV/m) whereas the maximum electric field stress i i i Figure 7 Electric field stresses of original and optimized insu: lator dimensions for the third model under uniformly potluted condition pe Dn a Figure 8 Electric fild stresses of original and optimized ins lator dimensions for the fourth model under uniformly polluted condition of the typical model is 355 (kV tion in the maximum eleetre field stress is about 28% means that the optimized model has excellent performance ‘The percentage of redue- snd this1384 ESM. ELRefaie etal Figure 9 Eletrc field stress of original and optimized insulator For water droplets on elean surface condition Fy insulator dimensions for water droplets on polluted surface condition. wre 10 Electric field strsses of original and optimized uring water droplet on a clean surfice compared to the typi- ceal model In power systems, composite insulators subjected to several weather conditions in service. The rugged service environment may consist in the form of Fog, dew, rain, and pollutants from the salty wind, ultraviolet radiation, chemicals from industries, ‘and other contamination. In this section, the effect of water {droplets on polluted insulator surface is studied for both typ- ical and optimized insulator designs. Fig. 10 shows the maximum electric field stress of the third model when there were water droplets on the polluted surface Of the insulator. AS seen, the maximum electric field stress of ‘optimized model is 120 (kV/im) whereas the maximum electric field stress of the typical model is 130 (kV m). This means that the optimized model still has a good performance during water droplet on the polluted surface with compared to the typical model = [-=a5mn| om ‘oem : ten fo bw io io doy fe i a 2 ’ Figate 11 The sation between cstre ld stress and pollution layer thickness, Figure 12 The relation betwoon electric fc stress and pollution layer electrical conductivity 4.3. Infuence of pollution layer thickness on electric field stress ‘To investigate the effect of pollution layer thickness on the clectie field siress, model 1 was used with different pollution, thickness of 0.5mm, 1.Smm and 3mm. The conductivity of pollution is held constant. Fig. 11 shows eleettc field stress Of the first model for different pollution layer thicknesses. AS seen, when the pollution layer thickness increased the electric, field stress is increased, 44, Influence of pollution layer conductivity on electric field ‘The pollution severity has been classified into four categories, according to its corresponding pollution layer conductivity and equivalent salt deposit density (ESDD) levels as light, ‘medium, heavy and very heavy pollution. The first model is, used {0 study the effect of electrical conductivity of pollution layer on the electric field stress, as shown in Fig, 12, and theElectric 1d distribution of optimized composite insulator profiles under different pollution conditions ass ‘Table 3 Maximum clectricfckl stress along the insulator surface during clean and polluted condition. Model no. nus (Vim) for original nus (RV) for optimized insulator models Gean Clean Polluted Model t v1 19 1 Model 2 207 Is us Model 3 155, 9. 133 Model 4 1 Ls. 12 electric field stress increased with the increase in pollution Acknowledgments lager conductivity, Thre valus of conduraivily ae wed 0.5 (um), 1 (Sim) and 5 (xS/m) to simulate different polation “The results of simulation modes for clean and uniformly polluted conditions applied vo both typical and optimized insu fetor dimensions are lated in Table. I can be noice” the following «The maximum electric field stress for optimized models dur= ing dry and clean conditions is less than the corresponding value for typical models. ‘The maximum electric field stress for optimized models dur- ing dry and polluted conditions is less than the correspond- ing value for typical models The maximum electric field stress of polluted conditions is, Jess than that of clean conditions for both typical and opti- ‘mized models. However, this distribution is very dynamic, and is expected to change rapidly as soon as leakage current starts to flow through the pollution layer. This result has been reviewed and confirmed with previous researchers [6,10 5. Conclusion In this paper. electric field distributions on different profiles of ‘composite insulators under clean and various contamination conditions were investizated by using software package of Comsol Multiphysics based on Finite Element Method (FEM). Comparison between the electric field of typical and optimized insulators during clean and polluted conditions was accomplished. In the dry and uniform polluted surface ‘conditions the value of masimum electtc field stress for opti ‘mized models was les than the corresponding value for typical models. In case of clean surface with water droplets the value ff maximum electric field stress for optimized model was less than the corresponding value for typical models. The percent= age of reduction was about 23%, In case of polluted surface with water droplets the value of maximum electric field stress for optimized model was less than the corresponding value for typical models. ‘The effects of pollution layer thickness and conductivity on the electric field have been investigated and when the pollution layer thickness increased the electric field stress increased, and alo when the electrical conductivity of pollution layer increased the electric field stress increased. Its concluded from the above results that the proposed optimization technique presents good results during all tested conditions and this, shows clearly its effectiveness Many thanks and appreciations are due to Dr. Mohamed Abdellah and UMR Industries staff (factory of composite insulators in Industrial zone Badr City, Cairo, Egypt) for the supporting and providing the details of practical composite insulators data. References In] Hackam R. Outdoor HY composite polymeric insulators. IEEE. “Trans Dielct Elect Hse 19996557 8, DP] Philips A, KuflelJ, Baker A, Buraham J, Carrera A, Cheney E, ta. let ids on AC composite transmission line insulators, IEEE Trans Power Delivery 2008:25:823 30, [5] Murugan N, Sharmila G, Kaanayeram G. Design optimization of high voltage composite insulator using electric fed computations. In: 2013 International conference on circuits, power and compul- ing technologies (ICCPCT). p. 315 20. [a] ELRefaie EM, Abd Elrahman MK, Mohamed MKh. Composite inaulators profile optimization using particle swarm algorithm and fniteclement method n: The 19thinternational symposium on high voltage engineering (ISH), Pilsen, Czach Republic, August 2015 [5] Nekahi A, Memeckin S, Farzanch M. Eifet of pollution severity ‘on electric fel distribution along a polymere insulator. In: 2015 TEE. 11th international conference on the properties and applications of dclctie materials (CPADM), [6] Marungst B, Onchantuek W, Oonsivilai A, Kulworawanichpong, TT Analysis of eletic field’ and poventil distbutions along. surface of silicone rubber insulators under various contamination conditions using finite element method, World Aead Sci, Eng “Technol 2009:533):1359 63 [7] Challagondla NK, Schramm HH. Electric field behavior of water drop on composite insulator surface under DC stress. In: 2015 IEEE 1th international conference on the properties and applications of dielecrie materials ICPADM), p. 3368. IB] Souza AL, Lopes 1. Experimental investigation of corona onset in contaminated polymer surfaces. IEEE Trans Dielect Electr Tsu 2015:220)1321 31 1B] Gao H, Jia Z, Mao Y, Guan Z, Wang L. Elect of hydrophobicity ‘on electric field distribution and discharges along various wetted hydrophobic surfaces. IEEE Trans Dielet Bletr Insul 2008:15 Oy85-43 [10] Abd-Rahman R, Haddad A, Haid N, Griffiths H. Sizes control ‘on polymeric outdoor insulators using zine oxide microvaristor ‘componies, IEEE Trans Dielct Electr Intl 2012:192)705 13, [11] COMSOL Multiphysics, "User's Manual", Version 5 [12] Muniraj C, Chandraschar S. Finite clement modoling for etic field and voltage distribution along the polluted polymeric insulator. World J Model Simul, England, UK 2012:84310.20. [U3] TECITS. 60815 3, 2008. Selection and. dimensioning oF igh ‘voltage insulators intended For use in polluted conditions Part 3 polymer insulators for a.1356 ESM. ELRefaie etal EbSayed M. M. BlRefuie received the Be, degree from Mansoura University, in 1971 Hee obtained the MSe. deuree from Helwan University and. PhiD. degsee from Slovak Technical University, Bratislava, in 1977 and 1984 respectively, all in clectrical power en- neering. At present he is professor with the electial power and machines enginering department, faculty of engineering, Helwan University, Cairo, He has many publications jin different local and international technical periodicals and. conferences. His fick! of imterest includes high voliage engineering, power systems, electrical safety and environment engineering. MLK. Abd Erahman was born in Giza, Egy, jn 1963, He received the Se. and MSc, degrees in elecrieal engineering from Helwan University, Caio, Egypt, in 1986 and 1992, respectively, and the Ph.D. degree in class= cation of paniallscharge pattems using {extre analysis algorithms from the Indian Institue of Technology, Kanpur, Tai, in 2000. From 2004 0 2009; he wast Lactuerin (Omar AL Mukhtar University, ALBayds, Libya, Currently, he is associate profesor in the Eececal Poser and Machine Engineering Department, Faculty of Engineering, Helvan University IMAKh. Mohamed was bora in Giza, Bey, in 1980, He received the B.S. and MSe, degrees in electrical engineering from Helwan University, Cairo, Exyp, in 2002 and 2009, respectively, snd he i now a PhD. stent in lectical engineering at Helwan University Caio, Egypt. He has worked as teaching asistnt since 2005 at the Elstrical Power and Machines Engineering department, Fac ulty of Engincoring, Helwan University Cairo, Egypt. His fk! of interest incindes high voltage engineering, insulation systems, carting systems, power systems, electromagnetic engingstng