36-WG11/Gdansk/198 Draft January 23, 2005 Selection of Insulators for AC Overhead Lines in North America With Respect to Contamination Principal authors: A. Baker, M. Farzaneh, T. Grisham, G.G Karady Working Group on Application and Inspection of Insulators* IEEE Lightning & Insulator Subcommittee

Abstract First published 25 years ago [1,2], this selection guide has been updated to modify the contaminated insulator flashover model to include composite insulators and developments in site severity assessment techniques. Standardization of artificial contamination test methods is also discussed. Finally leakage distance recommendations for the selection insulators with respect to contamination are given. Key Words: Contamination, Flashover, Insulators. 1. Introduction The power frequency voltage performance of outdoor electrical equipment can be de-graded by any of a wide variety of atmospheric particles deposited on the exposed insulation surfaces. The problem becomes more severe at high system voltages, and can indeed be the determining factor for properly selecting insulators at 230 kV and higher. However, even at distribution level voltages, in especially severe cases, contamination caused outages occur. Around the world there are a great variety of sources of the befouling materials; some manmade (e.g. industrial areas), and some natural (e.g. seacoasts, desserts, etc.), and utility line and station designers can sometimes have a very difficult task in choosing the proper insulation to obtain satisfactory service. For North America, “spot contamination” and “area contamination” are convenient classifications [1]. The former is characteristic for most utilities experiencing difficulty where only one substation, or a short section of a line, is subjected to contamination levels threatening service continuity. The latter is characteristic for utilities located along some warm area coastlines, or in large industrial areas. In other areas of the world, area contamination could apply to large desert regions. The application of ceramic insulators in a contaminated environment was addressed in an earlier Working Group paper [2] but needs to be re-visited, and the guidelines revised to reflect current knowledge. Non-ceramic (also referred to as polymer or composite) insulators have now been available for many years and account for a significant percentage of new insulators being installed [3,4]. These insulators may significantly differ from ceramic ones in their ability to resist contamination induced flashovers.

*G. G. Karady, WG Chairman, A. Baker, R. Bernstorf, G. Burnham, E. Cherney, W. Chisholm, R. Christman, E. del Bello, J-F Drapeau, M. Farzaneh, R. Gorur, T. Grisham, S. Grzybowski, S. Gubanski, R. Hartings, R. J. Hill , J. Kuffel, S. Mara, T. Rozek, H. Schneider, A. Schwalm, D. Shaffner, G. Stewart, R. Sundarrarajan, George G Karady, Arizona State University, [email protected]

Ceramic insulator surfaces are easily wetted, and contamination induced flashovers can be described in terms of this characteristic. An easily wetted material is said to have a high surface free energy. Some non-ceramic insulators are not easily wetted and a description of the contamination induced flashover process must be expanded to include the effect different surface wettabilities on the flashover mechanism. In addition, unlike ceramic insulators, some non-ceramic materials are not inert compared to ceramic insulators with regard to surface layers of pollution and this characteristic must be accounted for a complete description of the contamination flashover process. This update of the earlier paper includes modification and updates of: • Insulator surface characteristics • Contamination flashover process explanation. • Site contamination severity evaluation. • Laboratory contamination test standardization. • Insulator selection for contamination. • Maintenance practices. 2. Contamination Flashover Process Contamination induced flashovers occur when insulators become sufficiently coated with a wet conducting film containing dissolved salts. Rain water is not usually conductive enough by itself to cause a problem, nor is a dry salt deposit a problem. Dissolved salts may arrive simultaneously with the water, as in the case of sea spray, or the salts may build up slowly over many weeks or months with flashover occurring when the pollution film is sufficiently wetted and conductive to allow surface leakage current to flow. The insulation material can have a significant affect on this process. Every material has an associated surface free energy. If the insulator material has a higher surface energy than water, any moisture deposited on its surface will tend to uniformly spread out and dissolve the pollution layer salts, providing a low surface resistance and allowing a leakage current to flow across the insulator surface. If on the other hand, the insulator material has a surface free energy lower than water, any moisture deposited on the surface tends to bead up, inhibiting the formation of leakage current by maintaining a high surface resistance. A low surface free energy material is said to be hydrophobic. Examples of materials that exhibit low surface resistance when contaminated and wet (hydrophilic) are ceramics (porcelain or glass) and some non-ceramic such as aged ethylene-propylene diene terpolymer rubber (EPDM). It is important to differentiate between a material that exhibits hydrophobicity only when new, and one that maintains this characteristic over a long service life. Silicone rubber (SR) is a hydrophobic material and service experience has demonstrated a properly formulated silicone rubber retains this characteristic in service over many years [5]. Some materials, e.g. EPDM, can exhibit hydrophobicity initially because of

1

remnant processing oils on the rubber surface, but it is quickly lost in service as result of aging and/or contamination, and does not recover. The degree of hydrophobicity a material exhibits is determined by measuring the contact angle of a water drop at the water-air-surface triple point. This is a difficult and time consuming procedure, not well adapted for field use. In addition an evaluation of the hydrophobicity of the overall insulator surface is of real interest. A convenient visual method of rating materials with respect to overall surface hydrophobicity into seven different classes has been developed by the Swedish Transmission Research Institute (STRI) and is shown in Figure 1.

Light wetting conditions such as fog, dew, or drizzle are primarily responsible for contamination flashovers. Heavy rain is usually beneficial, as it washes away the surface deposits. If the insulator surface is polluted and easily wetted (i.e. a hydrophilic material), the pollution layer becomes a conducting film and allows a leakage current to flow. As the resulting heat is dissipated, some of the moisture is evaporated and dry bands form. Dry bands will form first on those sections of the insulator with the smallest diameters and the highest leakage current densities. Because these dry bands have much greater resistance than the wet surface, the voltage concentrates across them, occasionally leading to small, intermittent scintillating electric discharges bridging the dry band as illustrated in Figure 2. These scintillations are visible and range in current from a few milliamps to about one ampere (peak). As the surface is further wetted, the discharge current increases until at some point the discharges elongate, join together to bridge the entire insulator and trigger a power arc. Because the flashover discharges grow along the insulator surface, the hot power arc which follows their path may damage the insulator.

Insulator top surface

Wet zone (low resistance)

Dry zone (high resistance)

Figure 2. Illustration of dry band formation and arcing on a porcelain insulator [8]. courtesy of STRI Class HC 1………………. HC 2……………… HC 3……………… HC 4……………… HC 5……………… HC 6……………… HC 7 (not shown)…

Description Discrete drops, circular shape Discrete drops, less circular Discrete drops, not circular Completely wetted areas < 2 cm2 Some wetted areas > 2 cm2 Wet areas covering > 90% of total area. Continuous water film covering total area

Figure 1. STRI Hydrophobicity Classification Guide [6]. Silicone rubber typically exhibits an overall surface hydrophbicity consistent with HC1 through HC3, while ceramics and aged EPDM are normally HC7 [7].

For a hydrophobic insulator, the wetting process is much more complicated as surface water tends to coagulate into separate isolated areas, and formation of a continuous leakage current path is inhibited. In addition if the material is silicone rubber, low molecular weight chains of the polymer material are mobile enough to diffuse from the material surface into the pollution layer and the surface layer itself becomes hydrophobic. The resistance of the layer depends on the water that migrates into the underlying contaminant layer, and the solubility of any salts that may not be protected by the low molecular weight polymer. The combined effects of ohmic heating in the layer and the applied electric field on the water drops produce high resistance filaments on the insulator surface. Spot discharges emanating from the filaments causes localized loss of hydrophobicity, increased filament lengths and, if the resulting filament electric field strength exceeds that of the arcs, flashover can result as illustrated in Figure 3.

2

Contamination collection. A major part of the contamination layer, for both conductive and inert components, is transported to the surface of the insulator by air movement. The significance of wind direction on the contamination pattern is easily seen in most locations by observing the underside of insulators. Accumulation of contaminant on an insulator is usually greater on surfaces where air turbulence exists. This effect is particularly noticeable in locations where wind is fairly constant in direction. The electrostatic field has a secondary effect on the collection of wind-borne particles in regions of high AC stress such as the pin area on porcelain suspension insulators. This effect is much more pronounced for high voltage DC insulation where the polarity remains fixed with time. Particles are held through a process of dielectric polarization once they have contacted the surface of the insulator. The heating effect of leakage currents in areas of high stress hinders natural washing and increases the rate of contaminant accumulation. Figure 3. Illustration of filament formation and spot discharge on a silicone rubber insulator. [9]. 3. Site Severity Evaluation An assessment of the actual contamination level in a particular area must be made to properly select insulators. The rate at which a contamination layer builds-up depends on local contaminant characteristics, the effect of insulator material and configuration on contaminant collection, and local wetting conditions. Contaminants. Atmospherically deposited contamination material can be considered as having of two components – one conductive (when dissolved in water) and the other inert. The most common conductive component consists of ionic salts such as sodium chloride, magnesium chloride, sodium sulfates, etc. These salts, when in solution, can influence the flashover voltage by providing either a continuous, or partially continuous, conductive path on the surface of the insulator depending on the insulator material type. For the application of hydrophilic insulators, the severity of contamination due to these salts is generally quantified in terms of the equivalent salt deposit density (ESDD) of the pollution layer stated in units of mg/cm2 of NaCl. ESDD is the equivalent amount of NaCl that would yield the same conductivity in complete solution as that of the wetted layer. ESDD considers only the steady state solubility of the contaminant. Another important consideration is the rate at which the salt goes into solution. [10]. The inert component of the contamination is that portion of the solid material which does not go into solution as ions. This inert material may, however, decrease the electrical withstand strength of the insulator. Non-conducting materials such as silicone dioxide, clays, cement, etc. may form a mechanical matrix in which particles of a conductive component are embedded. The protection of this matrix generally decreases the effect of natural or mechanical washing of the insulator. The density of the inert material is quantified in terms of the non-soluble material deposit density (NSDD) also with units of mg/cm2. Inert materials may be hydrophilic and increase the wetting rate of the insulator, or hydrophobic and decrease its effective wetting. Typical hydrophilic inert contaminants include clay, silicone dioxide, and inorganic cements. Hydrophobic materials, for example grease and oil, may cause water beading and hence a discontinuous conducting layer.

Ceramic insulators have a hard glassy surface and typically remain unchanged over many years. Composite insulator surfaces do change over time. EPDM surfaces tend to increase in surface roughness and become chalky due to ablation surface polymer and exposure of the aluminum tri-hydrate (ATH) filler over time. Silicone rubber contains a percentage of low molecular weight polymer chains on the surface that can diffuse through the contaminant layer causing the layer itself to become hydrophobic. Therefore a contaminated and discolored silicone rubber insulator is not usually a cause for concern. Insulator wetting. The other factor determining site severity is moisture deposition on the insulator surface. This occurs mainly by three mechanisms – condensation, adsorption, and impingement. Condensation on a surface occurs when its temperature falls below the dew-point temperature. Ceramic and non-ceramic insulators differ in their thermal properties. For two different materials of comparable wettabilty (same hydrophobicity classification), at the same temperature, the one with lower specific heat will become wet first as a result of condensation. Therefore porcelain with a specific heat of 0.22 cal/g/ ºC would become wet from condensation before silicone rubber or EPDM with specific heats of 0.35 and 0.55 cal/g/ ºC respectively. However silicone rubber, due to its hydrophobicity, would tend to bead the water rather than wet-out like EPDM [11]. On a clear still night an insulator with a high surface emissivity, such as porcelain, loses heat through radiation to the night sky faster than heat can be supplied to it by air currents. If the insulator surface temperature drops below the dew-point, moisture forms on the surface of the insulator. Dew condensation wetting is a major cause of flashover on in-service insulators. Salts are hygroscopic and adsorb moisture at a relative humidity higher than about 80%. Consequently salt coated insulators can start to become wet even before dew forms. Water impingement on insulator surfaces occurs during fog and rain conditions. Depending on the wind, water impingement can wet the underside of an insulator more effectively than condensation. There is a critical amount of water required on the insulator surface to produce the minimum flashover voltage. The most severe condition requires a sufficient quantity of moisture and a time duration of wetting

3

sufficient to dissolve the majority of the conductive contaminate without dripping. Larger amounts of water can wash contaminants off allowing the insulation to recover. Severity measurement. The establishment of site severity can be accomplished in two ways: • Experience with existing lines. • Measurement of a parameter representative of the pollution level which can be correlated with expected insulator performance. Expected performance of insulators can be estimated from historical data of flashovers from existing lines and stations in the given vicinity, including test racks; however, historical data must span several years to be representative. One disadvantage of using existing lines for future designs is that if the reference lines never flash over, it is unknown if, or by how much, the insulation is over-designed. On the other hand if current line insulation must periodically be maintained, for example by washing, their insulation strength in that application may be better known. Site severity measuring methods involve measuring a significant parameter over a sufficient time period to provide a realistic estimation of the local contamination levels. Several methods have been used around the world for many years [12]. One method is to determine the pollution layer ESDD for an insulator either removed from an existing line or a “test” insulator exposed for a suitable time in the area. If the material surface characteristic of an insulator makes it difficult to remove a pollution layer to obtain a valid ESDD level, another method, such as leakage current measurements, for measuring site severity should be considered. [13]. The measured parameter must be correlated to a representative and repeatable test method that prospective insulators may be subjected to for a particular application. As explained later, the Clean Fog Test represents service conditions in which the dry contaminant is subsequently wetted. For this test insulators are uniformly contaminated by immersion into, or spraying on, a slurry mixture of water, salt, and inert material. The level of contamination is related to the ESDD of the dried artificial contamination layer. If the natural polluted insulator surface is such that the ESDD level can be accurately determined over a sufficiently long period of time, by removal and measurement, and be suitably modeled by the artificial contamination test, ESDD measurement on field exposed test insulators is a good way to evaluate site severity. Hydrophilic insulators such as porcelain are candidates. Surface pollution collection characteristics for non-ceramic insulators can be very different from those for porcelain. The surface wettability of some non-ceramics, such as EPDM, can change with age. Silicone rubber insulators, in addition to the long-term retention of surface hydrophobicity, also contains some low molecular weight polymer chains that can interact with the pollution layer to make it hydrophobic as well. In such cases removing the pollution layer to obtain a measure of its ESDD is problematic and some other measure of site severity is needed. Measurement of the surface leakage current for a test insulator exposed at the site of interest may be an alternative way of estimating site severity with respect to contamination level. A recently

completed series of round robin tests at several laboratories demonstrated that, though more work is needed, a standard for surface resistance measurements on non-ceramic insulators was feasible [14]. It was concluded that this method is a more sensitive indicator of the true contamination severity for non-ceramic insulators than ESDD. Critical aspects for conducting leakage current measurements, to insure repeatable and reproducible results, are the type and rate of surface wetting and test voltage stress. A voltage high enough to obtain a measurable current, but low enough not to cause drying of the contamination layer is required. For new insulators it was found that a low wetting rate gave the lowest surface resistance values. While only a limited number of tests with the surface artificially contaminated at one ESDD level were performed in the tests, and more work is needed the result is of interest and is given in Table 1. Recommended guidelines for conducting leakage current measurements are given in reference 15. Table 1 Lowest Surface Resistance (in MΩ) for Fog Wetting Voltage Stress = 10 v/mm EPDM Silicone Rubber Surface Condition New & uncontaminated 30-80 4000-8700 0.1 1 -1.7 New & contaminated ESDD=0.1 mg/cm2

Porc. 6-20 ------

4. Artificial Contamination Test Methods Laboratory tests must be representative of conditions occurring in nature as much as possible to be useful in deciding the suitability of an insulator for a given application. Over time many types of artificial contamination test methods have been studied and used, however, most were determined to not be sufficiently representative of nature. Two methods now widely used and standardized are the Salt Fog Test and the Clean Fog Test. The Salt Fog Test is applicable to coastal areas with little, if any, insoluble material. The Clean Fog Test represents areas where contaminants accumulate on the insulator surface and become wetted by natural moisture. The latter is the considered to be most representative for conditions in non-coastal areas, and coastal areas with non-soluble materials. Clean fog tests involve applying a contaminant layer on a test insulator using a slurry consisting of water, an inert material, and an appropriate amount of sodium chloride to achieve a required salt deposit density or layer conductivity. The insulator is energized after the layer has dried and before wetting is applied. The type and rate of wetting, as well as many other parameters, can have a profound effect on results obtained for artificial contamination tests and requirements for performing such tests have been standardized by a series of round robin tests [15] and are given in the latest edition of the IEEE Standard for Techniques for High-Voltage Testing [16]. Obtaining a uniform contamination layer on a hydrophobic insulator such as silicone rubber requires special techniques not yet covered by the standard. The inert material typically used in North American laboratories is Kaolin and specifications for this material are given in the standard [16]. The standard slurry contains 40 g of inert material per liter of water with an appropriate amount (5-40 g/l) of NaCl to achieve the target ESDD. Other inert materials such as Tonoko may be used as an

4

alternative to using Kaolin, however as noted in the standard doing so may give different test results and this must taken into account when comparing results. Comparing clean fog test results using Kaolin or Tonoko it was found that to obtain the same non-soluble deposit density (NSDD) on a porcelain insulator surface, a Kaolin slurry with 1 ½ times the density of a Tonoko slurry was required. Even with this correction, the flashover of the Kaolin contaminated insulators were 20-25% lower than with Tonoko contaminated insulators in the case of DC and about 10% in the case of AC [17]. Critical (50% probability) flashover voltage (CFO) test results for different contamination levels for an IEEE Disk porcelain insulator (see Fig 5) are given in Figure 4. These results were taken from the earlier paper [2] with additional points taken from the references noted. For these tests 40 g/l of Koalin meeting the requirements given in IEEE Std 4 was used as the inert material. 5. Design practice A. Ceramic Insulators The factors for selecting an insulator for application in a contaminated area are: • Ability of the insulator material to withstand the electrical stresses imposed. • Manufacturing limitations on the size, shape, and forming characteristics of the material. • Use of increased leakage distance. • Spacing, shape, diameter ratio, etc. of insulator sheds.

Suspension Type

Dimensions*

IEEE Disk

10” x 5 ¾” x 12”

ANSI C29 2 Class 52-3

10” x 5 ¾” x 11 ½”

Typical Fog Type

10” x 5 ¾” x 17”

Obsolete Fog Type

10” x 6 ½ ” x 17”

* Shed Diameter x Unit Spacing x Leakage Distance Figure 5. Porcelain suspension insulator units.

2 [2]

1.8

Using techniques described in section 3 the pollution severity for an area can co-related to an ESDD level and, Figure 4 can provide guidance in selecting an appropriate specific leakage distance for porcelain insulators for that application.

1.6 Flashover Voltage (50%) (kV/in leakage)

requirements for an ANSI C29.2 Class 52-3 suspension [ 20]. Typical high leakage distance suspensions, sometimes called a “Fog type” suspension currently available, along with an obsolete configuration are also shown.

IEEE Standard Disk

1.4

Porcelain Suspension Insulator

1.2

[17]

Assuming a standard deviation of 10%, a factor of 0.7 can be used to convert the average critical flashover stresses in Figure 4 to withstand values at a -3σ level. Based on results obtained in various laboratories for the Clean Fog Test, this is a conservative assumption [15].

[15]

1

[2] [18]

0.8 0.6

The curve approximating the data points in Figure 4, corrected to withstand voltage stress vs. contamination level, is given by

0.4 0.2 0 0.01

0.1

1 2

Salt Deposit Density ( mg/cm )

Figure 4. Power Frequency Critical Flashover Voltage (average stress) per unit leakage distance as a function of contamination (ESDD) level for porcelain IEEE Disks. Over the long period of time that porcelain insulators have been available, many designs have evolved, especially for suspension type insulators, in an attempt to obtain high leakage distances. Some of these high leakage distance suspensions are no longer available due to their high manufacturing cost. Several different configurations of porcelain suspensions are shown in Fig 5. The IEEE Disk was developed by the Working Group several years ago to be representative of several manufacturer’s units meeting the

WS = 0.32 (SDD) -0.36 kVl –g / inch of leakage distance In the earlier work ESDD levels for different site severities were given as shown in Table 2, to which has been added stress withstands expressed in kV/ inch of leakage distance for vertical strings of IEEE Disk porcelain suspensions for light and moderate contamination areas. Table 2 HVAC Specific Leakage Distance for site severity based on ESDD estimation Stress Limit Recommended ESDD Site for Withstand Leakage Distance (mg/cm2) (kV/in.) (in./kV) Severity Very light < 0.03 Light 0.03 – 0.06 1.13 – 0.88 0.88 – 1.13 Moderate 0.06 – 0.10 0.88 –0.73 1.13 – 1.37 Heavy > 0.10 -

5

More recent ESDD measurements of the pollution layers accumulated on station posts in several HVDC converter stations showed that an ESDD level of 0.05 mg/cm2 was representative of the most heavily contaminated station, located in a high density urban area, and a level of 0.005 mg/cm2 was representative of rural areas [21]. Though higher specific leakage distances are required for HVDC as compared to HVAC insulators, these results suggest that Table 2 provides is a reasonable baseline site severity classification for most applications. Common insulation ranges in use for transmission class overhead lines in North America are shown in Table 3 along with the calculated leakage distance per unit voltage. Comparing these values to Table 2 indicates most lines are located in areas that would be classified as light contamination. Table 3 Common Insulation Levels 138 kV – 765 kV Transmission Lines Insulation Range Typical Number of System 10” x 5 ¾” x 11 ½” Leakage Distance Voltage ANSI C29.2 per unit Voltage Class 52-3 Suspensions (kV) (in. / kVL-g ) 138 7–9 1.01 – 1.30 230 10 – 12 0.86 – 1.04 345 15 – 18 0.87 – 1.04 500 24 – 27 0.96 – 1.08 765 30 -35 0.78 – 0.91

Silicone rubber composite insulators, with their long-term hydrophobicity retention characteristic, perform significantly better in contamination flashover tests than hydrophilic insulators such as porcelain or EPDM. Clean Fog test results for silicone rubber insulators, taken from the references noted, along with the curve for porcelain insulators from Figure 4 are given in Figure 7. Though the improvement in contamination performance with silicone rubber insulation is dramatic compared to porcelain, and provides an option for applications in areas with heavy to very heavy contamination levels, the IEEE Application Guides for non-ceramic insulators recommends using leakage distances equivalent to those recommended for porcelain given in Table 2 [23].

If the ANSI Class 52-3 suspensions are replaced with “Fog Type” suspensions as shown in Figure 5, the number of units given in Table 3 would be suitable for moderate to heavy contamination areas. Also replacing the vertical strings with vee-strings would improve the contamination performance by 25-30% [22]. Flashover Voltage (50%) (kV/in leakgae)

B Non-ceramic Insulators A typical non-ceramic suspension insulator is shown in Figure 6. Single unit insulators of appropriate length and leakage distance are available for all voltages through EHV levels. Individual shed diameters may be uniform or staggered. Shed spacing is an important consideration for good performance. Care must be taken to avoid shed-to-shed arcing, in effect shorting out some portion of the leakage distance under severe contamination or heavy wetting conditions.

3.00

2.50

[19] [26]

[26]

2.00

[19]

1.50

1.00

0.50

0.00 0.01

0.1

1

Salt Deposit Density (mg/cm 2)

Figure 7. Clean Fog Test results for Silicone Rubber Insulators compared to those for porcelain suspensions. C. Distribution Voltage Insulators Specific leakage distances for distribution class deadend and line post insulators (≤ 69 kV) from the ANSI C29 series of insulator standards are shown in Figure 8.

Fig. 6 Typical Non-ceramic Suspension Insulator

Except for severe contamination areas, distribution class insulators are typically selected based on considerations other than contamination performance. Cost is usually a primary concern for these products.

6

Leakage Distance/ kV

1.8 1.6

Suitable

C29.13 Composite Deadends

1.4

C29.7 Porcelain Line Posts

1.2

C29.18 Composite Line Posts

1 0.8 66 151

0.6 0.4 0.2 0

System Voltage 15 KV

25 KV

35 KV

46 KV

69 KV

Figure 8. Leakage distance per unit line-to-ground voltage for ANSI C29 series of insulator standards [24]

A specific leakage of about 1 inch / kVl-g is suitable for areas classified as light to moderate contamination. For porcelain distribution class insulators common practice has been to use insulators 1 or 2 voltage classes higher than normal for heavy contamination areas such as along seacoasts. In such applications silicone rubber insulators of the proper voltage class could be used at the leakage specified in the ANSI standards. 6. Extreme Contamination. Though most areas of the North America can be classified as light-tomoderate contamination areas, some heavy-to-very heavy contamination areas may be encountered. Also the contamination characteristic of an area may change over time. Line performance may deteriorate from satisfactory to unsatisfactory if local contamination conditions increase, for example due to increased manufacturing or agricultural activity. Unusual environmental conditions such as forest fires, droughts, etc may also be a temporary source of contamination. Performance of existing lines may be improved by periodically cleaning the insulators. Methods and procedures for cleaning both energized and de-energized contaminated insulators are described in IEEE STD 957, “IEEE Guide for Cleaning Insulators” [25]. As cautioned in the guide, insulator cleaning must be done so that the insulators are not damaged, or that the cleaning activity itself does not cause a service interruption, such as can occur with improper water washing. Damage to insulators can occur during compressed air / dry abrasive cleaning if impingement of the abrasive material is allowed to erode away the glaze surface of ceramic insulators, or roughen the surface, or tear the sheds of non-ceramic insulators. Non-ceramic insulators can have a variety of construction options and characteristics, and it is best to consult the specific manufacturer for advice on cleaning their insulators. In particular consideration should be given to whether the polymer weathersheds are bonded to the fiberglass rod or not. If not, and hot-line water washing is employed, only low pressure (below 300 psi) washing techniques should be used. The general recommendations for hot-line water washing insulators taken from the Guide are given in Table 4. Table 4 Hot-line Water Washing Recommendations [25]

Type High Pressure Medium Pressure Low Pressure

Pressure range

Ceramic

For Bonded Polymers

Washing Unbonded Polymers

400-1000 psi

OK

OK

NO

300-400 psi

OK

OK

NO

Below 300 psi

OK

OK

OK

Coating ceramic insulators with silicone grease has been used satisfactorily as a means of preventing contamination flashovers in extreme environments for almost fifty years. The grease coating serves to provide a hydrophobic layer and encapsulates contamination particles. They are not permanent and must be

replaced at intervals dependent on local conditions, usually from a few months to a very few years. Since the grease layer, which is typically 1/16” to 3/16” thick when first applied, encapsulates pollution material overtime, eventually channel arcing through the layer may occur and lead to tracking. As a result greasing of non-ceramic insulators is not advisable [23]. Room temperature vulcanized (RTV) liquid silicone polymer can also be applied to ceramic insulators to provide a water repellent surface and improve their contamination performance. Guidelines for the selection and application of RTV coatings are given in IEEE Std. 1523 [27]. 7. Conclusions & Recommendations. The inherent material surface free energy of an insulator has an effect of the flashover performance of an outdoor insulator with respect to contamination. Some non-ceramic insulators have a low surface energy, and the contamination induced flashover model presented earlier has been modified to account for the superior performance of such insulators. Site severity evaluation techniques must be broadened to accommodate the differences in pollution collection characteristics of non-ceramic insulators compared to porcelain or glass. Artificial contamination test methods have now been standardized. Previous test data published in the earlier paper are valid with respect to the standardized test. Corrections for results obtained by different laboratories using different inert materials for the contamination slurry are known. For transmission class insulators, guidance for the selection of a total leakage distance is given in Section 5. In extreme contamination areas it may be necessary to resort to periodic maintenance such as insulator washing to maintain satisfactory line performance. Comprehensive guidance on cleaning insulators is now available [23]. References: [1] IEEE Committee Report, “A Survey of the Problem of Insulator Contamination in the United States and Canada – II”, IEEE Transactions on Power Apparatus and Systems, Vol. PAS- 91, No. 5, September / October, 1972 .

7

[2] IEEE Working Group on Insulator Contamination, “Application of Insulators in a Contaminated Environment”, IEEE Transactions on Power Apparatus and Systems, Vol. PAS-98, Vo. 5, September/October, 1979. [3] H.M. Schneider, J.F. Hall, G. Karady, & J. Renowden, “Nonceramic Insulators for Transmission Lines”, IEEE Transactions on Power Delivery Vol. 4, No. 4, October, 1989.

[17] R. Matsuoka, K. Kondo, K. Naito, & M. Ishii, “Influence of Nonsoluable Contaminants on the Flashover Voltages of Artificially Contaminated Insulators” IEEE Trans. On Power Delivery, Vol 11, No. 1, January, 1996. [18] F.A.M. Rizk & M. Bourdages, “Influence of AC Source Parameters on Flashover of Polluted Insulators”, IEEE Transactions on Power Apparatus and Systems, Vol PAS-104, No. 4, April, 1985.

[4] T. Kikuchi, S. Nishimura, M. Nagao, K. Izumi, Y. Kubota, & M. Sakata, “Survey on the Use of Non-ceramic Composite Insulators”, IEEE Transactions on Dielectrics & Electrical Insulation, Vol. 6, No. 5, October, 1999.

[19] J.N. Edgar, J. Kuffel, & J.D. Mintz, “Assessment of the Contamination Performance of Transmission Class Composite Insulators Using the Clean Fog Test Procedure”, CEA Project 280T621, Canadian Electrical Association.

[5] T. Sorqvixt & A.E. Vlastos, “Outdoor Polymeric Insulators Long-Term Exposed to HVDC”, IEEE Transactions on Power Delivery, Vol. 12, No. 2, 1997.

[20] American National Standard for Insulators ANSI C29.2– Wet-Process Porcelain and Toughened Glass Suspension Type. American National Standards Institute, 1430 Broadway, NY, NY 10018. [21] H.M. Schneider, “Measurements of Contamination on Post Insulators in HVDC Converter Stations”, IEEE Transactions on Power Delivery, Vol. 3, No. 1, January, 1988.

[6] STRI Guide 92/1, “Hydrophobicity Classification Guide” Swedish Transmission Research Institute.

[22] Transmission Line Reference Book – 345 kV and Above. Electric Power Research Institute, Palo Alto, CA.

[7] T. Sorqvist & A.E. Vlastos, “Performance and Ageing of Polymeric Insulators”, IEEE Transactions on Power Delivery, Vol. 12, No. 4, October, 1997.

[23] IEEE Std 987 – 2001, IEEE Guide for Application of Composite Insulators., IEEE 3 park Ave. NY, NY 10016.

[8] M. Shah, G.G. Karady, & R.L. Brown, “Flashover Mechanisms of Silicone Rubber Insulators Used for Outdoor Insulation – II” IEEE Transactions on Power Delivery Vol. 10, No. 4, November, 1995. [9] G.G. Karady, “Flashover Mechanism of Non-ceramic Insulators”, IEEE Transactions on Dielectrics & Electrical Insulation, Vol. 6, No. 5, October, 1999. [10] K.C. Holte, J.H. Kim, T.C. Cheng, Y.B. Kim, & Y. Nitta, “Dependence of Flashover Voltage on the Chemical Composition of Multi-Component Insulator Surface Contaminants”, IEEE Transactions on Power Apparatus and Systems, Vol. PAS-95, No. 2, March/April, 1976. [11] A. De La O, R.S. Gorur, & J. Cheng, « AC Clean Fog Tests on Nonceramic Insulating Materials and a Comparison with Porcelain, IEEE Transactions on Power Delivery, Vol. 9, No. 4, October, 1994.

[24] American National Standards for Insulators: ANSI C29.13 “Composite Distribution Deadend Type‫ ״‬. ANSI C29.7 “Wet-Process Porcelain Line Post Type”. ANSI C29.18 “Composite Distribution Line Post Type”. American National Standards Institute, 1430 Broadway, NY, NY 10018. [25] IEEE Std 957-2003, IEEE Guide for Cleaning Insulators. IEEE, 3 Park Ave. NY, NY 10016.. [26] Sediver Application Guide for Composite Suspension Insulators - 1995. Ref. TG95005. [27] IEEE Std 1523 – 2002, IEEE Guide for the Application, Maintenance and Evaluation of Room Temperature (RTV) Silicone Rubber Coatings for Outdoor Ceramic Insulators. IEEE 3 Park Ave. NY, NY 10016..

[12] CIGRE TF 33.04.03, “Insulator Pollution Monitoring”, Electra No. 152, February 1994. [13] A.E. Vlastos & T. Orbeck, “Outdoor Leakage Current Monitoring of Silicone Composite Insulators in Coastal Service Conditions”, IEEE Transactions on Power Delivery, Vol. 11, No. 2., April 1996. [14] IEEE Working Group on Insulator Contamination, “Surface Resistance Measurements on Nonceramic Insulators’, IEEE Transactions on Power Delivery Vol. 16, No. 4, October 2001. [15] IEEE Working Group on Insulator Contamination, “Final Report on the Clean Fog Test for HVAC Insulators”, IEEE Transactions on Power Delivery, VOL PWRD-2, No. 4, October, 1987. [16] IEEE Std 4 – 1995, IEEE Standard Techniques for High-Voltage Testing. IEEE, 3 Park Ave., NY ,NY 10016.

8