R. Houlgate November 2002

36-WG11/Capetown/136

Orientation and Size 1.

Introduction Wetted pollution on the surface of any high-voltage insulation produces substantial reduction in its electric strength [1],[2]. However, the effect of orientation and size on the flashover performance of polluted insulation is not necessarily subject to simple rules. The insulator type directly affects the performance of the polluted insulation in different orientations. In addition, the pollution severity at a site, and the time taken for maximum contamination levels to deposit, may determine the affect of orientation. The nature of the subsequent wetting process and the flashover mechanism (e.g. surface flashover or intershed breakdown) are also important in understanding the influence of orientation and size. Hence, the flashover strength of different insulator types and orientations is a balance between the different processes that directly influence performance. The following mechanisms may contribute or be dominant for each design and orientation: (i) improved natural cleaning as the orientation changes from vertical to horizontal, (ii) directional effects of pollution deposition for angled/horizontal orientation from a localised (directiondefined) pollution source, (iii) inter-shed breakdown due to heavy rain and pollution, (iv) inter-shed breakdown due to pollution and poor profile, (v) reduced flashover strength due to pollution concentration to the lower surface on horizontal or near horizontal insulation during rain. As an example, the long-term contamination of cap-and-pin insulators in tension may be less than that on the same insulators in suspension due to better washing of the tension units due to rain. However, the tension insulators may also be subject to a direction effect if the major source of contamination is from a well-defined direction. In this case there can be an influence of orientation and direction in determining the natural polluted insulator performance for a particular location or type of location. For other locations where contamination levels can accumulate rapidly, or the frequency of natural cleaning by rainfall is very low, the above influence of orientation may be significantly altered for the same insulator type. In reality there is no real substitute for testing insulators under the appropriate pollution and wetting conditions to determine how actual insulator designs will perform in different orientations. However, information to provide guidance to insulation design engineers is clearly warranted. Unfortunately, there is a lack of published data, quantifying these effects. This paper presents and discusses the results of some investigations into the influence of orientation and size on the flashover strength of polluted insulation of various designs. Experimental results from artificial pollution tests and from outdoor marine testing stations for various insulators and orientations are analysed to investigate if some simplified conclusions can be drawn from the data.

This review only considers insulators under a.c. energisation since the d.c. performance of insulation in various orientations may involve different pollution/wetting conditions and distinct/separate flashover mechanisms, and is outside the scope of this document.

ERA TECHNOLOGY LTD CLEEVE ROAD LEATHERHEAD SURREY KT22 7SA UK TEL: +44 (0) 1372 367000 FAX: +44 (0) 1372 367099 E-MAIL: [email protected] www.era.co.uk Approved to BS EN ISO 9001: 1994. Reg No FM1303. Company registered in England No 4044609. Registered Office as above.

2 2.

Insulators The influence of orientation and size are analysed and discussed for the following types of insulation:- (i) cap-and-pin insulators, (ii) interrupter head porcelains with open profiles, (iii) tapered bushing porcelains with ALS profiles, and (iv) substation post insulators with ALS and multicone type profiles. More details of insulator designs, if available, are contained in the referenced publications. The hollow porcelains were sealed with end-flanges and pressurised with dry-nitrogen or SF6, to avoid internal surface discharges. In general, these insulators were tested without their internal grading components because such items complicate the test assembly, but do not affect the pollution flashover process, which is not an electrostatic-field problem. Salt fog test results on a complete SF6/Air bushing confirmed the validity of testing only the porcelains.

3.

Test Procedures 3.1 Artificial Pollution Tests 3.1.1 Salt Fog Test The main test procedure was in general agreement with that specified in IEC 507 [1]. However, the "Quick Flashover" method [4] was used to obtain flashover voltages at each salinity, rather than the withstand salinity prescribed in the standard. The advantage of this method, particularly when used to compare the relative performance of insulation is that the gradient of flashover voltage against salinity allows the safety factor to be estimated for the design insulation at each pollution level. At each value of fog salinity, the flashover voltage was determined by energising the insulator for a conditioning period at 90% of the estimated flashover voltage. The voltage was then raised in 5% steps every 1 minute until flashover. The insulator was re-energised at 90% of this flashover voltage, and the process was repeated in approximately 3% steps every 5 minutes until flashover. This latter process was repeated several times (typically 10), until a stable set of flashover voltages were obtained and the final 5 readings were used to determine the mean value and the range. 3.1.2 Heavy-Wetting Test This test was an adaptation of a procedure developed to determine the performance of polymeric sheds fitted as supplements to porcelain barrel insulators [5,6]. The test investigates if inter-shed breakdown due to pollution and heavy rain bridging sheds is responsible for reducing the flashover voltage of insulators with poor profiles. Failure of this test has correlated very well on insulators that have a poor service history during heavy rain or live line washing (e.g. Tapered CTs, inclined transformer bushings, inverted-V substation insulators).

3 The insulator prepared and preconditioned as for the salt fog test above, was energised at the specified test voltage and subjected to the specified fog salinity for 15 minutes. The pollution phase is then followed by a 15-minute drain period, during which dryband discharge activity would establish a stabilised salt fog deposit. The insulator surface was then wetted at approximately 45o with a rain rate of 2mm/minute and water conductivity of 1000uS/cm, to simulate torrential rain and wind, at the onset of a storm. The test is considered a withstand if no flashover occurs as the pollution is washed from the insulator surface and the leakage current subsequently declines. If flashover occurred, the procedure was repeated at a lower salinity, and conversely, following withstand, then the salinity was increased. The withstand pre-applied salinity (W.P.S.) was determined when 3 withstands out of 4 tests were obtained at the minimum fog salinity. Note: W.P.S. values are not equivalent to IEC 507 withstand salinities, since the salt fog deposit decreases during the drain period in the heavy wetting test, and the wetting process favours inter-shed breakdown. 3.2 Natural Pollution These investigations were largely conducted at the Brighton Insulator Testing Station (BITS) where flashover was the criterion of performance. The test procedure has been described elsewhere [7,8] and involves determining the relative lengths of insulation at flashover, for each insulator design tested. Explosive fuses were used to automatically lengthen the insulator string that experiences flashover. Overall results were presented as the frequency of flashover versus flashover stress, normalised to a reference 400kN cap-&-pin suspension insulator. Each insulator on test is assigned a figure of merit (FOM) which relates its performance to that of the reference insulator, and is calculated from the mean length of control insulators at the time of flashover, divided by the length of the insulator across which flashover had occurred. Figures of merit were thus dependant on the insulator shape, and the position and orientation at the test station. The FOMs were normalised by dividing each value by the mean FOM of the reference insulator.

4.

Results and Discussion 4.1 Artificial Pollution Test Results 4.1.1 Cap-and-pin insulators Salt fog tests on several types of cap-&-pin insulator strings in the vertical and inclined orientations had established that an improved performance, corresponding to 10-20% increase in effective length, resulted from inclination [9]. However, inclination in excess of 20o to the vertical, gave no further improvement. The axial flashover stress, E(kV/m), and the fog salinity, S(kg/m3) can be expressed in the usual power form [1] as: (1) E = E0.S-p where E0 and p are constants. For cap and pin insulators E0 values will typically range from 120 to 140kV(rms)/m of axial length between fixing points, and p values from 0.14 to 0.21. The value of p for an insulator can be considered as a weighted average of the value for an electrolyte surface (p=0.33) and that for the air breakdown (p=0), between parts of the insulator surface. Hence, surface discharge is just the dominant breakdown mechanism for cap-&-pin insulators and this is confirmed by WPS vales in excess of 80kg/m3 at a specific creepage of 25mm/kV in the heavy wetting test.

4 IEC publication 60815 [10] provides guidelines for choosing ceramic insulators for polluted environments. Four pollution severity levels are defined and the minimum specific creepage distance (MSCD), corresponding to each level is specified, as shown in Table 1. Also shown is the representative value of pollution severity, expressed by the equivalent salt (NaCl) deposit density (ESDD) on the insulator surface, the corresponding withstand salinity in the artificial salt fog test, or the surface conductivity. From equation (1), results from the salt fog test for cap and pin insulators were converted to specific creepage, for direct comparison with IEC 60815 guidance, via:

C = C0.Sp

(2)

and, C0 = (1000/E0√3).(lpl/l) where (lpl/l) is the ratio of leakage path to axial length (i.e. spacing), and C0 is a constant expressed in specific creepage for each insulator design.

(3)

A best fit of QFO data to expression (2) was then used to establish the MSCD for each pollution severity range shown in Table 1. The results for three designs of cap-and-pin insulator in suspension are shown in Table 2, and generally support IEC 60815. However, it should be noted that the specific creepage at flashover shown, were calculated for the minimum salinity quoted in each pollution level range. Hence, specific creepage at flashover could exceed the MSCD values quoted in IEC 60815 at higher salinities in each range.

4.1.2

Interrupter head porcelains

A paper on the pollution flashover performance of various types of interrupter heads has recently been published [11]. Results from that paper, which gives full details of the insulators, were processed as in 4.1.1 above, and are shown in Table 3.

Only the vertical interrupter V1 from Table 3 supports the current guidance within IEC 60815, and again only at the lower salinity values within each pollution level. The performance of the horizontal interrupters, unlike the case for cap-and -pin insulators, was much inferior to that of the vertical V1, with C0 values between about 13 and 16 for horizontal orientation, compared with a value of 11.7 for V1. The slope of specific creepage versus fog salinity relationship for horizontal interrupters (p = 0.23 to 0.27) was also appreciably greater than for V1 (p = 0.17). The two principal factors influencing this large difference in performance between horizontal and vertical interrupter heads were diameter and orientation. The mean diameter of the vertical interrupter was 322mm., compared with 383-498mm. for the horizontal interrupters. From general experience [1], this difference in size should not account for the increase in electrical strength of between 1.3 and 1.6 witnessed from horizontal to vertical orientation. Indeed, IEC 60815 does not apply a mean diameter correction in determining MSCD over the above mean diameter range. The orientation effect was hence reasoned to be associated with the way that the wetted pollution drains from the insulator surface, and a simple mathematical model partially explained the process [11]. The resulting concentrated pollution deposit on the lower surface of the horizontal interrupter was then responsible for increased surface discharge, resulting in lower flashover voltages and an increased gradient versus fog salinity.

5 Although, no heavy wetting test results have been reported for interrupter porcelains, from other results reported below, it was concluded that withstands would be expected for the horizontal interrupters. The performance of V1 cannot be predicted and measurement would be required to confirm acceptable resistance to inter-shed breakdown during heavy rain following pre-pollution.

4.1.3 Tapered Bushings Details of the tapered bushings cannot be given at the present time due to commercial reasons. However, the performance confirmed the orientation effect in the salt fog for post type insulators, as witnessed for interrupters. Thus, the pollution flashover performance of the SF6/Air horizontal tapered bushing was much inferior to that when vertical. QFO Tests at 0o revealed a very flat slope in the flashover voltage versus fog salinity relationship (p = 0.06), indicating significant air-breakdown. Results at 90o again demonstrated a significant increase in gradient (p = 0.19), with increased surface discharge activity along the bottom insulator surface. Withstand salinities were also measured at angles of 0o, 20o, 45o and 90o to the vertical, with values of 28, 28, 14 and 5 kg/m3, respectively. Heavy wetting tests at these same angles gave WPS values of 160kg/m3, 56kg/m3, 160kg/m3 and 160kg/m3, respectively. These heavy wetting test results confirm that this process does not significantly affect horizontal bushings, but can reduce the inter-shed breakdown capability of some bushings, at angles close to the vertical. Service experience with inclined transformer bushings would confirm this observation. The heavy wetting withstand of 160kg/m3 when vertical probably confirmed that the large component of air breakdown, indicated by the QFO Tests, were not from shed to shed. Results presented in Table 4 indicate that correction factors of from 1.2 to 1.3 were required to support IEC 60815 for the bushings tested in salt fog. Of note: although the SF6/Air bushing was complete with full internal components, it had a similar performance to the Air/Air bushing porcelain, which was gas-filled, but hollow.

4.1.4 Substation Post Insulators Results for post insulators are presented in Table 5 for an ALS and multicone design, and again confirm the orientation effect in the salt fog test. For P1, C0 changed from 13.1 to 21.5, from the vertical to horizontal orientations, although, uniquely, p remained unchanged at 0.11, indicating significant air breakdown in both orientations. The WPS values from the heavy wetting test were 160kg/m3 in both the vertical and horizontal orientations, probably confirming that the air breakdowns were not directly from shed to shed. A correction factor of 1.4 was required to support IEC 60815 for this post insulator in the salt fog. The multicone post (details from Don Swift) had the smallest orientation effect of the post type insulators tested, since although the gradient of the flashover voltage versus salinity relationship (p =0.10), was increased when horizontal (p = 0.17), conversely, C0 was reduced from 14.7 to 12.8. The results in Table 5 show that for the minimum salinity in each pollution severity level, guidance in IEC 60815 does not require correction. WPS values of 160kg/m3 were recorded in both vertical and horizontal orientations. From the limited data above, there is some evidence that ALS profiles may not represent the most efficient insulator shape for horizontal posts.

6 4.2 Performance under Natural Pollution Figures of merit from natural testing at Brighton Insulator Testing Station have been analysed in detail to determine the values of parameters affecting performance [8]. The factor for angled cap-and-pin insulators at 45o to the vertical was 1.17, and for tension insulators, at 75o to the vertical, was 1.29. However, a directional effect, from a maximum of 1.08 to a minimum of 0.91, had to be combined with the effect of angle at Brighton. The above factors are relative to the same insulator type in suspension. Results from Noto Testing Station, at a coastal location in Japan, also concluded that tension strings have almost the same or a little higher anti-pollution strength in comparison with suspension strings [12]. These tests were similar to those at Brighton in that insulator strings were systematically increased in length until approximately equal flashover frequencies were established for all test insulators. Limited data from the interrupter-head porcelain, H1, on test at Brighton, showed Figures of Merit of between 0.7 and 1.0, during a short test period. These results indicate that pollution concentration to the lower surface may be as important a process under natural conditions, as witnessed in the artificial salt fog tests. However, more data is required, to establish if better washing by rain, on average, negates the pollution concentration effect under natural conditions.

5.

Conclusions The improved performance of cap-and-pin insulators when inclined compared to vertical has been confirmed by natural and artificial pollution tests. Test results in all orientations for these insulators, generally supports the application of IEC 60815 in selecting MSCDs, although the minimum nature of the resulting values needs more emphasis. The artificial pollution performance of horizontal post type insulators is inferior to the same insulators when vertical. IEC 60815 cannot be successfully applied in the selection of these insulators, if inclined, unless a correction factor is applied. There is some evidence that ALS profiles may not represent the most efficient insulator shape for horizontal post and bushing type designs.

Correction factors for orientation have been identified for various insulator types. More natural pollution test data is required for inclined post type insulators to establish relative flashover performance compared to the same insulators, when vertical. 6.

References [1] P.J. Lambeth, "Effect of pollution on high-voltage insulators". Proc. IEE, Vol 118, No. 9R, pp. 1107-1130, 1971. [2] J.S.T. Looms, "Insulators for high voltages". Peregrin Ltd., 1988.

7

[3] [4] [5] [6]

[7]

[8]

[9]

[10] [11]

[12]

IEC Standard 507. "Artificial pollution tests on high-voltage insulators to be used on a.c. systems". Second Edition, 1991. P.J. Lambeth, "Variable voltage application for insulator pollution tests", IEEE Trans. on Power Delivery, Vol.3, No.4, October 1988. C.H. Ely, P.J. Lambeth, J.S.T. Looms and D.A. Swift, "Discharges over wet, polluted polymers: the "Booster Shed", CIGRE Paper 15-02, 1978. C.H.A. Ely, P.J. Lambeth and J.S.T.Looms, "The Booster Shed: prevention of flashover of polluted substation insulators in heavy wetting", IEEE Transactions on Power Apparatus and Systems, Vol. PAS-97, Nov/Dec, 1978. P.J. Lambeth, J.S.T. Looms, W.J. Roberts and B.J. Drinkwater, "Natural pollution testing of insulators for UHV transmission systems", CIGRE Paper 33-12, 1974. R.G. Houlgate, P.J. Lambeth and W.J. Roberts, "The performance of insulators at Extra and Ultra High Voltage in a coastal environment", CIGRE Paper 33-01, 1982. P.J. Lambeth, J.S.T. Looms, M. Sforzini, C.Malaguti, Y. Porcheron and P. Claverie, "International research on polluted insulators", CIGRE Paper 33-02, 1970. IEC 60815: 1986, "Guide for the selection of insulators in respect of polluted conditions". R.G. Houlgate and D.A. Swift, "AC Circuit Breakers: Pollution flashover performance of various types of interrupter head", IEE Proc. – Generation, Transmission & Distribution, Vol. 144, No. 1, January 1997. Y. Taniguchi, N. Arai and Y. Imano, "Natural contamination test of insulators at Noto Testing Station near Japan Sea", IEEE Transactions on Power Apparatus and Systems, Vol. PAS-98, No. 1.Jan/Feb 1979.

Table 1: Minimum creepage distance for pollution severity levels I

II

III

IV

LIGHT

MEDIUM

HEAVY

V. HEAVY

ESDD mg/cm2

0.03 - 0.06

0.1 - 0.2

0.3 – 0.6

> 0.8

Salt Fog kg/m3

5 to 14

14 - 40

40 - 112

> 160

Conductivity µS

15 - 20

24 - 35

36 - 40

> 80

MSCD mm/kV

16

20

25

31

Pollution Level

8

Table 2: Cap and Pin Insulators Specific creepage mm/kV at F/O

Insulator Type I

II

III

IV

Supports IEC 60815

Correctio n Factor

Ref A

15

17

20

24

Y

1

IEEE

14

17

21

28

Y

1

Std Disc

14

18

22

30

Y

1

Table 3: Circuit Breaker Interrupters Specific creepage mm/kV at F/O

Insulator Type I

II

III

IV

Supports IEC 60815

o

19

24

31

43

N

1.3

o

22

29

38

55

N

1.5

o

24

30

38

52

N

1.5

o

H4 (90 )

23

31

22

60

N

1.6

V1

15

18

22

28

Y

1

Correctio n Factor

H1 (90 ) H2 (90 ) H3 (90 )

Correctio n Factor

Table 4: Bushings Specific creepage mm/kV at F/O

Insulator Type SF6/Air V (0o)

I

II

III

IV

Supports IEC 60815

22

24

25

27

N/Y

1.2/1

SF6/Air (20o)

24

N

1.2

SF6/Air (45o)

24

N

1.2

SF6/Air (90o)

22

27

32

42

N

1.3

Air/Air (75o)

21

24

28

34

N

1.2

Correctio n Factor

Table 5: Post Insulators Specific creepage mm/kV at F/O

Insulator Type I

II

III

IV

Supports IEC 60815

P1 (V) (0o)

16

18

20

23

Y

1

P1 (H) (90o)

26

29

32

38

N

1.4

17

19

22

25

Y

1

Multicone (90 ) 17

20

24

31

Y

1

Multicone (0o) o