diagnostic of nonceramic insulators aged in a salt fog chamber by using electric field sensor

7/29/2019 Diagnostic of Nonceramic Insulators Aged in a Salt Fog Chamber by Using Electric Field Sensor

1/4

ConferenceRecordof the2004IEEE International SymposiumonElectrical Insulation,Indianapolis,IN USA, 19-22September 2004DIAGNOSTIC OF NON-CERAMIC INSULATORS AGED IN A SALT FOGCHAMBER BY USING ELECTRIC FIELD SENSOR

Isaias Ramirez-Vbzquez Ramiro Hemhdez-Corona Gerard0 Montoya-TenaInstituto de Investigaciones Eltctricas ( I E )AV. Reforma 113, Col. Palmira, CP 62490Cuemavaca, Morelos, M [email protected]

Abshocr: The objective of this paper is to validate the electricfield measurement technique for the diagnosis of uon-ceramicinsulators (NCI). To validate this technique, fifteen NCI wereaged in a salt fog chamber by applying uon-staudard methods.Before ageing theNCI , the electric field distribution aloug eachinsulator was measured (sound field curve). When an insulatorflashed over during the ageing test, the insulator was removedfrom the salt fog chamber and its electric field distribution wasagain measured (aged fieldcurve). AU the insulators that flashedover had visible defects and showed an electric field distributiondifferent to that initially Obtained. The comparison between thesound field and aged field curves for these insulators validatedthe location of the defects along the insulator. On the otherhand, the electric field distribution measured on insulatorswhich had no visible defects was similar to that initiallyobtained. This means that for these cases the flashover was onlycaused by the pollution phenomenon, and not by any defectFurthermore, the electric field distribution measurements on allthe insulators at the middle and at the end of the ageing testwere performed, in order to mainly determine if some internaldefect in the insulators, without visible defects, had occurred.According to the results, uo-defects were found in the insulatorswithout visible defects, being as their aged fieid curves equaledtheir sound field curves. It is concluded that this diagnostictechnique may be used io service to evaluate theperformanceofnon-ceramic insulators if the measurement methodology issystematically applied and the electric field detector is used in aproper way. The insulators that did not suffer damage duringthe ageing test show a good correspondence between the electricfield curves measured before and aner ageing.INTRODUCTIONThe main comDonents whose deterioration could interrupt

field distribution [5], the daylight corona discharge imager[6], the leakage current measurement [7] and the partialdischarge measurements (PD).In Mexico, as a test plan, NCI having different designs havebeen installed on several transmission lines of 115 kV,230 kV and 400 kV. On some insulators, their leakagecurrent is being morutored by a leakage current systemdeveloped in Mexico [SI . This is in order to apply the electricfield technique in evaluating the state of the insulators anddetermine whether there is or not a relationship with theleakage current and the insulator ageing. To achieve thispurpose the study developed is presented in this paper.Test MethodFifteen NCI from different manufacturers were aged in a saltfog chamber by applying both non-standard methods as wellas the standard solid layer method 191. First, the electric fielddistribution along each insulator to be.aged in the fogchamber was measured, obtaining an average electric fieldcurve for each insulator (sound field curve). Second, someNCI were polluted with different pollutants by the solid layermethod, and then, the ageing test started. During the ageingtest, the insulators were visually inspected. When damage wasobserved on the insulator surface and such damage causedthree times a flashover on the insulator, the insulator wasremoved from test and left ahout 24 hours its surface to dry.The next step was to measure the electric field distribution ona damaged insulator and also to compare the field curvescanned with i ts sound field curve.

transmission l i e operation and to which can be applied on-line diagnostics, are conductors and insulators. Thetechniques used to detect damage in these components can bedivided into three erouus 111: &tical. acoustic and electrical . . .The device used to measure the electric field distributionalong the insulators evaluated was developed by Hydro-Quebec (IREQ) [ IO] and at present is manufactured by the- . _ _ .techniaues. Electric techniaues are the leakaee current pos'tron lnduStneSnc.

~measurement and the electric field measurement along theinsulator. By using electric field measurement, the defect islocated where the typical electric field distribution changes ina way moreor less abrupt.The sensor installation should be mounted on a strongmechanical base, the conditions of the distance of theinsulator-sensor and the sweeping velocity of the sensor arealways repetitive; another condition is to have approximately

Diagnosiic techniques or non-cerumic nsulatorsTo determine ageing in NCI, simple and complicatedtechniques areused. A simple technique isthe visual aid byusing either binoculars or hydrophobic visual observations[2]. A complicated technique is the measurement of thecontact angle [3], which is difficult to apply in field. Othertechniquesused arethe infrared thermography [4], the electric

an equal relative humidity for each measurement (lower than80%. To obtain a graph, 5 tests on each sample were doneand the average of the 5 setsof readings obtained are plotted.Besides the standard deviation for the 5 readings at each shedwas calculated, being in most cases standard deviationssmaller than 2%. Of the total readings performed for all theinsulator sheds, 8.35% of all the readings had a greater

0-7803-8447-4/04/$20.0002004 EEE. 41 1


2/4

standard deviation than 2% for which the repetitive of themethod is right. Normally, in the readings having greaterstandard deviation involved the hot line sheds.As in the circuit shows, the insulator is placed in a verticalposition, at the lower sheds the connector of high tension isfound, while the upper sheds are located at the ground. TheUFP (universal field probe) is connected to a pole, which isinstalled in a manual mechanical crane (Fig. 1).

the electric field after 1670 hours ageing and the curve Ccorrespondstothe test ending at 3277 hours(fmal condition).Insulatorsdamaged during test.I) Insulator#2 contaminated withkaolinThe Fig. 3shows the electric field distribution and insulatorphysical condition when it failed during test (at 1430hours).In curve B some variations are observed from shed 1 until

,

Figure1 -Equipmentforelectricfielddistributionmeasurement.The sweeping of the insulator was done downward (side

j0 1 I 1 4 I 8 7 I s / o ( 3 (2 11 I . 15 l b


3/4

show a deformation among sheds 6 and 12, in this section,besides erosion and chalking, the . insulator shows largeformations of tracking and cracks involving the rod and thesheds. Like insulator # 2, this unit did not return to thechamber and the curve C was also taken to the end of the test.CurveC shows a similar behavior to that of curve B.3) Insulator#12 cleanThe insulator # 12 was aged previously in another test anduntil this time, the insulator had tracking (13 cm of length),which at the present evaluation was deepened and extended to14,3 mm. This insulator was removed from the salt fogchamber after 2165 hours of test when it had accumulated 3flashovers. The electric field distribution and the physicalcondition of the insulator are shown in figure 5 . TheUFP does not register the electric field when the insulatorhas alternate sheds, since the infrared sensor detects duringsweeping the presence of the sheds. If they are of differentdiameter, the small ones are not detected.1"."1.1.1.,1 '

c*VAL% =

Figure 5 - Electric field distribution and physical condition ofinsulator#12.The tracking among sheds 7 and 9 could not be detected inthe electric field distribution curve, since the most severedamage was found in shed 8 and that is a smaller shed thanthe sheds 7 and9.The recommendation is to have provision that the UFPregisters for each shed. In this case, the electric field curveswould indicate that there was not any damage, because thereis not a great variation among the curves A,B andC.8 nsulator#I4 cleanThis insulator was the fourth that failed. Its duration insidethe chamber was 2464 hours. The electric field distribution

along this insulator is shown in figure 6.For this insulator thevariation of the electric field due to the damage suffered isvery clear (curve C). Physically, it represents degradation bytracking with formation of cracks involving the rod and thesheds; the depth of the cracks propagate until the glass fibre.bar.The damaged zone is found among sheds I O and 19 with atotal length of 46.5cm. In the graph a variation in the electricfield is shown starting from shed IO, which is in agreementwith the visual inspection. The maximum electric fieldmaximum point was displaced until shed 13 due to damagethat there was between this and the hot l ine side forming aconductive medium.

Figure 6 ~ Electric fied distribution and physical condition ofinsulator#14.Insulators not damaged during the evaluationI ) I nsubtor#I contamnatedwith ertilizer.The electric field distribution curve of this insulator is shownin figure 7.The three curves present the same behavior (the"U" shape), except at the end of the graph B where isobserved a decrease inthe electric field magnitude inshed 19,however, physically no.abnormality was seen near this shed.

ImY1 I, ,*dlli.,

0 , I 1 4 6 . 7 8 . 0 I , TI 3, (1 ,I 1 I 7


4/4

This insulator and insulator # 6 consisted of two modularweathershed housings with 16sheds each, that cover the rod.These modules are mechanically bonded to the adjacentmodule by an external polymer collar. This insulator has 32sheds; however, a similar arcing distance compared with theother insulators was considered (25sheds). In curve B asmallvariation is observed among sheds 13 and 18 (the polymercollar is among sheds 16 and 17), but in the last curve (C) thisvariation is not observed (Fig. 8). Physically, only theinsulator has erosion and chalking.3) Other insulators hatdid not ail in the testInsulator #4 with natural pollution, insulator #5 clean,insulator #6 contaminated with lime, insulator #7contaminated with lime, insulator #8contaminated with lime,insulator #9 contaminated with l i e , insulator # O clean,insulator #11 contaminated with lime and insulator #13 clean.

I.t. . I,...F 1.0f

0 ! 2 , . , , * 10 I , >*!,. C I. $ 7 1) I. 10 1% 21 I I * I SEh..,

diagnostic of nonceramic insulators aged in a salt fog chamber by using electric field sensor

Documents