12
Classification of discharge patterns during ageing of insulation Abstract Introduction Discharge detection and reco gnition Short-term ageing of cavitie s Long-term ageing till breakd own of a 12kV current transformer

Classification of discharge patterns during ageing of insulation

Embed Size (px)

DESCRIPTION

Classification of discharge patterns during ageing of insulation. Abstract Introduction Discharge detection and recognition Short-term ageing of cavities Long-term ageing till breakdown of a 12kV current transformer Conclusions. Abstract. - PowerPoint PPT Presentation

Citation preview

Page 1: Classification of discharge patterns during ageing of insulation

Classification of discharge patterns during ageing of

insulation

Abstract Introduction Discharge detection and recognition Short-term ageing of cavities Long-term ageing till breakdown of a 12kV current transformer Conclusions

Page 2: Classification of discharge patterns during ageing of insulation

Abstract

This paper investigated discharge distributions during ageing of artificial cavities.

Conventional discharge detection with statistical

processing of discharge signals analyze a 12kV current

transformer.

Using various mathematical techniques a data base of

discharge patterns.

Recognition of discharges in HV components.

Periodic testing of HV equipment.

Page 3: Classification of discharge patterns during ageing of insulation

Instroduction

Recognized in the past that the degradation of insulation by discharge takes place in stages.

Recent research on discharges in cavities in polyethylene => At least three consecutive stages In third stage : formation of pits on the cavity = final breakdown of the insulation Discharge measurment => estimate of an ageing stage of HV comp

onent Observe discharge patterns during ageing of artificial cavities and a

12kV current tansformer and to classify the patterns according their ageing stage.

Page 4: Classification of discharge patterns during ageing of insulation

Discharge detection and recognition

PD measurement : statistical discharge analyzer(TEAS 570 by Haefely ; bandwidth 40-400kHz)

The shape of the maximum pulse height distribution : The shape of the mean pulse height distribution : The shape of the pulse count distribution : The number of discharge as a function of the discharge magnitude :

H(q) The number of discharges as a function of the discharge energy : H

(p) H(q), H(p) are described by statistical parameters => skewness, kurtosis

)(max qH

)(qnH

)(nH

Page 5: Classification of discharge patterns during ageing of insulation

In this way a set of 29 parameters : fingerprint => a basic element for the recognition The centour score method : indicates the match between fingerprint

s. => 100% for a perfect fit, 0% for a complete lack of resemblance

Discharge detection and recognition

Page 6: Classification of discharge patterns during ageing of insulation

Short-term ageing of cavities

Polyethylene(diameter 5-9mm, height 0.4-0.5mm) Tree stage of ageing due to PD; (a) virgin : first 2 min. after reached test voltage (b) conditioned : 5-10 min. from the beginning (c) aged : 90 min. from the beginning Fingerprint collecte : each aged stage – test voltage 50-80% Pattern recognition purpose : single classification category, repres

ented by a number of finger

Page 7: Classification of discharge patterns during ageing of insulation

Short-term ageing of cavities-Simplicity phase-related

distributions

-Significant change in the

distribution : short time

(a) virgin stage : atypical patterns, equal discharge magnitude in both half-cycles

, metallic oxide layer on the surface of a metallic electrode.

(b) Conditioned stage : asymmetry, ‘burn-out’ of a metallic oxide layer.

(c) Aged stage : rapid changes in discharge patterns.

Page 8: Classification of discharge patterns during ageing of insulation

Short-term ageing of cavities

-Total of 26 fingerprints were classified

-Cluster analysis of fingerprints

-The group average method

-Sorts fingerprints in the form of a tree

-’branchs’ can be identified

-Fingerprints of each stage : used for the creation of a data(an assessment of

condition in discharge site)

Page 9: Classification of discharge patterns during ageing of insulation

Long-term ageing till BD of a 12kV current transformer

-A 12kV current transformer

-Discharge at a 28kV

-Cause of discharge : cavities, cracks situated

-900 hours, increased in step 45~90kV

-H(q) : three distinct peaks

-The test voltage of 40kV

-After few hours : discharge extinguish

=> (sensitivity of 1pC)

Page 10: Classification of discharge patterns during ageing of insulation

Long-term ageing till BD of a 12kV current transformer

-No discharges detectable at 40kV

-Everyday 40~90kV increased

-After stage 2 : No meaurable discharges

-After 850hour : reappeared after 50hours

=> about 110pC, test voltage(65kV)

-Fig 5(b) : group(ageing stage) as a function of

the ageing time

-Three different groups of fingerprints

-The collecyed fingerprints : cluster analysis

-No discharges stage : sensitivity(1pC)

Page 11: Classification of discharge patterns during ageing of insulation

Long-term ageing till BD of a 12kV current transformer

-Discharge distributions

Page 12: Classification of discharge patterns during ageing of insulation

Conclusions

Discharge patterns during ageing - artificial electrode bounded cavities - A 12kV current transformer1. The discharge patterns changed several times during ageing period.2. Cluster analysis, the group average method : ageing stages during

the ageing tests3. Centour method : classification of fingerprints to ageing stage

4. Recognition tools(the group average method in combination with the centour score method) have a good potential for industrial application(recofnition of discharges in HV components, periodic testing of HV equipment)