DocumentCode :
874402
Title :
Statistical approach for determining breakdown voltage of gas-insulated cables
Author :
Yildirim, Fetih ; Korasli, Celal
Author_Institution :
Dept. of Stat., Middle East Tech. Univ., Ankara, Turkey
Volume :
27
Issue :
6
fYear :
1992
fDate :
12/1/1992 12:00:00 AM
Firstpage :
1186
Lastpage :
1192
Abstract :
Several goodness-of-fit test procedures were applied to DC-breakdown conditioning test results, obtained in SF6 at 0.4 MPa in concentric-coaxial systems, and to other published data to determine whether the three-parameter Weibull distribution is the only match. The three-parameter Weibull distribution was found to be the best suited among probabilistic exponential distributions representing the area-dependent statistical characteristics of compressed-gas insulated systems. Parameters of the three-parameter Weibull distribution were estimated by simultaneous solution of maximum-likelihood equations for all conditioning data. In contrast to the experimental results given in the relevant literature, the estimates of the parameters were found to be dependent upon the pressure and geometry of the gap. The value of the shape parameter β≃7 seems to hold only for the conditioning data at 0.6 MPa. The lowest asymptotic value of the threshold estimator-minimum breakdown field strength was observed to be ~141 kV/cm
Keywords :
electric breakdown of gases; electric strength; gaseous insulation; power cables; sulphur compounds; 0.4 to 0.6 MPa; DC-breakdown conditioning test results; SF6; area-dependent statistical characteristics; breakdown field strength; breakdown voltage; compressed-gas insulated systems; concentric-coaxial systems; conditioning data; experimental results; gas pressure cables; gas-insulated cables; geometry dependence; goodness-of-fit test procedures; maximum-likelihood equations; pressure dependence; probabilistic exponential distributions; shape parameter; simultaneous solution; statistical approach; three-parameter Weibull distribution; Electric breakdown; Equations; Exponential distribution; Geometry; Insulation; Maximum likelihood estimation; Parameter estimation; Shape; System testing; Weibull distribution;
fLanguage :
English
Journal_Title :
Electrical Insulation, IEEE Transactions on
Publisher :
ieee
ISSN :
0018-9367
Type :
jour
DOI :
10.1109/14.204870
Filename :
204870
Link To Document :
بازگشت