The Determination of Switching Surge Withstand Voltages for EHV Insulation Systems

New algorithm for determining critical flashover voltage and standard deviation from flashover probability data

“…6 and 7). weight assigned to the kth point assuming known CT (v5()), measured V, , using V, and assumed known cr uh function related to w, C , , C2 constants Vso( 1) critical flashover voltage, assuming the first point is fixed VMXJ( 1) mean voltage value at 84.13 % probability, assuming the first point is fixed V, , (2) critical flashover voltage assuming the second point is fixed VMw(2) mean voltage value at 84.13 % probability, assuming the second point is fixed V,, (3) critical flashover voltage, assuming the third point is fixed vM,4(3) mean voltage value at 84.13 % probability, assuming the third point is fixed V,, (4) critical flashover voltage, assuming the fourth point is fixed VMa(4) mean voltage value at 84.13 % probability, assuming the fourth point is fixed V, , accurate voltage at 84.13 % probability…”

“…The two parameters are determined such that L is maximum. It is usually more convenient to work with the natural logarithm of L, thus (4) The critical flashover voltage is given by Performing the partial differentiations and denot- The proposed algorithm avoids the computational difficulties that arise in the two methods (Maximum Likelihood 'and Graphical Method). Referring to the Maximum Likelihood Method, it has been noticed that exact solution for the partial derivations (aS,/aV5,, as,lao, as2ldv5,, as,ldo) are very complex sincep,, z k and Yk depend on V50 and o i n complicated fashion (see eqs.…”

“…Determination of V, ,(4) and vM84(4). assuming the fourth point is tixedE TEP number of voltage applications causing flashovers of The steps ofthe program are as follows: each level, and -number of voltage applications of each level.…”