Discussion on

Short-line fault tests on the CEGB 275 kV system

A.Ametani: The authors have contributed valuable informa-tion on short-line faults in an e.h.v. transmission line. Al-though analytical results agree well with the field test re-sults, several questions may be raised concerning the analy-sis.

The earth resistivity has been taken as 20 ftm. In my ex-perience, earth resistivities from field tests are in the range50-300 ftm; in comparison 20 fim seems to be too low. If theactual resistivity is higher, then the calculated line inductanceand surge impedance increase. This causes further dis-crepancies in the comparison between the measured and thecalculated results. Also, resistivity may vary with fre-quency, since the earth is not homogeneous. Clarification ofthe following points would be helpful:

(a) Is an earth resistivity of 20 flm the measured value ?(b) If so,at what frequency was the measurement taken?(c) If not, on what basis was the figure of 20 Om chosen?

In calculating the transient recovery voltages on the sourceside, the authors concluded that it would appear that the inclu-sion of frequency dependence caused the increase in peak,value and time, leading to the conclusion that the differencesbetween the test and calculated results cannot be attributedto frequency-dependent effects (Section 12.2.1). This conclu-sion is based on a comparison of the calculated results atconstant line parameters with those of method 2 (Fouriertransform), when frequency dependence at 20 Om earth resis-tivity is included. Usually, however,the inclusion of fre-quency dependence causes attenuation and retardation, andhence a decrease in peak value would be expected.

With a somewhat different example, it has been my experiencethat a higher resistivity causes a decrease in peak value, andthat initial conditions significantly affect the first peak of thetransient recovery voltage.

This arises from calculations on a simple single-phasetransmission line using a computer technique similar to theSchnyder-Bergeron method, though including frequency de-pendence. Would the authors please comment on the follow-ing points:

(i) Were calculations carried out for different earth re-sistivities, using method 2?

(ii) If so, does a higher resistivity cause further increase inpeak value ?

H.Heimer: One of the topics which is of interest to designengineers is the distribution of the earth-fault current. Wereany measurements taken during the tests to determine whatproportion of current was carried by the overhead-lineearth wire,and what proportion went through the ground?Were any checks made as to whether this current distribu-tion conformed to the theoretical distribution, such as theone given by Carson and Pollaczek?B

One is also very impressed by the prodigious amount of workthat has gone into the source-side transient-recovery-voltagecalculations. There must therefore be concern as to whetherthis sort of detailed analysis is necessary every time a cir-cuit breaker is specified for a transmission-system exten-sion. How detailed, in the authors' opinion, should these cal-

Paper 6101 P by BOLTON, E., BATTISSON, M .J., BICKFORD,J.P., DWEK, M. G., JACKSON, R. L., and SCOTT, M.[see 1970,117, (4), pp. 771-784]

Read before the IEE North-Western Centre, at Manchester,27th February 1973

culations be when such extensions are contemplated?

E. Bolton, M. J. Battisson, J. P. Bickford, M. G. Dwek, R. L.Jackson, and M. Scott (in reply): In reply to Mr. Ametani, weshould first point out that it is the apparent total earthimpedance that varies with frequency, and not the earth re-sistivity, which is a nonvariable parameter. We took an earthresistivity of 20 fim for our test-line 50 Hz and surge-impedance calculations, as a result of measurements madeprior to the short-line fault tests. The measurements weremade using direct current. Line parameters, chosen for thesource-side t.r.v. assessment using methods 1 and 3, werebased on the predominant transient frequency of 1 kHz.Whether the elaboration of assuming complete frequencydependence of the parameters causes an increase or reduc-tion in the peak t.r.v. as compared with fixed-frequencymethods will depend on the frequency spectrum of the over-all wave. We agree with Mr. Ametani's general commentthat higher earth resistivities will cause a decrease in thepeak t.r.v. Earth resistivity is, however, only one parameterof many that interact in determining the t.r.v. waveform.One source-side line parameters were based on an esti-mated earth resistivity of 20 fim, and, over the wider areacovered, perhaps a higher value of about 50 £2m would havebeen more appropriate. Reference A shows the effect ofincreasing both frequency and earth resistivity on line surgeimpedance. The effect is greatest for the last pole to clearthe bottom-conductor condition, but, for the average condi-tions assumed in the source-side t.r.v. calculations, theeffect of earth-resistivity variations on the source-sidet.r.v. is small, and for this reason was neglected. ToMr. Ametani's final question, studies were not performed bycalculation method 2 (Fourier-transform method) fordifferent earth resistivities.

In reply to Mr.Heimer, measurements were made on the testline during the fault tests to determine the proportion offault current that was carried by the overhead-line earthwire. This amounted to about 60% at the faulted tower whenthe adjacent circuit was out of service, and about 55% whenthe adjacent circuit was on load (tests 9a-12). Calculations'were not made to check this measured current distribution.

We do not think that such detailed analysis of source-sidet.r.v. is necessary every time a circuit breaker is beingconsidered for application in a transmission system. Ourobject in the work reported was to develop and validatemodelling techniques by use of test results. In the main,simplified assessment methods can be used to check t.r.v.conditions based on the immediate source-network configura-tion, local infeeds, circuit parameters and circuit lengths.Source-side t.r.v. conditions so derived can be comparedwith national or international standard specifications, and, iffound to be within the specified levels, no further analysis isrequired. Detailed analysis is, however, required both whenoverall specifications are being prepared and also when cir-cuit breakers in service may become overstressed, owingto planned changes in system conditions, and decisions onuprating must be made.

References

BOLTON, E., BIRTWHISTLE, D., BOWNES, P., DWEK,M. G., and ROUTLEDGE, G.W.: 'Overhead-line para-meters for circuit-breaker application1, Proc.IEE, 1973,120, (5), pp. 561-573LACKEY, C.H.W.: 'Fault calculations' (Oliver & Boyd,1951), p. 152

PROC.IEE, Vol.120, No. 11, NOVEMBER 1973 1417