Principal Research Results
Reasonable Selection Method of Test Voltage in Gas Insulated Substation
In the insulation design (sizing) of gas-insulated switchgear (GIS), which is the main equipment of the substation, the high-
est lightning surge in many kinds of overvoltages is important to the test voltage. However, the lightning surge which invades the GIS
is a complex waveform including steep and oscillation element, and it differs greatly from the standard lightning impulse waveform
which is test waveform of the equipment. Therefore, the establishment of a reasonable evaluation method of the test voltage value to
the lightning surge is needed. We have proposed the evaluation method of the test voltage level by a quantitative estimation method of
the V-t characteristics which uses the equal-area criterion based on the square impulse (Fig.1).
To verify the equal-area criterion which uses the parameter quantified by square impulse. To evaluate the equivalence of a s-
tandard lightning impulse and the lightning surge using the equal-area criterion and to examine the reduction possibility of present
lightning impulse withstand voltage (LIWV).
I In the GIS model electrode at practical SF6 gas pressure (0.5MPa), the short time range V-t characteristics were measured by the
square impulse, and the equal-area parameters were quantified. The equal-area criterion was applied to the lightning impulse wave-
form and the nonstandard lightning impulse waveform which modeled the lightning surge (up to 1.45MHz 1) and the disconnector
switching surge (up to 5MHz: included the double frequency oscillating impulse 2) based on this result. As a result, the minimum
sparkover voltage which becomes the most important point in design of the GIS was excellently evaluated (Fig.2), and it was shown
that the equal-area criterion was useful as the V-t characteristics quantitative evaluation method to any waveform (In 0.75MHz os-
cillating impulse, there is an error margin of about 200ns in 600kV/cm from 500kV/cm. This factor of error is thought of as the mea-
sured time to sparkover include the statistical time lag. However, this error margin does not become a problem because it does not
use the time to sparkover as a parameter in this method.) .
II The lightning surge which invades the standard substation layout 3 for each nominal voltage(66-500kV) was analyzed by EMTP 4,
the equal-area criterion was applied to this result and converted into the value of standard lightning impulse voltage (testing wave-
form). The principal results are summarized as follows.
(1) As a result of analyzing all nodes of standard substation layout, the node which shows maximum peak value of the lightning surge
by EMTP does not necessarily show the maximum value when converted into the standard lightning impulse (Fig.3, Table 1). This
result differs from the past method which is judged by only peak value of surge waveform, and is a typical result of the equal-area
criterion which considers even wave duration.
(2) The lightning surge converted into a standard lightning impulse with most cases is lower than present LIWV (Table 2). There is a
possibility of LIWV reduction in the GIS of 220kV or less in nominal voltage especially. In cases of over 275kV in nominal volt-
age, there are cases which exceeded present LIWV, but LIWV reduction is possible by the appropriate arrangement of the arrestor.
When "JEC-0102-1994 Test voltage standard" is revised, we propose the V-t characteristics estimation method (equal-area
criterion based on the square impulse) and submit draft proposal of the present LIWV reduction value.
Main Researcher: Hiroyuki Shinkai,
Research Assosiate, Electrical Insulation Dept., Komae Research Laboratory
E H.Shinkai, H.Goshima and H.Fujinami, 2002, "Study of Lightning Surge Region Voltage-time Characteristics Evaluation Method
of GIS -Quantitative Evaluation method by Steep-front Square Impulse Voltage-", CRIEPI Report, No.T01017 (Japanese only)
E H.Shinkai, H.Goshima, H.Matsubara and H.Fujinami, 2003, "Possibility Evaluation of Test Voltage Reduction Based on
Equivalence of Lightning Surege in Gas Insulated Substation and Standard Lightning Impulse Test Voltage Waveform -Evaluation
by Equal-area Criterion Based on Square Impulse-",CRIEPI Report, No.T02013 (Japanese only)
1 "V-t Characteristics of SF6Gas for Lightning surges", CRIEPI Report, No.T91092 (1991) (Japanese only)
2 Okabe et al.: Insulation characteristics of GIS under oscillatory waveforms of non-standard lightning impulse-Insulation characteristics under
single- and double-frequency oscillatory waveforms and evaluation of disconnector switching surges-", Proc. of the 11th ISH, Vol3
3 "The Suggestion of the Useful way to Analysis of Lightning Overvoltages and Application", CRIEPI Report, No.T90068 (1990) (Japanese
4 Electro Magnetic Transients Program
4. Power delivery - Reducing costs of electric power transmission facilities
<Equal-area criterion> <OSI=0.75 > <OSI=5.4 >
The concept of equal-area criterion is that the time integral
value in the part which exceeds V0 dominate sparkover.
0.0 1.0 2.0 3.0 0.0 1.0 2.0
Time[ µs] Time[ µs]
Applied electric field (E) [kV/cm]
500 parkover 0.75MHz oscillating impulse 5.4MHz oscillating impulse
400 E0 Af 700
OSI =0.75 OSI =5.4
300 Analyzed value
200 600 by square impulse>
E 0=355 [kV/cm]
550 A f=5.72E-3 [V*s/cm]
0 tf Lower limit of analytical
450 value (OSI =5.4)
0 200 400 600 400 Lower bound of analytical
time[ns] value(OSI =0.75)
5 6 7 2 3 4 5 6 7 2 3 4 5 6 7 2
<Equal-area parameters> 0.1 1 10
E0 : the minimum sparkover voltage Time to sparkover [ µs]
Af : Constant decided by formative time lag and E0 Fig.2 The result of the measured V-t characteristics and
(These two parameters were decided by square impulse) estimated V-t curve by equal-area criterion (The
minimum sparkover level of the analytical value
Fig.1 Equal-area criterion based on the square impulse. and measurement value fitsexcellently).
Peak voltage 817.5[kV] Peak voltage 869.5[kV]
0.8x10 G1-0R 0.8x10 G1-6R
0 2 4 6 8 10 0 2 4 6 8 10
Time[ µs] Time[ µs]
(a) G1-0R (b) G1-6R
Fig.3 Analysis example in EMTP, Nominal voltage is 275kV. The length of service entrance
conductor is 30m (GIL). The length of Conductor to transformer is 50m(GIL). The ar-
restor is set up on the service entrance and the bus side of the circuits to the trans-
former. The decrease rate of the discharge voltage of the arrestor is 10%.
Table 1 Peak value of lightning surge of fig.3 and standard lightning impulse conversion voltage by equal-area criterion
(a) G1-0R (b) G1-6R # The peak voltage of G1-6R is higher than G1-0R, but the dura-
Peak voltage of lightning surge 817.5kV 869.5kV tion of peak voltage of G1-0R is longer than G1-6R. Therefore,
Peak voltage of standard when both surge waveforms are converted into a standard
lightning impulse (Conversion 744.6kV 742.7kV lightning impulse by the equal-area criterion, G1-0R becomes
value by equal-area criterion) consequentially high.
Table 2 Result of conversion into standard impulse voltage value by equal-area criterion
Nominal voltage [kV]
66 350 250kV or less 250kV or less 250kV or less
77 400 300kV or less 300kV or less 300kV or less
750kV or less (700kV or
154 750 650kV or less 600kV or less
less in most cases)
750 700kV or less (650kV or
220 650kV or less 650kV or less
900 less in most cases)
950 950kV or less in most ca-
275 850kV or less ses, but 1050kV or more 900kV or less
1050 in rare case.
1300 1550kV or less (1425kV 1425kV or less in more
1425 or less in most cases, cases (1300kV or less in 1425KV or less (1300kV
500 1 3 0 0 k V o r l e s s i n a r- some cases), 1550kV or or less in most cases)
1550 restor performance 30% more in rare case.
5 Value in this analytical case (standard substation layout 3). When the composition and the arrangement of the substation are
different, it is not the same as this result. Therefore, it is necessary to analyze EMTP individually, and apply the equal-area