Beruflich Dokumente
Kultur Dokumente
4 , July/August 1975
Harvey E. Spindle
ABSTRACT
end supports and a single ground strap goes from the top of the end
support to the substation ground mat as shown. The first step in the
sounding analysis of this system involves the calculationofthe
inductance.
TheGroundingof
Gas InsulatedSubstations
(GIS) presents
several unique
problems
not encountered
in
the
grounding<
of
paper,
Gas
conventional
substations.
Forthe
purposes of fhis
Insulated
Substations
are defined as high-voltage, compact,
metal-enclosed minisubstations which are compressed g a s insulated at
maximum rated voltages of 72.5 kV and above. The metal enclosureis
grounded
and
consists
of theoutersheathforthe
substation
equipment and the coaxial gas insulated bus. The maximum voltage
developed on thegrounded outersheath during faultconditions is
calculatedusing
couplingtechniquesandatransformerequivalent
voltage to
circuit. By settinglimits on themaximumoutersheath
station ground, the maximum distance between grounding points can
be determinedfor
thesafety
ofpersonnel
in a Gas Insulated
Substation.
INDUCTANCE
Outer Sheath
The inductance per unit length of the outer sheath with respect
to ground (assuming an infinite ground plane) can be calculated from
the following equation1 :
P = 4.61 x
where
INTRODUCTION
and
Now
&= inductance/length
(henrys/m)
(1)
P =henrys/m
7.84
x
XP = 2nfP = inductivereactance(ohms/m)
where
f = 60Hz
Thus
Xp =ohms/m
2.96
x
(2)
Ground Strap
The inductance per unit length of the horizontal ground straps
withrespect to ground (assuming an infinite groundplane) can be
calculated in the same manner:
-7
where
nus
ag
= inductive reactance
(henrys/m)
(3)
Pg = 1.17
henrys/m
x
= 21rfpg = 4.41 x
And
ANALYSIS
ohms/m
It is assumed that the 3.96 m vertical ground strap has the same
inductive reactance per meter as the two 2.13 m horizontal ground
straps.
MUTUAL INDUCTANCE
Themutualinductanceperunitlengthbetweentheinner
three phases can be calculated
conductorandoutersheathforthe
from the following equationsz:
TheinstallationatSouthern
CalirorniaEdisonconsists
ofa
center cablefeed, two circuit breakerpositions,and two overhead
transmissionline
connections(one at each end). Thethreeouter
sheaths of the system are connected together at the top of each set of
m = 4.61
where
loglo
-7
GMD
10loglO==
mutualinductance/length
(henrys/m)
(4)
1183
1.Determinetheapproximateinductivereactance
equations ( I ) , (21, and (3).
Location of
Maximum Voltage
Breaker
Circuit
GMD=
where
- d,.,
dab=distancebetween4Aandq5B=2.13m
GroundStraos
/&Circuit
m = 5.68 x l d 7 henrys/m
And
Breaker
3.96m
Now
And
,,X
.,
= 2.14 x
-k
4.27m
2.1 3m
ohms/m
b-
dac
/ J
9.75m
Tab
4-
8.89cm
0.953cm
Fig. 3. 242 k V bus dimensions.
1184
7-
External Fault
FAULT CONDITIONS
The outer sheath of the system can be exposed to two types of
fault conditions: 1. An internal fault within the SF6-insulated system;
2 A fault external to the SF6-insulated system in which fault current
flows through the CIS systemandinducescurrents
in theouter
sheaths. The internal faultw
libe investigated first.
Internal Fault
Z m u =~mutual impedance
follows:
where
Zoutenheath = 9.75
3.96 m ) x ~= 4.06 x
ohms
sheath
of outer
r2 = inner radius
It can be. shown that the maximum outer sheath voltage is developed when the internal fault current divides such that 25 kA flows in
each direction in the outer sheath on either side of the fault location.
Such conditions exist whenever the internal fault is located approximately at the top of the outer breaker tank. Since the distance from
the center cable feed to the top of the outer breaker tank is approximately 10.1 meters, the impedance and maximum
voltage to station
ground developed are as follows:
(XQ)x(length) = (2.96xohms/m)(10.lm)
v,
= (25 k ~ (2.99
)
x
= 2.99 x
= 15.2 cm
ohms
ohms) = 75 volts ms
conductor = 1.27 x
ohms
50kA
outer sheath
outer sheath
ground strap
vertical
outer sheath
--
--
&
.
outer sheath
outer sheath
z ground strap
vertical
outer sheath
---
- Z mutual
- Z mutual
0
outer sheath
t
ground
strap
horizontal
outer sheath
z ground strap
vertical
support insulators were used to obtain the clampon current measurements.The field testmeasurements made on the GIS installation
yielded the results shown in Table I which also shows the theoretical
current values.
Table I - Field Tests
11= 1080amperes (primary current ininner conductorof center phase)
622
Theory
Test
Outer
Sheath
Outer
Sheath
Outer
Center Phase
(12)
Outer Phase
Sheath
Outer Phase
amperes
amperes
129
475 amperes
130 amperes
129 amperes
173 amperes
The significance of the field test data is the fact that the induced
currentontheoutersheath
of thecurrent carrying phase (475
amperes) is verifiid to be on the order of 50% of the inner conductor
current (1080 amperes) rather than the 9 0 7 ~ 9 5 %usually predicted
(see IEEE Guide for Calculating Losses in Isolated - Phase Bus, ANSI
C37.23).Thenext
stepinthe
f i l d tests was to make voltage
mekurementsonthe
GISsystem; however,becauseof
induced
voltages and currents from the overhead power transmission lines in
thesubstation,it
provedimpossible to obtain meaningfulvoltage
measurements
despite
several attempts which employed various
measuring techniques. Therefore it was deemed necessary to build an
electrical scale model of theGIS installation in ordertoobtain
meaningfulvoltage measurements for verificationof thepredicted
voltagevalues. Figure 7 shows a photograph of the test model and
Table I1 compares it with the actual GIS system.
1186
Actual
System
h
GMR
XP
GMRg
%
GMD
Xam
2
1
overall length
dab
dac
dbc
0.305 m (1 ft)
4.32 cm (1.7 in)
200 p ohms/m
(60.9 p ohms/ft)
0.635 cm (0.25 in)
345 pohms/m
(105 p ohmslft)
0.768 m (2.52 ft)
218pohms/m
(66.4 p ohms/ft)
4.51 cm(1.78in)
1.33 cm (0.524 in)
1.52 m (5 ft)
0.610 m (2 ft)
1.22 m(4 ft)
0.610 m(2 ft)
Theory
Test
Thefirst stepinthelaboratorytest
using theelectrical scale
model involved themeasurementofcurrentsinduced
on the outer
sheath due to current flowing in one of the inner conductors. The
same procedure was followed in the laboratory current test as in the
GIS system. However, inthelaboratorytest,the
field testofthe
current was first applied to the inner conductorof the center phase of
the model and then to one of the outer phases. The results are given m
Table In.
CONCLUSIONS
amperes
57 327
Theory
Test
Outer Phase
amperes
amperes
amperes
67275
Outer P h a ~
amperes
57
1 13 amperes
amperes
amperes
amperes
amperes
37
146
61
20.1 mV
25.3 mV
Outer
Sheath
Outer
Sheath
Outer
Sheath
Center phase (12)
33.8mV
34.5 mV
Center Phase
Outer Phase
60.8 amperes
28.2 amperes
1187
2. Determinethemutualinductance
per meter using equation(4).
Resistancevalues arealmost alwaysnegligible compared to the
inductive reactance.
3. UseFigure 4 to determinethe outersheathcurrentandthen
voltage per meter for an internal
fault.
4. Use thetransfomer
equivalentcircuit
given in Figure 6 to
determinethe outersheathcurrent
asafunction
of theinner
conductor, current for an external fault. The voltage per meter on
the outer sheath can then
be calculated for an external fault.
5. Compare the outer sheath voltage per meter for the internal and
external fault to determine the worst
case.
6.Based uponthedistance
between ground points, themaximum
voltage on the outer sheath can be determined. Conversely if the
distancebetweengroundpoints
is desired, the maximum outer
sheath voltage allowed is divided by the voltage per meter.
Currently there are no standards or limits for maximum allowable outersheath voltage on Gas Insulated Substations.The limits
presently used Seem to be as numerous and widespread as the utilities
employing them. The IEEE Guide for Safety in AlternatingCurrent
Substation Grounding stated: Some utilities establish more or less
arbitrary voltage limits, e.g. 150 volts . . . due to faultgradients . . . for
duration of shock not exceeding1.2 seconds.Althoughthe maximum value of 150 volts rms is not a recommended level (but just a
limit used bysomeutilities), the maximumvalue ontheSouthern
California Edison installation of 75 volts rms Seems to be within the
limits of establishing a safe system.
Thus a method has beendeveloped fordetermining the maximum outer sheath voltage and the maximum distance between ground
points for SF6-insulatedsystems. This information is then used to
determine the location ofthe ground points in the CISsystem. Several
simplifying assumptionsaremade
to enable calculation ofthe
inductances which areafunctionofrathercomplicatedgeometric
configurations. Field tests on an actual, GIS
installation and laboratory tests onan electrical scale model have proved that the simplifying
assumptionsare permissible and thatthe resulting theoretical apis valid. Thusa simplified,
proach to theGISgroundingproblem
for GISsystems hasbeenprovided
practicalgroundingprocedure
which will insure the
safety
of personnel
in
a Gas
Insulated
Substation.
ACKNOWLEDGEMENT
Theauthors wish to thank Mr. P.E.
Martin, Westinghouse
ElectricCorporation, East Pittsburgh, Pa., for his assistance inthe
Mr. D. Craghead, Southern
laboratory
setup
and
testing,
and
CaliforniaEdison
Companyand Mr. W. J. Melilli,Westinghouse
ElectricCorporation,Trafford,
Pa. fortheir assistancein the field
testing.
9
REFERENCES
D.
Evans,
Symmetrical
Components,
1. C. F. Wagner and R.
McGraw-Hill Book Co., Inc. New York, 1933, p. 144.
2. CentralStation
Engineers,Westinghouse
Electric corporation,
ElectricalTransmission
andDistributionReferenceBook,
East
Pittsburgh, Pa., 1950, p. 71.
Discussion
D.0. Craghead and A. Mateuchev (Southern CaliforniaEdisonCompany,Rosemead, Calif.): Theauthors havedeveloped in this paper
methods that provide valuable and timely guidance for the grounding of
Gas-Insulated Substations. Their efforts in verifying the adequacy of
the grounding scheme employedat ourVista 220 kV SF6 installation is
very much appreciated. Nevertheless, the usefulness of this papercan be
increased bydirect discussion of maximum touch voltages, which
provide a better measure of safety forpersonnel.
It should be noted,that
considerationoftheground
grid
potential rise., with due regard to polarities, may result in a higher touch
potential. As the workonstandards
for Gas-Insulated Substations
continue through the efforts of the IEEE Substation Committee WorkingGroup70.1,thelimitsoftouchpotentials
will eventually be
derived. Until a voltage level is agreed to for Gas-Insulated Substations,
the IEEE Guide for Safety in AlternatingCurrent Substation Grounding appears to be theacceptablesource
forthe150 voltlevel. A
subsequent investigation to determine the actual potentialvalue may be
proposed; however, it is not expected that the 150
volt level will be
exceeded at Vista.
In the text of the authors discussion, in the section pertaining to
10.1
internal faults, it is not specifically shown how the distance of
meters to the internal fault providing maximum outer sheath voltage
was derived. It should be remembered that a 14 - ground fault is being
considered. Thus, the current
divides in accordance with the
various
combinations of parallel paths available to the station ground. Therefore, 2.99 X 10-3 ohms equals 10.1 meters of sheath or is equivalent to
3.61 metersof sheath, a parallel combination of impedances in the
other two phase sheaths, horizontal ground straps and the single vertical
ground strap. Calculations confirm the authors figures reasonably well.
It does, however, seem that the ground straps which connect the center
cable feed to the station ground system have been neglected. From the
common sheath tie, there are three connections
to the station ground
mat. One exists through approximately 10 meters of strap and two are
throughthetwo
breakerframegrounds.Thelatterincludes
ap
proximately three or four meters of ground strap.
For simplicity, it may be assumed that these three parallel paths
couldaddpproximateiy
one third of (4.41 X l e ) X (10 meters) or
1.47 X 10- ohms. Recalculating, using the added impedance, relocates
the fault within the sheath toa point where the impedance to ground is
equal to 3.63 X 10-3 ohms. Therefore,the
voltagedeveloped
is
(3.63 X 10-3) X (25 kA) or 90.7 volts.
M. W. El Menschawi(BrownBoveri
& Cie AG,Mannheim,West
Germany):Theauthors
havediscussed grounding in a gas insulated
substation and shown a method of
calculating the maximum voltage
developed on the outer sheath of the system as well as a method to
derive the grounding points of a system. It is essential today with GIs,
high voltage subwhich is compactingconsiderablythelayoutof
stations, to take into consideration the voltage drop in case of faults
which could lead to new considerations of fencing. The question arises
if the earth mesh should not be extended outside of the fencedarea. In
case of indoor substationsbecause of the better earthshield possibilities
this problem can be coped with.
Theauthors have calculatedthecurrent
flowing ontheouter
enclosure in a range between 41-70% of theinner conductor current and
the test shows values between 30 and 60% which is approximately the
value expected in such stations.
Our tests show also values in the range of approximately 50% of
the inner conductor current. We are also of the opinion that the IEEE
Guide for calculating losses in isolated phase bus, ANSI C37.23, do not
apply for the compact GIS. We have made successful measurements in
indoor installations andit would be useful to continue the investigations
helping to clarify the matter and lead to suitable standards in the near
future.
Manuscript received February 18,1975.
4. CentralStation
Engineers,Westinghouse
ElectricalTransmission
andDistribution
Pittsburgh, Pa.,
paper
524.1950,
p.
ElectricCorporation,
Reference Book, East
1188
~~
that
However, I cannotfully agree withtheAuthorsopinion
currently there is no guidance established for determining the permissible sheath voltage to ground.
When deciding how good or bad it is to have the reported 73volts
between the enclosure and the station ground, it isuseful to recall that
not the contact voltage itself, but the amount of current perceived by
thehumanbodyduringthecontact
is whatcounts physiologically.
IEEEGuide No. 80 indicatesthatthetreshold
ofcurrent perception is about lmA, and a lack of muscular control may occur in the
9-25mA current range.
When the danger of loss of human life is considered solely as a
consequence of heart fibrillation, the Guide presently provides us with
a time dependent limit of body current, equal to the value of 165mA/
f l w i t h the time t in seconds.
No. 80, the well-known
In thecontemplated revisionofIEEE
Dalziels equation for the permissible electric shock energey constant
will be set equal t o the value
Consider the case of a man inside the GIS area, standing on some
thin ground surface layer, which can be a concrete floor, or anything
from a gassfrichsoil combination t o crushed stone, overlaying both the
foundations and the surrounding earth. Touching a groundedenclosure
or structure, heestablisheshimselfas
an alternate grounding system
consisthg of two series components;
1. Resistance of a human body, for hand-to-foot contact, assumed
to be
which will result in a somewhat lower limit of body current for a given
time, approximately 1 1 6 m A w
Incident y, this lowervalue has already been implemented in a
recent report prepared by the
Working Group on Electrostatic Effects
of Transmission Lines. As the report also includes detailed reasoning for
such a change, it is advisable to follow this reference for the meantime.
RBody =
Ohms
2.Resistance
of an unprotectedfoot
surface, approximated by
standing atthe
ground
measured between the touched enclosure andthe level of ground underneath the thin surface layer.
This last point deserves a bit of attention: In IEEE No. 80, the
resistanceofa
single unprotectedhumanfoot
is assumed as the
resistanceof ametallic circular plate of approximately 5 diameter
(r = 7 to 8 cm), buried at zero depth in homogeneous ground. If the
surface is covered by a layer of different material, the resistivity Ps is
equal t o the resistivity of that surface material. According to P. Lament,
the resistance of such aplate is
Ps
(0.28 to 0.32)
P 3
x (Ps)ohms
Based onsucha
premise, thecalculated
touch voltage limit
rise
implicitly represents the difference between the ground potential
UF, as carried by the contacted enclosure, and some average value Uo
of potentials beneath the first layer, which would be indicative of the
body circuit current flow through the boundary between the surface
layer and the earth.
As the earth becomessimultaneouslysaturated by thecurrents
emanating
from
the
other, permanent
grounding
electrodes,
the
resulting gradient field has aprimaryeffect
onthe value of Uo.
Typically, if the permanent electrode system forms a grid of few
conductors horizontally buriedin the earth, the saturation
is uneven
thepotentialsattheboundary
vary withdistance, as shown below,
Fig. 3.
Then, in this interpretation, the Guide considers the design of a
permanent grounding system adequate,if the formula (5) is satisfied for
the worst boundary potential EB-,
which is usually found above the
center of theperipheral mesh,i.e.
Etouch-hit = UF - uo; for u o = EBmin
Fig. 2
error.
includes
point,
certain
EB ( e
(kV)
distance (m)
(6)
Fig. 3
1189
Fig. 4a
Fig. 4b
Surface material
Resistivity
Body Current
Criterion
Ps
IB
For thisreason,theformula
( 6 ) may prove to be overly conservative. Furthermore, grounding of the reinforcing steel contained in
concrete foundations represents another, significant factor to consider.
Therefore, in my opinion, we must look into this matter in more
detail,
preferably
within the frarmw~rkof W.G. 69.2, aiming to
the
Of IEEE No. 8o in the next few ye**
At present, the typical touch voltagelimitscalculated for 30 ohmmeter, or for 3000 ohm-metersurface layers, are:
@ Touch Voltage L i m i t
1 sec.
0 . 2 sec.
0 . 1 sec.
3000 ohm-meter
Let-go l i m i t 16mA
88.0 V
196.9 V
278.5 V
(crushed stone)
F i b r i l l . lim.1169
638.0 V
1,427.0 V
2,619.0 V
30 ohm-meter
Let-go l i m i t
fl*)
16mR
F i b r i l l . lim.116*
REFERENCE
[ 6 ] IEEE Working Group Report, Electrostatic Effects ofOverhead
Transmission Lines, IEEE Trans.,
Vol. PAS-91,No.2, pp. 422426, March/April 1972.
16.72 V
121.2 V
37.4 V
52.9 V
383.6 V
271.2
rived. The distance was determined by solving the electrical circuit for
the specific fault location
whichcaused the internal fault current t o
divide equally as shown inFigure 4 of thepaper. When the current
divides equally (i.e., 25 kA in each direction), the maximum voltage is
developed on the outer sheath. The 10.1 meters is measured from the
center cablefeed to theouter breakertank (location ofmaximum
voltage) as shown in Figure 1 of the paper. The ground straps from the
center cable feed to the station ground were neglected by the authors
and the recalculation including these ground strapsby Messrs. Craghead
and Mateuchev is appreciatedandrepresentsamoreaccuratedetermination of the maximum outer sheathvoltage. Although the voltage of
these ground straps increases the maximum outer sheath voltage from
75 volts rms to 90.7 volts rms, the latter value remains well below the
150 volts rms level discussed in IEEE Guide No. 80.
1190
Since the bus runs in GIS normally have a height above ground of
several meters, a phase separation on the orderof a meter or more, and
overall lengths which are relatively short compared to isolated phase
bus; the authors agree with Dr. El Menschawi that ANSI C 37.23 on
Isolated Phase Bus should not apply t o compact GIs. It is the recommendation of the authors that the station ground should be extended
beyond the fenced area of the substation. Of greater importance is the
fact that all personnel within the substation area and
in contact with the
fence itself should be isolated (i.e., insulated) from the actual
earth
ground in order to prevent excessive voltages during an internal fault.
Mr. Sverak has stated thatguidance for determiningthe permissible
outer sheathvoltage is available. If Mr. Sverak is referring to themethod
ofdeterminingsuch
levels (i.e., IEEE Guide No. 80 or Dalziels
equation), the authors agree. However, if Mr. Sverak is referring to an
actual recommended voltage level, the authors disagree. Although the
methods of IEEEGuide No. 80 and Dalziels equation are generally
accepted, in order to determine a permissible body current the duration
&e., time or 2) of the electrical shock must be established. Further
the permissible touch voltagelevel isdependentonthisestablished
current as well as the body resistance and soil resistivity. It is upon this
voltagelevel
that uniformagreement
can notbe
established. Mr.
Sveraks tables illustrate the dependence of the touch voltage limit on
time and resistivity as no less than twelve voltages are given for various
shock durations and surfaceresistivities.
As a general worst case calculation of themaximum touch
voltage, a surface resistivity of zero ohm-meters and a body resistance
of 1000 ohms will be assumed. Using the body current equation given
in IEEE Guide No. 80 (165 mANT) and assuming a shock duration of
1 second yields:
1191