Earthing Calculation PDF
Earthing Calculation PDF
Earthing Calculation PDF
by Arun Balakrishnan
A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in
Electrical Engineering
Montana State University
© Copyright by Arun Balakrishnan (1990)
Abstract:
An important aspect of substation grounding system design is the calculation of the ground resistance
offered by the grounding grid, and the values of the mesh and step voltages on the surface. The values
of these parameters have to be kept within certain established Emits, keeping in view the safety of the
personnel present within and without, the substation area.
Simple expressions for calculation of these parameters for square and rectangular grids are already
available. Computer programs for designing substation grids, regardless of their shape, are available,
but tend to be tedious in usage when the grids are not symmetrical, since a large amount of data has to
be input to describe the grid configuration. Thus simplified expressions are preferred.
Another important factor in grounding grid design, is the footing resistance, which depends on the
resistivity and thickness of the surface rock layer. The footing resistance, at present, can be determined
by use of an infinite series, which is not very easy to evaluate. There is hence need for a simplified
finite expression for calculating the footing resistance. The value of footing resistance also affects the
safe values of the mesh and step voltages.
If two substation grounding grids are intertied by a bare conductor, when a fault occurs at one
substation, it is important to determine, by a simple analytical procedure, the effect of this fault current
on the other grounding grid, and the response of the system as a whole, to the fault current.
The objectives of this thesis may be summarized as follows: (i) To develop a new set of expressions,
which may be used for calculation of the ground resistance, and the mesh and step voltages, for all
practical shapes of grounding grids, once information regarding the grid configuration and soil
characteristics is available.
(ii) To develop a finite expression, for calculating the footing resistance in a substation.
(iii) To design amethod to calculate the total groundresistance of a system of two substation grounding
grids intertied by a bare conductor, as seen from the station where the fault occurs, and the ground
potential rise at both the substations.
The development of the expressions for the calculation of the ground resistance, mesh and step
voltages, footing resistance, and for evaluating the performance of intertied grids, is described in this
thesis. Through a comparison with existing expressions for the calculation of these parameters, the
considerable reduction in error through the use of the newly developed expressions is demonstrated.
The scope for future work in this area is also discussed briefly.
C A L C U L A T IO N O F IM PO R TA N T D E SIG N PA R A M E TER S FO R
G RO U N D IN G SY STEM S IN SU BSTATIO NS
by
Arun Balakrishnan
of
Master o f Science
in
Electrical Engineering
June 1990
/I W /
ii
APPROVAL
of a thesis submitted by
Arun Balakrishnan
This thesis has been read by each member o f the thesis committee and has been found to
be satisfactory regarding content, English usage, format, citations, bibliographic style, and
consistency, and is ready for submission to the College o f Graduate Studies.
,/W AW
Date Head, Major Department
ST A TE M EN T OF PE R M ISSIO N TO USE
In presenting this thesis in partial fulfillment o f the requirements for a master’s degree at
Montana State University, I agree that the Library shall make it available to borrowers under
rules o f the Library. B rief quotations from this thesis are allowable without special permission,
Permission for extensive quotation from or reproduction o f this thesis may be granted by
my major professor, or in his absence, by the Dean o f Libraries when, in the opinion o f either,
the proposed use o f the material is for scholarly purposes. Any copying or use of the material
in this thesis for financial gain shall not be allowed without my written permission.
Signature
Date 14-1 O .
Z
A C K N O W L ED G E M E N T S
Throughout the duration o f the author’s work at Montana State University, Dr. Baldev
Thapar was a constant source o f inspiration and encouragement. His support, accessibility, and
active participation, without which this work would not have been possible, are acknowledged
with gratitude.
The author is also grateful to Dr. Victor Gerez and Dr. Murari Kejariwal for their unstinting
support and encouragement. Their valuable suggestions made during the time this thesis was
written are appreciated. Mr. Prince Emmanuel is responsible for encoding parts o f an initial
version o f the program RESIS. His help during development o f the later version of the program
is acknowledged.
The Department o f Electrical Engineering at Montana State University, and the Bonneville
TABLE O F CO NTEN TS
Page
LIST OF TABLES vi
ABSTRACT viii
1. INTRODUCTION I
Program Description 14
Development o f the Equations 16
U se o f the Program 20
4. VOLTAGE CALCULATIONS 29
Basic Considerations 41
Ladder Circuit Solution 42
A n Approximate Solution to the Problem 44
Voltage Decay A long a Very Long Buried Conductor 46
REFERENCES CITED 53
vi
LIST OF TABLES
Table Page
LIST OF FIGURES
Figure Page
A B ST R A C T
A n important aspect o f substation grounding system design is the calculation o f the ground
resistance offered by the grounding grid, and the values o f the mesh and step voltages on the
surface. The values o f these parameters have to be kept within certain established Emits, keeping
in view the safety o f the personnel present within and without, the substation area.
Simple expressions for calculation o f these parameters for square and rectangular grids
are already available. Computer programs for designing substation grids, regardless o f their
shape, are available, but tend to be tedious in usage when the grids are not symmetrical, since
a large amount o f data has to be input to describe the grid configuration. Thus simplified
expressions are preferred.
Another important factor in grounding grid design, is the footing resistance, which depends
on the resistivity and thickness o f the surface rock layer. The footing resistance, at present; can
be determined by use o f an infinite series, which is not very easy to evaluate. There is hence
need for a simplified finite expression for calculating the footing resistance. The value o f footing
resistance also affects the safe values of the m esh and step voltages.
U two substation grounding grids are intertied by a bare conductor, when a fault occurs
at one substation, it is important to determine, by a simple analytical procedure, the effect of
this fault current on the other grounding grid, and the response o f the system as a whole, to the
fault current.
The objectives o f this thesis may be summarized as follows:
(i) To develop a new set o f expressions, which may be used for calculation o f the ground
resistance, and the mesh and step voltages, for all practical shapes o f grounding grids,
once information regarding the grid configuration and soil characteristics is available.
(E) To develop a finite expression, for calculating the footing resistance in a substation;
(Ei) To design am ethod to calculate the total groundresistance o f a system o f two substation
grounding grids intertied by a bare conductor, as seen from the station where the fault
occurs, and (lie ground potential rise at both the substations.
The development o f the expressions for the calculation o f the ground resistance, mesh and
step voltages, footing resistance, and for evaluating the performance o f intertied grids, is
described in this thesis. Through a comparison with existing expressions for the calculation o f
these parameters, the considerable reduction in error through the use o f the newly developed
expressions is demonstrated. The scope for future work in this area is also discussed briefly.
I
C H A PTER I
IN T R O D U C TIO N
(1) To carry electric currents to ground under normal and fault conditions, without
service;
(2) To ensure that a person in the vicinity o f the grounded facilities is not exposed to the
The ground return circuit is defined [1] as the circuit in which the earth or an equivalent
conducting body is utilized to allow current circulation from or to its current source. The ground
or som e conducting body o f relatively large extent that serves in place o f the earth. A system
or apparatus is provided with ground for purposes o f establishing a ground return circuit and
Whenever a fault occurs in a substation, the earth becomes saturated with currents flow ing
to ground from the grounding grid and other ground electrodes buried below the surface o f the
earth. The substation grounding grid is a system o f horizontal, interconnected bare ground
conductors, spaced a few meters apart, buried in the earth at a depth varying from 0.25 to 2.5
meters, which provides a common ground to all the equipment present in the substation.
A ll grounding systems offer a finite resistance (to remote earth) to the fault currents. It is
very important to have a very low value o f this resistance, referred to as the ground resistance
2
o f a substation grid, which is preferably about I ohm for large substations, and between I o h m
and 5 ohms for distribution substations. W hile under normal conditions, the grounded apparatus
in a substation operates at near zero ground potential, under fault conditions, the portion o f the
fault current that is conducted into the earth by the grounding grid causes the potential o f the
grid to rise with respect to remote earth. This rise in voltage of the grounding grid is referred to
as the ground potential rise (GPR), and is proportional to the magnitude o f the fault current, and
Two other important design parameters, which depend upon the value o f the ground
potential rise and the ground resistance o f the grounding grid are the m esh voltage and the step
voltage. The conductors o f a grounding grid divide it into a number o f m eshes, and the potential
on the surface o f the earth, above the grounding grid, is not the same at all points. The touch
voltage is defined as the potential difference between the ground potential rise, and the surface
potential at the point where a person is standing, while being in contact with a grounded structure.
The maximum value o f the touch voltage within the substation yard is known as the mesh
voltage. The step voltage is defined as the difference in the surface potential experienced by a
person, with his feet one meter apart, with no part o f his body being in contact with any grounded
structure.
The human body, for dc, and ac power frequencies (50 and 60H z), can be represented as
a finite noninductive resistance, o f value in the range 500 to 5000 ohms, and is usually
approximated by a value o f 1000 ohms [1]. There is hence a critical lim it to the amount o f shock
energy that can be absorbed by a human being. Crossing this limit, which varies from person
to person depending upon factors including the body weight, can prove to be dangerous. The
m ost comm on physiological effects o f electric current on the body are perception, muscular
contraction, unconsciousness, fibrillation o f the heart, respiratory nerve blockage, and burning.
The fibrillation threshold, which can prove fatal, is reached when currents o f magnitude 60 -
3
100 niA are reached. The duration for which the fault current can be tolerated by m ost people
is given as [1]
(Jb) = (11)
where I8 is the rms value o f current, in amperes, flow ing through the body, ts is the duration of
the current flow , in seconds, and S8 is the an empirical constant related to the electric shock
Based on studies conducted by Dalziel [2], a current I8 that can be tolerated by 99.5% of
Here, ^[s^ has a value o f 0.116 based upon the fact that the shock energy that can be survived
by 99.5% o f people weighing 50 kg, results in a value o f S8 of 0.0135. Thus for a 50 kg body
weight,
/ B = 0 .1 1 6 / ^ (1.3)
4 = 0 .1 5 7 /^ 4 (IA)
Since the above equations are based on tests for the 0.03-0.3 seconds time range, they are
not valid for very short or long times. Ferris, e ta l [3], suggest a value o f IOOmAas the fibrillation
threshold if duration o f the shock is not specified. Beigelm eier [4] has suggested values o f 500
m A and 50 hiA respectively, for shock durations o f less, and more, than one heartbeat duration.
It is thus important that the grounding system be designed in a manner to keep the shock currents
below the value mentioned above. The resistance o f the human body, R8, is assumed to be about
1000 ohms for defining the limits on the step and touch voltages.
Substations usually have a layer o f crushed rock spread on the soil. This usually provides
a high resistivity layer below the feet o f personnel in the substation. A highly simplified approach,
which neglects the mutual ground resistance between the two feet o f the person, and assumes
4
a very large depth o f the crushed rock layer, gives the value o f the footing resistance (which is
defined as the resistance o f the ground below the feet), Rfoon as 3ps, where ps is the resistivity of
the crushed rock layer. This expression is derived by modelling the foot as a conducting disc
o f radius 8 cm. In accidental circuits for mesh and step voltages, the two feet are in parallel and
series respectively; therefore, the total footing resistance is taken to be LSp5 and 6ps. The layer
o f crushed rock greatly increases the contact resistance between the feet and the substation
surface, which greatly reduces the current flow ing through the body o f a person present in the
substation.
The effect o f the layer o f crushed rock spread on the surface o f the substation area on the
value o f the fault current flow ing upwards into the body depends on its thickness, the relative
resistivity o f the crushed rock layer and the lower soil, and the resistance o f the foot.
The safety o f a person depends on prevention o f his absorbing the critical amount o f shock
energy before the fault is cleared and the system re-energised. The driving voltage o f any
With R b = 1000i2, and Rfoot = 3ps, equation (1.5), for a 50kg body weight, can be written as
Again, with R b = 1CKXX2, and Rfoot = 3ps, equation (1.7), again for a 50kg body weight, can be
written as
The factor C5 is included to account for the layer o f crushed rock which is spread on top
o f the soil in the substation area, and is o f resistivity different from that o f the soil. It is a factor
5
applied for compensating for the finite thickness o f the surface crushed rock layer. This factor
and its calculation form an important aspect o f substation grounding grid design, and are dis
cussed in detail later. For a uniform soil, with no upper crushed rock layer, the value of Caequals
unity. pais then the resistivity o f the soil in iQ-m, and tais the duration o f shock current in seconds.
The calculated value o f mesh voltage should not exceed the maximum allowable touch voltage,
Etouch. Similarly, the calculated value o f the step voltage should not exceed the maximum
allowable step voltage, Estep. The values OfEatep^ and Etouch(50) are calculated assuming the body
Calculation o f the mesh and step voltages, and the ground resistance, as part o f a design
m ethodology, can be made using different methods. A number o f computer programs have been
developed for this purpose, and give fairly accurate results. Unfortunately, programs are not
always available in field sites. A lso, preliminary design strategies do not require the high
accuracy offered by these programs. It is hence important to have a simplified set o f expressions,
M ost o f the work so far has concentrated on grids which are o f shapes approaching a
square or a rectangle. Though simplified equations for making calculations o f the ground
resistance, and mesh and step voltages for such grids have been developed, and have been in
use for som e time, not all grids can be approximated by a square or rectangle, and for the design
o f such grids, the use of the existing expressions leads to erroneous results.
The follow ing site parameters are known to have a substantial impact on the grid design:
maximum fault cu rren t^ ), fault duration^), soil resistivity(p). The substation grid design
depends on the area o f the grounding system, the conductor spacing, the depth o f burial o f the
The objectives o f this thesis are to develop a new set o f expressions, which can be applied
(i) To develop a new set o f expressions, which may be used for calculation o f the ground
resistance, and the m esh and step voltages, for all practical shapes o f grounding grids.
(ii) To develop a finite expression, for calculating the footing resistance in a substation,
(iii) To design amethod to calculate the total ground resistance o f a system of two substation
grounding grids intertied by a bare conductor, as seen from the station where the fault
The results o f the research covered in this thesis have been presented in references [5],
Chapter I has thus far given a brief introduction to the areas o f substation grounding with
which the topics covered in this thesis are related. Som e work regarding calculation o f the
ground resistance, mesh voltage, step voltage and footing resistance is available in literature,
Chapter 2 covers the algorithm o f the program RESTS which has been used as the primary
development reference for the work involving the calculation o f ground resistance and mesh
and step voltages reported in this thesis. Relevant equations for developing the resistance matrix
In Chapter 3, the calculations involving the grounding resistance have been described,
and a comparison o f the new expression for ground resistance, presented in this thesis, has been
made with expressions already available in literature. The accuracy o f the simplified expressions
Chapter 4 covers the development of the expressions for the calculation o f mesh and step
voltages for grids o f irregular shapes. A comparison o f the new expressions with the methodology
Chapter 5 describes the development o f the expression for the calculation o f the footing
resistance. In addition, two fairly accurate analytical models to calculate the footing resistance
have also been presented. A comparison o f the new expression with the expressions recom
mended by the IEEE Std. 80-1986 is made, and the accuracy o f the new expression and the two
m odels demonstrated.
Chapter 6 provides details about the method for determining the performance o f two
substation grids interbed by a bare, buried conductor, when a fault occurs in one o f the sub-
stadons. From the suggested method, the total ground resistance o f the system, as seen from the
stabon where the fault occurs, and the ground potential rise at both substabons can be calculated.
A simplified method, derived using the transmission line equabons, is discussed in detail.
The thesis is concluded with a brief summary o f the work done, and its benefits, in Chapter
Literature Survey
This section gives a brief summary o f the expressions already available for the calculabon
o f the ground resistance, mesh and step voltages; and the footing resistance, and the performance
A minimum value o f the ground resistance offered by a grid buried in soil o f uniform
resisdvity, can be obtained from the expression for the resistance o f a circular plate, which may
(1.9)
where Rg is the resistance to ground o f the substabon grid in ohms, p is the resisdvity of soil in
The upper lim it to the ground resistance o f substation grids with square shapes can be
obtained by use o f the formula given by Laurent [9], and Niemann [10]:
( 1. 10)
The factor p/L accounts for the fact that the resistance o f a solid plate is always lower than
that o f a grid o f the same shape, composed o f a finite number o f conductors. A s can be seen
from the formula, the difference in resistance reduces, and approaches zero as the total length
o f the conductors gets infinitely large. W hile these expressions may be used with reasonable
accuracy for square grids buried in the soil at depths less than 0.25 meters, for grids buried
between a minimum o f 0.25 meters and a maximum o f 2.5 meters, Sverak [11], introduces a
correction factor to account for the variation in the depth o f burial, and develops the expression
( 1 . 11 )
The resistance of a grid, embedded in a soil o f uniform resistivity, can also be calculated
by use o f the expression presented by Schwarz [12]. Schwarz’ expression is valid for multi-grid
systems, and may be simplified for a single grid, without ground rods, in a soil o f uniform
resistivity p, as
( 1. 12)
where a * = ^2rh, for conductors buried at a depth o f h meters, and r is the radius o f the conductors
in meters.
A limitation o f the expression proposed by Schwarz is that the factors K1 and K2 can be
obtained only by use o f graphs, which prevents an analytical solution using a computer. This
9
deficiency may be overcome by use o f the expressions for K 1 and K2 derived by Kercel [13].
ab A a + ^ a 2+ b 2 I 1 b + ^ a 2+ b 2
K1=1.84- + —In
- 2 [a I b J b II 4 J
(1.13)
3b2 3 c 2 3aV
where b is the length o f the long side o f the grid, and a is the length o f the short side of the grid
in meters. These expressions, and the graphs, can only be applied to a lim ited set o f rectangles,
Simplified equations have also been proposed for the calculation o f the mesh and step
voltages. The best expressions available so far are those developed by Sverak, and recommended
The m esh voltage, for square grids with square meshes, is calculated as
plaKmK-i
(1.15)
L
where Ig is the maximum grid current that flow s from the grid to ground in amperes.
I D2 (D+2hf h
+ — In----------- (1.16)
2n 1 6 M + SDd Ad Kh n(2n - 1)
D is the spacing between parallel conductors in meters, h is the depth o f burial of grid conductors
in meters, and d is the diameter o f grid conductors in meters, and n is the number o f parallel
conductors in any one dirction. K1, Kji, and Khare correction factors, used to compensate for the
10
fact that the actual grids are different from the model, which is based on a system o f n parallel
conductors in one direction. The actual grids have conductors in two directions, and these
The corrective weighting factor, Kii, that adjusts the effects o f inner conductors on the comer
m esh is given as
K il = 1/(2« f ” (1.18)
Kh, the corrective factor that emphasizes the effects o f grid depth, is given as
Kh= - ^ l + h (1.19)
n P^aKsKi
( 1. 20)
' L
( 1.21)
s + z r b + ^ 1 - ° - 5“" )
For rectangular grids with square meshes, the value o f n is m odified as follows:
n = V«a« b (1.22)
where nAis the number o f parallel conductors along one o f the coordinate axis, and nBthe number
These equations are recommended for use within the follow ing limits:
A ll the expressions given above are recommended for use with grids with shapes which
may be approximated by a square, or a rectangle with length to width ratio limited to less than
8 : 1.
Expressions are also available for the calculation o f footing resistance [I]. They may be
summarized as:
£
' d.25)
Ii
F(K) = 1 + 2 £ Q (1.26)
M = I
Q = K nl [ \ + Q n X f f 2 (1.27)
^ = (P -P J/(P + PJ (1.28)
where p3 is the resistivity o f the surface crushed rock layer in ohm-meters, p is the resistivity
o f soil in ohm-meters, X = h/b, h is the thickness o f the crushed rock layer in meters, b is the
equivalent radius o f a foot equalling 8 centimeters, and K is the reflection factor. Figure I shows
the variation o f F(X) with h, with p/ps varying from 0.005 to 0.5.
Though the value o f footing resistance may be calculated from the above expressions, due
to the infinite nature o f equation (1.26), they are difficult to use without the help o f a computer
or programmable calculator.
Seedhar, Arora and Thapar [14] have proposed a finite expression to compute the potential
at any point due to a point current source located anywhere in a two layer soil. A finite expression
for calculating the footing resistance may be derived from this expression.
12
Sverak has also proposed a simplified formula for calculating the footing resistance, which
may be given as
(1.29)
C (X )= 1 -0 .1 0 6 l ~p. (1.30)
2/1,+ 0.106
S 0.6 -
Sobralj e ta l [15] have developed a method for studying interconnected grounding systems,
connected by overhead ground wires. In their analysis, the mutual effect o f the interconnected
grids are neglected, and the system is assumed to comprise o f a number o f grids, and overhead
ground wires, which transfer the potential o f the energised grid to other connected grids.
14
CH A PTER 2
TH E PR O G R A M RESIS
Program Description
With a view to have an available reference, with which any newly developed expressions
may be compared, a computer program RESTS, based on the finite element analysis, was
developed. This program can handle grid configurations composed o f straight linear conductors,
laid in three mutually perpendicular directions. For a given grid configuration, the grid con
ductors are divided into small straight segments, each about 5 meters in length. The self and
mutual resistances of these segments are then calculated to obtain the ground resistance o f the
grid.
For grids with a very large number of parallel conductors, the voltage drop may be
neglected due to the lower value o f current flow ing per conductor. For large grids with a small
number o f conductors, it becomes necessary to express the voltage on each segment as a function
o f the applied voltage, and the voltage drops due to the leakage currents flow ing through the
resistance and inductance, and mutual inductances o f the conductor segments. Since practical
grids are rarely o f this type, in the program RESTS, the conductors are assumed to be at the same
potential. This is a valid assumption, as at power frequencies, the resistance and inductance o f
the segments are negligible when compared to the resistance between the segments and the
earth. A t power frequencies, the voltage drop along the conductors becom es important only if
Whenever a fault occurs in a substation, a part o f the fault current flow s to the ground
through the grounding grid. This fault current flow ing through the grounding grid causes the
potential o f the grid, with respect to the remote ground, to rise. This potential rise of the conductor
segments is also known as the ground potential rise (GPR) o f the grid.
A part o f the fault current flow ing through the grid, is dissipated by each o f the conductor
segments. It is thus important to determine the current distribution from the segments o f the
grid conductors. Once the current distribution has been determined, the voltage at any point on
the surface o f the earth is determined by obtaining the sum o f voltages at that point due to the
current dissipated by each o f the segments. In this fashion, the voltage profile in any direction
Once the voltage profile is available, the m esh and step voltages can be computed. For a
given design, based on the configuration o f the grid conductors, the approximate area within
which the point where the minimum potential is likely to occur, can usually be identified. Within
the specified area, the potentials at a finite number o f points are determined, and the minimum
o f these values, and its location, are identified. The difference between the GPR and this
minimum voltage, gives the value o f mesh voltage for the specific configuration.For grids with
irregular shapes, it is likely that the point o f minimum potential may not be readily identifiable,
and a larger area, within which this point is likely to occur, may have to be chosen. The step
voltage is calculated as the difference in voltage between the two points on the surface o f the
ground, one directly above a comer o f the grid, and the other one meter away in the direction
o f the diagonal o f the comer mesh extended outwards. In case o f grids with irregular shapes,
the step voltage at all the comer points may be determined, and the highest of these values
The equation for the calculation o f potential at a point, and subsequently the equations
for the calculation o f the self and mutual resistances o f the various conductor segments are
derived on the basis o f the method of images [15]. This is done to account for the boundary
The potential at any point (x,y,z), produced by a linear conductor m parallel to the x-axis,
Similar equations can be written for conductors oriented along the y-, and z-, coordinate axes.
The nomenclature used in equation (2.1) above, and in equations (2.2), (2.3), and (2.4)
The self ground resistance Rmm, o f a conductor segment, is defined as the ratio o f the
voltage on the segment to the current flow ing out o f the segment, neglecting the effect o f other
conductor segments, and is determined by calculating the average potential on the surface of
( 2. 2)
* - = ^ * < 2‘ toy - 2 + t o i ^ r + f - T >
This equation includes the effect o f the image of the conductor segment under consideration.
The mutual ground resistance, Rmn, between two linear conductor segments m and n, is
determined by calculating the average potential on the surface o f conductor segment m, when
conductor segment n is discharging a unit current. To account for the effect of the image seg
ments, the computer is programmed such that the equation is applied twice, once for the two
17
segments, and once for one o f the segments and the image o f the other. The sum o f these values
gives us the total mutual ground resistance between the two conductor segments. The mutual
resistance o f segment m, as seen by the segment n is equal to the mutual resistance o f segment
a2+ 4h2= s2
The effect of the im age conductors may be taken into consideration by using the same
expressions with one modification. Since the image conductors are not on the same plane as the
source conductors, for calculation o f the mutual resistances, the distance separating the ith
conductor segment, and the image o f the jth conductor segment, along the z-axis, is equal to
18
tw ice the depth o f burial o f the source conductors. The distance between the i'hconductor segment
and the im age o f the.jth conductor segment, along the other two coordinate axes, is the same as
the distance between the i,h conductor segment and the j ,h conductor segment. This can be easily
When the two conductor segments are parallel to the x-axis, the value OfRinn is given as
If the two segments are parallel to the y-axis, the value o f Rmnis obtained by interchanging
If the two segments are orthogonal, and parallel to the x- and y- axes respectively, the
(b r+ lQ th + kd U h+ k^+ K )
2 (^2 + ^ 7) (62+ ^8) 2 (ti2+ /:7) ( t i 2+ /r8)
T T + ^k6)*
k9(ax T T T+ /r4)
kw(a2 ^
. - x jh + a A . _i s2+ a2k%
(2.4)
Sm + k2) S m kw{a2+ k^)
A resistance matrix R, composed o f the self and mutual resistances o f the various conductor
segments, is formed, and for a grid with j linear segments, is o f order (j X j). A lso, the sum o f
the currents being dissipated by the various conductor segments of the grid is equal to the fault
current, Ig . This is the (j+ l)th equation, and the system is o f order (j+1). The system is to be
solved for the current components of the fault current being dissipated by the various conductor
segments, and the voltage to which each conductor segment is raised due to these current
components. The voltage on a conductor segment is the sum due to the leakage current flow ing
through it, and the leakage currents flow ing through the other segments.
19
(2.5)
Rj2 Rji -I 0
Rn • Rjj Ij
I I . . I 0 , W
The potential at a point (x,y,0), on the surface o f the earth, is the sum o f the potentials at
that point due to the currents dissipated by the various conductor segments, and their images;
The potential at the point (x,y,0), due to each o f these segments, can be determined by use o f
equation (2.1). This equation can also be used to calculate the m esh voltage at a point, by
calculating the potential at a finite number o f points on the surface, and choosing the point with
the minimum potential. The difference in potential between the ground potential rise, and this
value, gives the value o f the m esh voltage. The step voltage may be calculated as the difference
in potential between two points on the surface of the earth, which are one meter apart.
A s mentioned earlier, the value o f the m esh and step voltages should be less than die
U se o f the Program
The values o f ground resistance, mesh voltage and touch voltage, obtained with the
program RESTS, were compared with those available in literature. Since m ost o f the available
analysis has been done for square shaped grids, only results for such grids are compared in Table
2.
It may be observed that the results from RESIS compare very w ell with those obtained
from other sources. Since m ost results available in literature are for square grids, only results
for such grids are compared. Once it was established that the results from RESIS were quite
accurate, calculations were made for more than one hundred and fifty grids o f different con
figurations and sizes. The shapes o f the grids, and the dimensions varied are shown in Figure
2.
TABLE 2. GROUND RESISTANCE, MESH AND STEP VOLTAGES FOR SQUARE GRIDS
Depth of burial of grid = h , meters.
Diameter of grid conductors = 0.01 meters.
Number of meshes in the grid = n.
Resistivity of soil = 100 ohm-meters
SGSYS: Reference [17]
J-P Reference [18]
Grid Size n h Ground Resistance Mesh Voltage Step Voltage
(m X m ) (ohms) (As a percentage of GPR)
RESIS SGSYS J-P RESIS SGSYS J-P RESIS SGSYS J-P
10X 10 I 1.5 4.97 5.09 45.52 46.89 ..
8.19 7.28
20X 20 4 1.5 2.45 2.49 ——
32.69 32.82 ___
7.25 6.49
30X 30 9 1.5 1.61 1.62 — 28.08 27.80 — 6.61 5.91 —
40X 40 16 1.0 1.21 1.23 1.40 26.13 26.16 25.00 8.40 7.55 7.50
50X 50 25 1.0 0.96 0.97 1.10 24.25 24.19 24.00 7.90 7.10 6.50
60X 60 36 1.0 0.79 0.80 0.85 22.84 22.94 22.00 7.50 6.39 6.00
70X 70 49 0.5 0.69 0.69 0.75 23.24 23.40 25.00 11.32 9.57 12.00
80X 80 64 0.5 0.60 0.60 22.27 22.41 10.87 9.20 —
During these computer runs, it was observed that once the grid shape and size are fixed,
the ground resistance and mesh and step voltages did not vary much with change in the radius
o f the conductors used, for the range o f conductor sizes (2.5 millimeters - 1 0 millimeters) usually
used in such grids. This is shown in Table 3. The radius of the conductors was thus fixed at an
average value o f 5 millimeters (AWG 4/0). The grid depth was varied from 0.25 meters to 2.5
SQUARE
TRIANGLE
C H A PTER 3
G R O U N D R ESIST A N C E C A LC ULATIO NS
Since Sverak’s formula is simple to use, and gives reasonably accurate results for square
grids, it was decided to use it as the basis for any modification and improvement. The basic
considerations which were kept in view when the formula was m odified are:
(ii) A factor to account for the variation in shape needs to be identified. One such
dimensionless factor is -^AJLpt which gives a measure o f the spread o f the area. It has
a maximum value o f 0.282 for a circular area, and decreases as the length o f the spread
o f the area increases. The ground resistance o f a conductor o f given surface area
decreases as the length over which the area spreads is increased. For example, the
ground resistance o f a circular plate is more than the ground resistance o f a strip
conductor o f the same surface area. A s the number o f meshes in a grid o f given
(iii) A number o f expressions, as a function o f ^[XlLp were tried, and the expression
[2 In(LpV^ZA) - IP(VAZLp) was found to give the best fit with the results obtained from
RESIS.
A comparison o f the expression which thus evolved, with those for a strip conductor and
a circular plate showed the newly developed expression to be in error for the two cases by factors
o f 1.437 and 1.594. A factor o f 1.52 was chosen as the average of these figures, and the new
24
expression was finally m odified by a factor o f 1.52, to give a new equation for the calculation
o f ground resistance as
A comparison o f equation (3.1) with existing formulas for the calculation of the ground
resistance o f grids o f various shapes, is given in Table 4. The values obtained through the use
o f Sverak’s expressions, Schwarz’ expressions, and equation (3.1), are compared with the
accurate values obtained through the use of the program RESIS. The follow ing observations
(i) Sverak’s formula is quite accurate if the shape o f the grid is a square. For grids with
other shapes, the error is much higher, and in a few cases, is as high as 40%.
(ii) Schwarz’ formula gives better results for square, and rectangular grids, but gives high
(iii) The errors obtained from use o f equation (3.1) are consistently below 10%.
To compare the results obtained from use o f equation (3.1), with scale model test results,
analog m odels o f grids o f the shapes under consideration were made, and their resistances
determined. The m odels are made o f styrofoam o f about 2cm thickness, are covered with alu
1. Square.
Three m odels o f each shape were tested in an electrolytic tank, with water o f resistivity
66.2 ohm-m used as the electrolyte. The ground resistance o f each m odel floating on the surface
o f the water was determined. The area o f each m odel in contact with the water was 3 14.2sq.cm.
The ground resistance in each case, as determined from equation (3.1), with Lt = °°, and
h = 0, is given in the second column of Table 5. It may be observed that the difference in the
calculated and measured values is not more than 8 % in any o f the cases. Since the models tested
represent plates o f the same shape, it may be concluded that equation ( 3 . 1 ) may be even applied
to grids having a very large number o f parallel conductors. This is also proved analytically in
For a given area, a circular plate has the minimum peripheral length, and a strip with
negligible width (as compared to its length) has the maximum peripheral length. For a given
area, the circular plate gives the maximum ground resistance, and the strip, the m inim um . These
two shapes, can thus be assumed to represent the extremes in shapes as far as the ground resistance
is considered. The ground resistance o f a circular plate o f radius r placed on the surface o f the
ground is given as
R = p/(4r) (3.2)
26
while the resistance to ground o f a strip o f length o f length L and width W, again placed on the
Lf = oo
Il
(3.4a)
¥
h =O
Il
(3.46)
/?=p/(4.15r) (3.5)
A comparison o f equations (3.2) and (3.5) shows that equation (3.5) gives a ground resistance
A -LW Lf = oo (3.6a)
(3.7)
0.947tL lnF - 1
A comparison o f equations (3.3) and (3.7) shows that equation (3.7) gives a ground resistance
Thus, from the data in Table 5, and from the above analysis, it may be observed that the
error for m ost shapes o f grounding grids is less than 10%. Equation (3.1) may thus be used to
C H A PTER 4
VO LT A G E CA LC ULATIO NS
Equations (1.9), (1 .10), (1 .14), and (1.15) are based on a model comprising equally spaced
infinitely long, parallel conductors with no cross connections, dissipating uniform current.
Correction factors Ki, Kii, and Kh are used to account for the difference in the actual grids and
the model. Ki, as given in equation (1.11), is based on the data obtained by Koch, from his
limited experimental work [I]. This value is recommended for use for grids with the number o f
conductors along one direction not to exceed 25, but it was determined that a better correction
factor may be obtained from a different expression for Ki. From the test cases for which RESlS
was run, more than forty examples for square and rectangular grids were selected, and a new
expression for Ki was derived from plotting the values o f Ki obtained from the expression
(4.1)
The best linear fit for Ki was obtained with the expression
Ki = 0.644 + 0.148*« (4 .2 )
The number o f parallel conductors in one direction, n, is one o f the factors used in the
simplified equations. The choice o f the factor n is simple for square grids, as the number of
conductors is the same in both directions if the number of meshes are assumed to be the same.
For rectangular grids, the IEEE Std. 80 recommends the use o f equations (1.22) and (1.23). The
value o f n obtained from this method gives reasonably good results, provided, the length to
30
width ratio is limited to less than 8:1. Such a strategy is not applicable to irregular shaped grids
(i) The number o f parallel conductors in one direction may be different from that in the
other direction.
(ii) The length o f all the parallel conductors in one direction may not be the same.
It is thus necessary to formulate an expression for n, which should be applicable for all
shapes o f grids. Such a expression would have to be a function o f the shape of the grid, its
dimensions, and the length o f conductor used in the grid. Unlike square grids where the total
length is an integer multiple o f the length o f one side, for irregular shaped grids, other dimensions
Dm, the maximum distance between any two points on the grid periphery;
U sing the data generated using the program RESIS, the follow ing expression, as a function
n=a*b*c*d (4 .3 )
where
a = TLJLp (4.3a)
b = ( L p/4-\[A)m (4.3b)
(4.3c)
The dimensionless expression for n given in equation (4.3), reduces to the number o f
parallel conductors in any direction for square grids having square m eshes. The factors b, c, and
d reduce to unity for square grids. The factors c and d reduce to unity for square and rectangular
grids. The factor d equals unity for square, rectangular and L shaped grids.
A comparison o f the method described in the IEEE Std. 80-1986, for the calculation o f
m esh and step voltages, as outlined in Chapter I, with the same method using the new
(i) f ° r mesh voltages, the new expressions give lower errors than the TEFR Std. 80
expressions, and are within 16% o f the actual values obtained using RESTS. Som e o f
the IEEE Std. 80 calculations for the m esh voltage are in error by as much as 50%.
(ii) for step voltages, the new expressions give lower errors than the TERR std. 80
expressions, and are within 30% o f the actual values obtained using RESTS. Som e the
IEEE Std. 80 calculations for the step voltage are in error by more than 100%.
Since the expressions for Ki have been optimized keeping the m esh voltage calculations
in view , the use o f the same expressions for the step voltage results in a slightly higher error.
This error in the calculation o f the step voltage is acceptable, as once a design has been made
safe for m esh voltage considerations, it will usually be safe from the step voltage point o f view .
33
C H A PTER 5
The footing resistance, which is defined as the resistance o f the ground below the feet of
a person present in a substation, is affected by the layer o f crushed rock spread on the surface
o f the soil. This crushed rock layer is usually IOcm - 20cm thick, and provides a high resistivity
layer. A s an approximation, the footing resistance is taken as 3ps, where ps is the resistivity o f
the crushed rock layer. But this expression is arrived at based on the assumption that the layer
o f crushed rock is o f a very large thickness, by virtue o f which the effect o f the boundary at the
rock-soil interface is neglected. The expression is derived by considering the foot to be a con
ducting disc o f radius 8 cm, and neglecting the mutual resistance between the two feet. In the
accidental circuits for step and mesh voltages, the total footing resistance is approximated to be
The expressions for a more accurate calculation o f the footing resistance recommended
by the IEEE Std.80 (equations 1 .25-1.28), take into consideration the mutual resistance between
the two feet. The effect o f the thin layer of crushed rock on the top o f the soil is dealt with by
using the two layer model o f soil. W hile the expression derived on the basis o f the above
grammable calculator. Thus, a simple set o f finite expressions for calculating the footing
Equations (1 .26) and (1.27), are used in calculating the footing resistance. Equation (1.26)
is given as
F(X) = I + 2 X Q (5.1)
n=1 x z
Q = K n/ [ I + ( I n X f f 2 (5.2)
Kn
F(X) = I + :+ 2 Z (5.3)
[l + ( 2 X f f M=2 [I + (2nXf]vl
The infinite series expressions for potential due to a point current source in a two layer
soil, when both the source S and the point P where the potential is desired, are in the upper layer,
V 1 + *-■
471/1 ^ d 2
h +(ds - d pf ^jd2h +(ds + d pf
” i I
+ I K n*( . -- +
"“1 + (2 « —ds —dpf 'fdf, + (2 zi + d s —dpf
where I is the current dissipated by the point source S, P 1 is the resistivity o f the upper layer o f
the soil, p 2 is the resistivity o f the lower layer o f soil, h is the depth o f the upper layer, ds =
(depth o f S from the ground surface)/h, dp = (depth o f P from the ground surface)/h, dh =
(horizontal distance between S and P)/h, and K is the reflection factor. AU the distances have
been normalized with respect to the depth o f the upper layer o f the soil. Equation (5.4) is the
basis for the development o f an empirical formula to represent the terms under the summation
sign in equation (5.3). Seedhar, Arora and Thapar have suggested the following equivalence
~ 2Kn _ I
[ -K —ln(l (5.5a)
"=2 [I + (2riX)2]112 13 + (1/X)2
35
The multiplying factor under the square root sign is approximately equal to one for most practical
2K I
H { X ) = l + : j T 7 ^ ~ x [ K + H X ~ K)] {55b)
where K is the reflection factor given as K = (p, - p)/(p„ + p), p, is the resistivity of crushed rock
layer in ohm-meters, p is the resistivity o f soil in ohm-meters, X = h/b, h is the thickness o f the
crushed rock layer in meters, and b is the equivalent radius o f foot equalling 8 cm. Calculations
for F(X) and H(X) were made for the following range o f usually encountered values:
In this model, as shown in Figure 3, the surface layer o f crushed rock is replaced by a
truncated circular cone having the same resistivity as the crushedrock. The height o f the truncated
cone is equal to the thickness o f the crushed rock layer, and the radius o f the top o f the truncated
cone is equal to the equivalent radius o f a human foot, which is usually approximated as 8 cm.
The angle Gj, is so selected, that a foot on top o f the cone has the same footing resistance as it
The resistance offered by the soil Rs, is calculated by assuming the bottom of the cone to
The footing resistance from this model is determined by adding the resistance o f the
(5.9)
o f the expressions for A (X ) and F(X), from equations (5.3) and (5.10). Figure4 gives the variation
in G1 with h, for various ratios o f p/ps. The values o f A (X ) computed for the range o f parameters
specified in equation (5.6), are given in Table 7. The errors are once again calculated to be less
than 4%.
37
65 -
64 -
62 -
59 -
57 -
56 -
55 -
54 -
The truncated cone model is shown in Figure 5. The soil and the crushed rock layer are
represented by a truncated cone. The angle of the cone in this case is different from the truncated
Again, the empirical formula for G2 is obtained by comparing the expression for B(X) given in
equation (5.14), with F(X), so that the footing resistance is the same in both cases.
38
2b
LI
The footin g resistance in this case is determined by the resistance o f the cone, and is given as
(5.13)
Figure 6 sh ow s the variation in G2 with h, for various ratios o f p/p,. The values o f B (X )
com puted for the range o f parameters specified in equation (5.7), are given in Table 5. The
com puted values are once again not very different from the values obtained through the infinite
The tw o m odels have been tested to be quite accurate for the range specified in equation
(5.7) for h and p/p,. Though the truncated cone and soil m odel m ay be used beyond this range,
the truncated cone m odel gives erroneous results for values not in this range.
39
FUNCTION %ERROR
h(cm) R, F(X) H(X) A(X) B(X) H(X) A(X) B(X)
7.5 2 0.733 0.734 0.737 0.709 0.14 0.55 -3.27
7.5 5 0.532 0.538 0:540 0.545 1.13 1.50 2.44
7.5 10 0.455 0.464 0:464 0.468 1:98 1.98 2.86
7.5 20 0.414 0.426 0.423 0.421 2.90 2.17 1.69
7.5 50 0.389 0.402 0.397 0.390 3.34 2.06 0.26
7.5 100 0.380 0.394 0.389 0.379 3.68 2.37 -0.26
7.5 200 0.376 0.390 0.384 0.373 3.72 2.13 -0.80
66 —
64 -
62 -
60 -
58 -
56 -
54 -
52 -
50 -
46 -
44 -
40
38 -
h (cm)
CH A PTER 6
Basic Considerations
The grounding grids o f two substations are likely to get connected unintentionally through
overhead ground wires on transmission lines, underground cables, metallic pipes etc. going from
one substation to the other. Sometimes, the interconnection may be intentional in the form o f a
bare grounding conductor, connecting the two grids. When a fault occurs in one o f the inter
connected substations, it is important to determine the ground resistance o f the system o f two
gnds, and the interconnecting conductor. A lso o f interest is the ground potential rise at the two
For an analysis to determine the potential rise, system ground resistance, and the voltage
(ii) The ground resistance o f the two grids, R 1 and R2, at the two substations.
(iii) The number n, length X , and radius a, o f the underground bare conductors going
from one station to the other. A lso required is the resistivity pc o f the conductors. It is assumed
that all the conductors are o f the same length, radius, and are made o f the same material.
The current injected at one substation gets divided over the interconnecting conductors.
It is thus logical to ignore the voltage drop along the conductors of the grid. The grounding grids
at the two substations are represented by equivalent circular plates o f radius T1and r2 respectively,
such that
Underground conductors are usually buried at a shallow depth, and their length is usually
much longer than the spacing between them. A system o f interconnecting conductors can thus
internal impedance Z, which may be represented as resistance R, and inductance L; and the
the shunt current drawn by the capacitance is negligible as compared to the current drawn by
the conductance. Thus, the shunt capacitance may be ignored for all practical purposes. The
Z = R + J ( I n f )L (6 .2 )
and
Y =G (6.3)
where
ohms/m (6.4)
nna
n
to ? -i siemens/m (6.5)
P A
2
to?-i
A
henries/m (6 6 )
where A is the geometric mean radius of the conductors o f the intertie in meters, and X is the
A solution to the problem of evaluating the potential rise, ground resistance and voltage
profile, can be obtained by dividing the equivalent intertie conductor into a number of sections,
o f equal length, each o f which may be represented as a p i section, cascaded to form the conductor.
43
X/m , has an internal impedance o f ZX/m, and an external admittance o f YX/(2m ) at each leg
o f the pi-sections. These m pi sections, and the two circular plates representing the grounding
-r sin" (6.7)
X + i'i + A12
where r —T1, if r 1>r2, and r —r2, if r2>r^, and p is the resistivity o f the soil in ohm-meters:
The mutual resistance between one grid and any node point p o f the ladder network is
given by
P sin"1 (6 8)
2nr r +x
where r is the radius o f the equivalent plate representing the grid, in meters and x is the distance
The mutual resistance between two nodes p and q, which are located at the end of the p 111
When current I is injected at grid I, it is dissipated to ground from each grid, and by each
node o f the pi-section. A computer program INTEG has been developed to determine the current
dissipated and the voltage at the two grids and each node o f the pi-sections fo rming the intertie.
For a ladder network composed o f N nodes, the current dissipated and voltage, at each o f the
N nodes, form 2N unknowns that are to be determined. The following 2N equations are available:
I. N equations giving the voltages at the nodes in terms o f the currents dissipated at each
node, and the self and mutual ground resistances o f the nodes.
44
2; N -I equations giving the difference in the voltage o f adjoining nodes in terms o f the
series impedance and the current flow ing between the two nodes.
3. One equation given by the sum o f all currents dissipated at the nodes, which is equal
to the known fault current injected into the system at the first grid.
U sing the program INTEG, more than 1000 cases were studied for systems in which the
follow ing variations were observed in the resistance o f the two grids, soil resistivity, and
R 1 and R 2 = I to 20 ohms,
p = 50 to 3000 ohm-meters
Interconnecting conductor:
Number = 2 to 6
The voltage at the two grids, as obtained for a few representative cases are given in Table
8 . The current dissipated and voltage at the grids and nodes, for a sample case, is given in Table
9. The resistance o f the first grid is I ohms. For a fault current o f 1000 amperes, the ground
potential rise o f grid I, would be 1000volts. Due to the intertie and the second grid, the effective
resistance o f the system, as seen by the fault.current falls to 178/1000 = 0.178 ohms.
The solution o f the ladder circuit is simplified if the mutual resistances between the nodes
o f the pi-sections are neglected. The transmission line equations can then be applied to the
45
intertie. The characteristic impedance Z 0 and the propagation constant y o f the intertie are then
given by
Z 0 = (ZIY)112 ( 6 . 11)
The ladder circuit may then be solved by adopting the foUowing methodology:
I-The ground resistance o f the intertie alone is determined by calculating the input
Z i = Z 0Ztanh(YX) (6.13)
2-The ground resistance o f grid 2, R2, is then connected at the far end o f the intertie. The
ground impedance o f the iritertie and grid 2 as seen from end I is given as [18]
R2+X0Ianh(YX))
Zi2= Z l (6.14)
Z0+ R 2Ianh(YX) J
3.The ground resistance o f grid I, R 1, is in parallel with Zi2. There is also mutual ground
resistance between grid I, and the ground system formed by the intertie and grid 2. An
4. The ground impedance o f the complete system comprising the two grids and the intertie
is then given by
V1=IpZsys (6.18)
1 R i+ Z i2-T R m <6-19)
8 .The voltage at any point on the intertie, can be determined by replacing X b y x in equation
A comparison o f the voltages at the two grids, obtained from this method, for a few
representative cases, is made with the voltages as obtained from the ladder circuit method in
For an infinitely long conductor buried in the ground near the surface, the shunt con
ductance and series inductance o f the conductor are so related that a simple approximate
Equations (6.4), (6.5), and ( 6 .6 ) can be simplified for a copper conductor to give
1.7 x KT
2, .74X + l 2 3 7 ( 6 . 21)
Mtrln
Under practical conditions, the real part o f ZY varies from 2.38 to 93.15, and may hence be
Z Y=J-IQ -* (6 .22)
P
47
which implies
Y =V zF (6.23)
(6.24)
(6.25)
N ow , Y=CH-Jp, where a and pare the attenuation constant andphase constant respectively.
a= l/(100V p) (6.26)
a is a measure o f the rate at which the voltage decays along the length o f the conductor.
The length at which the voltage drops to 36.8% o f its initial value is defined as the length
X = 1/ a = lOOVp (6.27)
The length constant may be observed to depend primarily on the resistivity o f the soil. At
a distance o f 2X, the voltage on the intertie w ill drop to 13.5% o f its initial value. Therefore, the
two station grids at a distance o f more than 2 X ihtertied with a buried bare copper conductor,
I 154 178
a 63 178
b 97 165
C 88 154
d 81 146
e 77 139
f 74 134
g 71 128
h 70 127
i 71 126
k 42 125
2 103 125
50
C H A PTER 7
C O N C LU SIO N S A N D P U T U R E W O RK
The expressions presented in this thesis, for the calculation o f ground resistance, and mesh
and step voltages, are easy to use, and should be useful in the design o f substation grids. Existing
formulas for the calculation o f these parameters, are applicable only to square grids, or rect
angular grids with a limited length to width ratio. When these formulas are used for the design
o f grids with other shapes, the results could be highly erroneous. Since the errors from the use
o f existing formulas are usually on the positive side, they lead to substantial wastage of material.
Since the new expressions reduce these errors, they should lead to considerable savings in
The validity o f the expression for ground resistance has been verified for grids whose
shapes and sizes are given in Figure 2. The use o f this expression has also been verified for the
two extreme cases o f a strip conductor, and a circular plate, and through the use of scale models.
It is thus expected that the equation should be valid for all intermediate shapes and sizes o f grids.
A ll the grid shapes given in Figure 2 have a common characteristic in that they use the factor
VA ILp as a measure o f the length over which the area is spread. This is only valid for grid shapes
for which a straight line drawn perpendicular to the periphery at any point, when extended
outside the area o f the grid, always remains outside. The suggested equation, thus, is not valid
for grid shapes which do not satisfy this criteria. Rarely are such grids encountered.
51
A similar limitation exists for the expressions developed for the calculation o f step and
mesh voltages. It may have been observed that the errors in the calculation o f the step voltage
are higher than those for mesh voltage calculations. A s mentioned earlier, this is usually
acceptable. In case higher accuracy is required, new expressions for step voltage calculations
may be warranted.
The purpose o f developing a new expression for calculating the footing resistance within
a substation yard, has not been reduction o f error, but the ease o f use. The expression developed
on the basis o f the two-layer soil model, and recommended by the TF.P.F. Std. 80, is not easy to
use, and requires the help o f a computer or a programmable calculator. The finite formula
suggested by Sverak, as mentioned earlier, gives errors as high as 50% in some cases, and is
thus not readily applicable. The new expression, and the models developed in this thesis, do not
give errors o f more than 4%, for any o f the cases considered, which cover the usual practical
situations. The two models presented give information regarding the region o f crushed rock
below tire foot that effectively contributes to the footing resistance. This information is useful
The method to evaluate the performance o f intertied grids is easy to use, and gives an
improved idea o f the likely behaviour of such a system. The simplified method, which is based
on the transmission line equations, should be used to make a quick analysis o f the system, and
should not be used as a final analysis tool. For comprehensive and acurate calculations, the
program INTEG is preferred. It is interesting to note that the length constant depends primarily
on the resistivity o f the soil. A t a distance o f 2X, the voltage on the intertie w ill drop to 13.5%
o f its initial value. Thus, two substation grids located a distance more than 2X apart will have
The expressions for the calculation o f step voltage have been adapted from the calculations
carried out for the m esh voltage. A s Suchi the errors for the step voltage calculations are higher.
More work needs to be done in this area. One possibility is the development o f a separate current
irregularity factor Kj for step voltage as a function o f n, and possibly, the depth o f burial.
The equations as presented in this paper cannot be applied to the practical design o f a grid,
as rarely is the grid configuration known ahead o f time. It is thus necessary to formulate a design
strategy, where, once the safety limits for the step and maximum touch voltages are established,
the addition o f grid conductors can be done on an incremental basis. Such a formulation can
incorporate the equations presented in this paper, and be applied to grids o f all practical shapes.
The results presented here have been collected on the basis o f computer runs and analog
m odel tests. It is very important to collect data on the basis o f measurements made in actual
substation sites, and make a comparison o f the measured and calculated data. This would also
give a good indication o f the direction in which future work should proceed.
53
R EFER EN C ES CITED
54
[1] "IEEE G uide for Safety in A C Substation Grounding" ANST/TF.F.F. SM Rn" 1986.
[4] Biegelm eier,U.G.,"Die Bedeatung der Z-Schwelle des Herzkammerfilimmems fur die
Fesdegung von Bemhrunggsspanugs greuzeu bei den Schutzma Bradhmer gegen
elektrische Unfalle", E&M. Vol.93, N o .l, 1976, pp.1-8.
[7] ThaparfB., V.Gerez, A.Balakrishnan, D.A.Blank, "Finite Expression and M odels for
Footing Resistance in Substations", Forwarded for presentation at the TRRR/PRS
International Meeting. India.
[11] Sverak, J.G., "Simplified Analysis o f Electrical Gradients A bove a Ground Grid: Part
I - H ow Good is the Present IEEE Method?", IEEE Transactions on Power Apparatus
and System s. Vol. PAS-103, pp.7-25, January 1984.
[13] Kercel,S.W ., "Design o f Switchyard Grounding Systems UsingM ultiple Grids", TPRR
Transactions on Power Apparatus and System s. Vol. PA S-100, pp.1341-1350, March
1981.
[14] Seedhar 1FLR., J.K.Arora, B .Thapar, "Finite Expressions for Computation of Potential
in Two Layer Soil", IEEE Transactions on Power Delivery. Vol. PWRD-2, 1987,
p p.1098-1102.
55
[15] Sobral, S.T., et al, "Decoupled Method for Studying Large Scale Interconnected
Grounding Systems Using Microcomputers- Part I - Fundamentals", TFRR T r a n s i
tions on Power Delivery. Vol. PW RD-3, 1988, pp.1536-1544.
[17] Joy,E.B., N.Paik, T.E.Brewer, R.E.W ilson, R.P.Webb, A.M eliopoulos, "Graphical
Data for Ground Grid Analysis", IEEE Transactions on Power Apparatus and System s.
Vol. P A S-102, pp.3038-3048, September 1983.
[18] Magnusson, P C., Transmission Lines and W ave Propagation. Allyn and Bacon,
M O N TA N A S T A T P Iiu a / cdcttv ■m n .
3 1762 10036045 0