WF—U-v ————~—

‘5..- :'
‘ 1‘3
~<

W

.3 3 . :_.
-J"< ‘yf‘

n‘k» 4": ~ ,1 "

.242?

J". xx.

‘5
:1":

I
1x
".

“V

I?“
' m
.

$51

#3:}, '3: r I
1. 4)

'7
“3|
.1

l 3"

i

'9»:

> 7
? 1: ‘
‘.. 1:71”
w 3

.g 1
"at
't‘

> i1
1
3‘3?" 4 4

4L

a 5

.
9‘?

5r

 

l Aw
-, .

.12
W

 

fir}
‘ "A .'

‘11}
1...

'1
>I

1
"‘31 ‘“Es .
1! k3"; 1‘

a«'

9‘3: 0
xi»?
IA.
11 ,1.

w l'
«1.1.4. ,

Q’Q‘k;
‘1’ ‘”

r
l

 

1.. ‘ 11.1.11 $1,119,111
{1 '1 v o 1'!

J
1 . . ,1
filialg§él
‘0?” d V “

 

 

' «31.:- v as, + -
n‘ I “'ll‘ . _
tn: fifilfi'ii “9.1 I
13.94. ,1... _
"'2’ , n ,i ’1‘
' ’ x 13
i , I!
wfi‘ 5mg :1
1‘. . J
3&1» '2}, ~:
”K -‘

5411);“. '.'~‘;,"

 

«T

2’35? -

.‘ , pg

‘ 1 - c.1' L ‘
I ‘1'; ' J 1 4336352;
' W 113 5;? VP
“‘4: u .~ . A
big—€11.55. -
‘kmwgb

FUR? -

{w
55:21' 1
M.

a

:l‘Af:
i"‘."

:5

Jaw:
( * 45;}. 1p
wr'Sfi’fi ‘12-’13; .‘
"NW1?! ‘

1 ' ,r'
.39—

1 1-54..
Q 1
I figs?! Wu
“Wk:

 

‘ ' I 1
‘ “if" ‘ ' {A . '
.1 ‘7, {$3117 , 235':
1 I 1'4“ 1’7"?” '7 ‘m‘
. WM 51
1‘" '

1,
*1 fir

   

.Z"

.
‘“ w a. P
AM»

, “or :1;

   

~
‘1 .“ A 1
i: ”if,“ f

‘ )

‘ It
.- w
1,

9&1,»

.r 14”
£5
2
PL
1'

1.
1.1 .1
:J

.3 “

'4": " 1

my
. . . WW? - .1 .
’15'3‘,~,-\?2i€452433fi . ,1
I»! - 1 ~.;- 1 3
‘1%.;J:=.a.~\-' Mr
1.5m!»

[TY

iii‘i

‘ \\‘i\\ W

liliiiiiili" iii

a
H
l
3 1293 01409 9372

ii

This is to certify that the

dissertation entitled

Analysis of Neutral-to-Earth Voltage Characteristics along
Rural Electric Power Distribution Systems through
Computer Simulation and Field Testing

presented by
(hanging Li

has been accepted towards fulfillment
of the requirements for

 

Ph.D. degree in Agricultural Engineering

QMWW C5510}

. ' v
Major professor

3/ K/fif

MSU is an Affirmative Action/Equal Opportunity Institution 0- 12771

____—_-—___._—._—-——__

 

LIBRARY
MIchlgan State
UniversIty

 

 

 

PLACE N RETURN BoxwmmmhMMMywm
TO AVOID FINES Mum on or bdon duo duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU IoAn Affirmative Adm/Emu Oppommlly Imam
mm:

ANALYSIS OF N EUTRAL—TO-EARTH VOLTAGE
CHARACTERISTICS ALONG RURAL ELECTRIC
POWER DISTRIBUTION SYSTEMS THROUGH
COMPUTER SIMULATION AND FIELD TESTING

By

Changming Li

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

‘6

Department of Agricultural Engineering

1994

ABSTRACT

ANALYSIS OF NEUTRAL-TO-EARTH VOLTAGE
CHARACTERISTICS ALONG RURAL ELECTRIC POWER
DISTRIBUTION SYSTEMS THROUGH COMPUTER
SIMULATION AND FIELD TESTING

By

Changming Li

The multi-grounded rural electric power distribution systems in
the U. S. can produce neutral-to-earth voltage (NEV) on a dairy farm
which may affect milk production of a dairy cow. In this study,
computer simulation models were developed and farm field tests were
conducted to facilitate the investigation and analysis of this NEV
phenomenon.

The computer models of nearly 4.3 kilometers long, 4.8 kV rural
power distribution systems simulated the NEV changes arising from
high resistance segments in the neutral of the primary line, primary
heavy loading demand, substation and other ground resistance changes,
different levels of the neutral line grounding, primary operating voltage-
change, secondary ground faults and primary phase-to-earth faults.
Based upon the analysis of the simulation results, guidelines were
established for power suppliers to identify the sources of NEV

originating from the parameter changes of the power distribution

system. It is expected that the overall NEV is substantially lower along
the electric power distribution system of a balanced three-phase
four-wire Yo/Yo circuit than that of the single-phase.

The farm field tests revealed the NEV gradient distribution near
the ground electrode systems on the primary and secondary sides of the
neutral isolated distribution transformer. The NEV gradient which the
isolated secondary ground electrode coupled from the primary side
ground system was determined and analyzed. The equivalent circuit of
the isolated secondary ground system was developed to study the
parameters that influence the neutral separation. Low farm grounding
system resistance to earth is important for effective neutral separation.
Increased separation distance between the primary and secondary
ground rods is a factor for separation effectiveness increasing.
Resistivity of the soil is also an important parameter. The low

resistance of the isolated farm ground will result in low NEV on farm.

In memory of the author’s mother

iv

ACKNOWLEDGMENTS

I would like to express my deepest gratitude to Dr. Truman C.
Surbrook, advisor and friend, for his academic guidance, moral
encouragement and technical support since 1988 throughout my
master’s and doctoral programs. His fruitful cooperation with Michigan
Consumers Power Company has made this study possible financially.

I would like to extend my cordial thanks to Dr. Larry J. Segerlind,
Dr. H. Roland Zapp and Dr. Howard L. Person for their serving on my
Ph. D. guidance committee. Their helpful academic guidances and
moral encouragements are greatly appreciated by the author. Special
thanks also goes to the faculty, staff and other graduate students of the
Department of Agricultural Engineering for their assistance and
cooperation in the recent years.

A very special thanks goes to my parents and all of my family
members for their expectation, encouragement, understanding and
moral support from the remote China.

Special thanks goes to the N ingxia T.V. University, China for
letting me attend the graduate school in the United States, and for its

subsequent moral encouragement of me.

TABLE OF CONTENTS

LIST OF FIGURES ........................................... ix
LIST OF TABLES .......................................... xvii
I. INTRODUCTION ...................................... 1
II. LITERATURE REVIEW ....................................... 5
2.1 Stray Voltage and Electric Current Effect on Dairy Cows ........... 6
2.2 Possnble Sources of the Farm N eutral-to-Earth Voltage ........... 14
2.3 Field Test and Diagnosis of Farm Neutral-to-Earth Voltage ....... 21
2.4 Mitigation of Farm N eutral-to-Earth Voltage Problem ............ 24
III. OBJECTIVE .................................. . ......... 31
IV. METHODOLOGY ..................................... 34
4.1 Computer Simulations of Primary N eutral-to-Earth Voltages ...... 34
4.1.1 The DC Single-Phase Distribution Simulation Model ....... 34
4.1.2 The AC Single-Phase Electric Power Distribution Model . . . .40
4.1.3 The AC Three-Phase Electric Power Distribution Model . . . .43
4.2 Simulation of Abnormal Conditions along the Distribution line ..... 45
4.2.1 Neutral Conductor Resistance Change .................... 45
4.2.2 Heavily Loaded Power Line on Farm ..................... 46
4.2.3 The Distribution System Ground Resistance Calculation . . . . 47
4.2.4 Substation Grounding Resistance Change ................. 49
4.2.5 Neutral-to-Earth Resistance Reduction ................... 49
4.2.6 Primary Operating Voltage Level Change ................ 50
4.2.7 Secondary Earth Fault .................................. 52
4.2.8 Primary Phase-to-Earth Fault ............................ 54

4.3 Field Tests of Neutral-to-Earth Voltage Gradient ................ 57
4.3.0 Data Collection of Voltage and Resistance of Ground Rod . .58
4.3.1 Test of NEV Gradient Distribution,
Primary One Ground Rod ............................... 61
4.3.2 Test of NEV Gradient Distribution,
Primary Two Ground Rods .............................. 62
4.3.3 Test of NEV Gradient Distribution,
Primary One Ground Rod, Secondary Isolated One Ground
Rod Without Farm Ground Connection ................... 63
4.3.4 Test of NEV Gradient Distribution,
Primary One Ground Rod, Secondary Isolated One Ground
Rod With Farm Ground Connection ..................... 65
4.3.5 Verification of the Neutral Isolation Circuit Model ......... 67
V. RESULTS and DISCUSSION ............................ 70
5.1 Results and Analysis of NEV Computer Simulation .............. 70
5.1.0 Normal Operational Condition .......................... 70
5.1.1 Neutral Conductor Resistance Change .................... 75
5.1.2 Heavily Loaded Power Line on Farm ..................... 83
5.1.3 The Distribution System Ground Resistance Calculation . . . . 90
5.1.4 Substation Grounding Resistance Change ................. 98
5.1.5 Neutral-to-Earth Resistance Reduction .................. 103
5.1.6 Primary Operating Voltage Level Change ............... 111
5.1.7 Secondary Earth Fault ................................. 114
5.1.8 Primary Phase-to-Earth Fault .......................... 130
5.2 Guidelines to Identify NEV along the Distribution Systems ....... 139
5.3 Field Tests of Neutral-to-Earth Voltage Gradient ............... 141
5.3.1 Test of NEV Gradient Distribution, _
Primary One Ground Rod ............................. 142
5.3.2 Test of NEV Gradient Distribution,
Primary Two Ground Rods ............................ 154
5.3.3 Test of NEV Gradient Distribution,

Primary One Ground Rod, Secondary Isolated One Ground

Rod Without Farm Ground Connection ................. 164
5.3.4 Test of NEV Gradient Distribution,

Primary One Ground Rod, Secondary Isolated One Ground

Rod With Farm Ground Connection .................... 169
5.3.5 Verification of the Neutral Isolation Circuit Model ........ 174
5.3.6 Parameters Influencing Effectiveness of
Neutral Separation .................................... 188
VI. CONCLUSION ....................................... 201
APPENDIX ................................................ 206
REFERENCES ............................................ 224

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

LIST OF FIGURES

Page
Behavioral and milk production responses to increasing
current levels ................................................. 8
Circuit Model for Estimation of Effect of Neutral Isolation on
N eutral-to-Earth Voltage (Althouse, 1990) ...................... 29
Transition Design for Equipotential Planes on Farm .............. 30
DC circuit model for single-phase electrical distribution system
(input voltage 4,800 V) ....................................... 35
AC circuit model for single-phase electrical distribution system
(input voltage 4,800 V) ....................................... 41
AC circuit model for three-phase four-wire wye connection
(Yo/Yo connection) distribution system (4800 V for each
phase-to-neutral voltage) ...................................... 44
DC circuit model for single-phase electrical distribution system
with secondary earth fault simulated ............................ 53
DC circuit model for single-phase electrical distribution system
with primary phase-to-earth fault simulated ..................... 55

AC circuit model for three-phase four-wire wye connection
(Yo/Yo connection) distribution system (4800 V for each
phase-to-neutral voltage) with phase A to earth fault simulated ..... 5 6

A variable transformer was used to drive electric current into
the earth at the test primary ground rod ......................... 59

The 40 x 40 Feet Square Shaped Test Plot for NEV Gradient
Testing, Primary One and Two Ground Rods .................... 60

Figure 12.

Figure 13.

Figure 14.

Figure 15.

Figure 16.

Figure 17.

Figure 18.

Figure 19.

Figure 20.

Page
The 4 x 10 Feet Square Shaped Test Plot for NEV Gradient
Testing, Isolated Primary and Secondary Ground Systems ......... 64

A secondary isolated ground rod and farm grounding system
were established to an area away from influence of earth
voltage gradients from other electrical systems ................... 66

The physical quantities measured in the neutral separation
circuit model ................................................ 69

Profile of the neutral-to-earth voltage along the primary
distribution line from substation (node 1) to 57th node for
the DC single-phase base model .......... , ...................... 71

Profile of the neutral-to-earth voltage along the primary
distribution line from substation (node 1) to 57th node for
the AC single-phase base model ................................ 73

Neutral-to-earth voltage profile comparison between the
simulations of single-phase AC and DC models .................. 74

Profile of the neutral-to-earth voltage along the primary

distribution line from substation to 57th node with variable

neutral conductor resistance RD9 between node 9 and node 10

and a normal customer load of 3A at node 9 (single-phase

DC model simulation) ........................................ 76

Profile of the neutral-to-earth voltage along the primary

distribution line from substation to 5 7th node with variable

neutral conductor resistance RD33 between node 33 and node

34 and a normal customer load of 3A at node 33 (single-phase

DC model simulation) ........................................ 77

Profile of the neutral-to-earth voltage along the primary

distribution line from substation to 5 7th node with variable

neutral conductor resistance RD56 between node 56 and

node 57 and a normal customer load of 3A at node 57

(single-phase DC model simulation) ............................ 79

Figure 21.

Figure 22.

Figure 23.

Figure 24.

Figure 25.

Figure 26.

Figure 27.

Page

The circuit connection to study the neutral-to-earth voltage

changes resulting from changing the neutral conductor

resistance RD9, RD33 and RD56 in three different

representative locations along the line at normal load demand
condition (3 A load) .......................................... 81

Profile of the neutral-to-earth voltage along the primary

distribution line from substation to 57th node with variable load
demand conditions at node 404-9 (single-phase DC model

simulation) .................................................. 84

Profile of the neutral-to-earth voltage along the primary

distribution line from substation to 57th node with variable load
demand conditions at node 416-33 (single-phase DC model
simulation) .................................................. 86

Profile of the neutral-to-earth voltage along the primary

distribution line from substation to 57th node with variable load
demand conditions at node 428-57 (single-phase DC model
simulation) ................................................. 88

The test circuit to determine the system resistance from
the farm at node 33 ........................................... 93

The simulation plot of changes of the distribution system ground
resistance (Thevenin Equivalent Resistance) and Thevenin
Equivalent Open Circuit Voltages at the location of node 33

to earth as well as neutral-to-earth voltage at node 33 with the
resistances increased from 0.086 ohm to 100 ohms in the neutral

line segments RD4, RD8, RD12, RD16, RD20 RD56,

respectively in the DC circuit model ............................ 95

The circuit connection to determine the changes of the

distribution system ground resistance (Thevenin Equivalent
Resistance) and Thevenin Equivalent Open Circuit Voltages at

the location of node 33 to earth as well as neutral-to-earth voltage

at node 33 with the resistances increased from 0.086 ohm to 100

ohms in the neutral line segments RD4, RD8, RD12, RD16,

RD20 RD56, respectively ................................... 96

Page

Figure 28. Profile of the neutral-to-earth voltage along the primary
distribution line from substation to 57th node with base case vs.
variable substation resistance RS (single-phase DC model
simulation) .................................................. 99

Figure 29. The substation Thevenin equivalent open circuit voltage is
divided in series by the substation ground resistance and the
system Thevenin equivalent resistance at this location ............ 104

Figure 30. Profile of the neutral-to-earth voltage along the primary
distribution line from substation to 57th node with base case vs.
ground resistance RFG33 change at node 33. (single-phase DC
model simulation) ........................................... 105

Figure 31. Profile of the neutral-to-earth voltage along the primary
distribution line from substation to 57th node with base case vs.
ground rod number increasing along the entire system
(single-phase DC model simulation) ........................... 110

Figure 32. Profile of the neutral-to-earth voltage along the primary
distribution line from substation to 57th node with base case
4.8 kV operating voltage vs. 2.4, 7.2 and 26.4 kV operating
voltages (single-phase DC model simulation) .................. 112

Figure 33. Profile of the neutral-to-earth voltage along the primary
distribution line from substation to 57th node with base case vs.
variable secondary ground fault attached to node 33
(out-of-phase, single-phase DC model simulation) ............... 115

Figure 34. Profile of the neutral-to-earth voltage along the primary
distribution line from substation to 57th node with base case vs.
variable secondary ground fault attached to node 33 (in-phase,
single-phase DC model simulation) ............................ 117

Figure 35. Profile of the neutral-to-earth voltage along the primary
distribution line from substation to 57th node with base case vs.
variable secondary ground fault attached to node 9
(out-of-phase, single-phase DC model simulation) ............... 119

Figure 36.

Figure 37.

Figure 38.

Figure 39.

Figure 40.

Figure 41.

Figure 42.

Figure 43.

Page

Profile of the neutral-to-earth voltage along the primary

distribution line from substation to 5 7th node with base case vs.
variable secondary ground fault attached to node 9 (in-phase,
single-phase DC model simulation) ............................ 121

Profile of the neutral-to-earth voltage along the primary

distribution line from substation to 57th node with base case vs.
variable secondary ground fault attached to node 57

(out-of-phase, single-phase DC model simulation) ............... 122

Profile of the neutral-to-earth voltage along the primary

distribution line from substation to 57th node with base case vs.
variable secondary ground fault attached to node 57 (in-phase,
single-phase DC model simulation) ............................ 124

Neutral conductor and grounding electrode current with
secondary out-of-phase ground fault attached at node 57
at the end of the distribution line .............................. 127

Neutral conductor and grounding electrode current with
secondary in-phase ground fault attached at node 57
at the eitd of the distribution line .............................. 128

Profile of the neutral-tooearth voltage along the primary

distribution line from substation to 57th node with an

ungrounded phase conductor to earth fault at node 416

compared with the normal line

(single-phase DC model simulation) ........................... 131

Profile of the neutral-to—earth voltage along the primary

distribution line from substation to 57th node with 12 A phase
conductor to earth fault at node 404, 416, or 428 compared with
normal line (single-phase DC model simulation) .- ............... 133

Profile of the neutral-to-earth voltage along the three-phase

primary distribution line from substation to 57th node with an
ungrounded phase conductor (phase A) to earth fault at node

416A compared with the normal line (three-phase AC model
simulation) ................................................. 135

Figure 44.

Figure 45.

Figure 46.

Figure 47.

Figure 48.

Figure 49.

Figure 50.

Figure 51.

Figure 52.

Profile of the neutral-to-earth voltage along the three-phase
primary distribution line from substation to 57th node with
12 A phase conductor (phase A) to earth fault at node 404A,
416A, or 428A compared with normal line (three-phase AC

model simulation) ........................................

Three dimensional graphic plot of the neutral-to-earth voltage
gradient distribution from single driven electrode rod (length

unit is foot) ..............................................

Equi-potential lines of the neutral-to-earth voltage gradient
distribution from single driven electrode rod

(length unit is foot) .......................................

Neutral-to-earth voltage gradient distributions from single

driven electrode rod along four directions (length unit is foot) . .

Comparison between theoretical value calculated from
equation (5.3) and field measured value of neutral-to-earth
voltage gradient distributions from single driven electrode rod

along east direction (length unit is inch) ......................

Three dimensional graphic plot of the neutral-to-earth voltage
gradient distribution from two electrode ground rods with 6

feet separation distance (length unit is foot) ..................

Equi-potential lines of the neutral-to—earth voltage gradient
distribution from two electrode ground rods with 6 feet

separation distance (length unit is foot) ......................

Graphic plot of the neutral-to-earth voltage profile data
comparison (per unit normalized) between one rod and two
rods of primary ground systems along the north direction

(length unit is foot) .......................................

Graphic plot of the neutral-to-earth voltage profile data
comparison (per unit normalized) between one rod and two
rods of primary ground systems along the other three

directions (length unit is foot) ..............................

xiv

Page

...137

...146

...147

. . .149

..152

..157

..158

. . .161

. . .163

Figure 53.

Figure 54.

Figure 55.

Figure 56.

Figure 57.

Figure 58.

Figure 59.

Figure 60.

The test circuit for the neutral-to-earth voltage gradient
distribution resulting from one primary ground rod, and one
secondary isolated ground rod without the farm grounding

system connected ..........................................

Three dimensional graphic plot of the neutral-to-earth voltage
gradient distribution from primary one ground rod, secondary
isolated one ground rod (6 feet separation distance) without

farm ground connection (length unit is foot) ..................

Equi-potential lines of the neutral-to-earth voltage gradient
distribution from primary one ground rod, secondary isolated
one ground rod (6 feet separation distance) without farm

ground connection (length unit is foot) .......................

The test circuit for the neutral-to-earth voltage gradient
distribution resulting from one primary ground rod, and an
isolated secondary ground rod connected to a 5.24 ohm farm

ground resistance .........................................

Three dimensional graphic plot of the neutral-to-earth voltage
gradient distribution from primary one ground rod, secondary
isolated one ground rod (6 feet separation distance) with farm

ground connection resistance 5.2 ohm (length unit is foot) ......

Equi-potential lines of the neutral-to-earth voltage gradient
distribution from primary one ground rod, secondary isolated
one ground rod (6 feet separation distance) with farm ground

connection resistance 5.2 ohm (length unit is foot) .............

Comparison between neutral-to-earth voltage gradient
distributions (per unit normalized) from secondary isolated
ground systems with and without farm ground connection along
the direction of primary and secondary isolated connection line

(length unit is foot) ........................................

The circuit model of Figure 14 of Section 4.3.5 is arranged

vertically so that all quantities can be visualized more easily .....

Page

.166

..167

. .168

..171

..172

. .173

. .177

. 182

Figure 61.

Figure 62.

Figure 63.

Figure 64.

Figure 65.

The circuit model of Figure 14 of Section 4.3.5 without farm
ground system connection is arranged vertically so that all

quantities can be visualized more easily ......................

Slightly altered form of the neutral separation circuit model

(Althouse, 1990) ..........................................

The ratio of farm neutral-to-earth voltage (EF) to primary
ground rod neutral-to—earth voltage (Es) was plotted against
the value of farm grounding system resistance to earth (R1,).
The data was from Table 18 and Table 19 for two levels of

primary grounding resistance ...............................

The ratio of farm neutral-to-earth voltage (BF) to primary
ground rod neutral-to-earth voltage (Es) was plotted against
the value of farm grounding system resistance to earth (Rr)'

The data was calculated from the circuit model in Figure 62 . . . .

The ratio of farm neutral-to-earth voltage (Er) to primary
ground rod neutral-to-earth voltage (Es) was plotted against
the value of farm grounding system resistance to earth (RF) for
6 inch and 6 foot separation distance. The calculated data

was from the circuit model in Figure 62 ......................

Page

..185

..191

..194

. . .197

..199

Table 1.
Table 2.

Table 3.

Table 4.

Table 5.

Table 6.

Table 7.

Table 8.

LIST OF TABLES

Page
Resistance of various electrical pathways through the cow .......... 7

Parameters for Base DC Circuit Model in Figure 4 ............... 37

Some typical values of the ground resistances for the farm and
residential distribution transformers on the field in the service
area of the Consumers Power Company ......................... 42

Primary Current and Load Resistance, RRL and RFL, for
Different Primary Voltage Simulations ......................... 51

Calculation results of the distribution system ground resistance
from the four different representative locations along the
single-phase DC distribution line model ......................... 91

The simulation data of changes of the distribution system ground
resistance (Thevenin Equivalent Resistance) and Thevenin
Equivalent Open Circuit Voltages at the location of node 33 to

earth as well as neutral-to-earth voltage at node 33 with the
resistances increased from 0.086 ohm to 100 ohms in the neutral

line segments RD4, RD8, RD12, RD16, RD20 RD56,

respectively in the DC circuit model ............................ 97

Numerical calculation results of substation system ground

resistance, open circuit neutral-to-earth voltage,

ground resistance and neutral-to-earth voltage of the

single-phase DC distribution line model ........................ 102

Numerical calculation results of system ground resistance, open

circuit neutral-to-earth voltage, ground resistance and

neutral-to-earth voltage at mid-point of the single-phase DC
distribution line model ....................................... 108

Table 9.

Table 10.

Table 11.

Table 12.

Table 13.

Table 14.

Table 15.

Table 16.

Table 17.

Page
Test measurement data of the neutral-to-earth voltage gradient
distribution from single driven electrode rod (length unit is foot). . . 144

Test measurement data of neutral-to-earth voltage gradient
distributions from single driven electrode rod along four
directions .................................................. 148

Data comparison between theoretical value calculated from

equation (5.3) and field measured value of neutral-to-earth

voltage gradient distributions from single driven electrode rod

along east direction (length unit is inch) ........................ 151

Test measurement data of the neutral-to-earth voltage gradient
distribution from primary two electrode ground rods with 6 feet
separation distance (length unit is foot) ........................ 155

N eutral-to-earth voltage test data (per unit normalized) from
primary ground systems of two electrode rods along the north
direction ................................................... 160

Test measurement data of the neutral-to-earth voltage profile

data comparison (per unit normalized) between one rod and

two rods of primary ground systems along the other three

directions .................................................. 162

Test measurement data of the neutral-to-earth voltage gradient
distribution from primary one ground rod, secondary isolated

one ground rod (6 feet separation distance) without farm ground
connection (length unit is foot) ............................... 165

Test measurement data of the neutral-to-earth voltage gradient
distribution from primary one ground rod, secondary isolated one
ground rod (6 feet separation distance) with farm ground

connection resistance 5 .2 ohm (length unit is foot) .............. 170

Data comparison between neutral-to-earth voltage gradient
distributions (per unit normalized) from secondary isolated

ground systems with and without farm ground connection

along the direction of primary and secondary isolated

connection line ............................................. 175

Table 18.

Table 19.

Table 20.

Table 21.

Table 22.

Table 23.

Page

Data for neutral separation test using one primary ground rod
and one secondary ground rod with a six foot separation ......... 178

Data for neutral separation test using two primary ground rods
and one secondary ground rod with a six foot separation ......... 179

Calculated values of grounding electrode and ground circuit
resistances from data in Table 18 and Table 19 ................. 183

Data for neutral separation test of Althouse (1990) using one
primary ground rod and one secondary ground rod with a six
inch separation and six foot separation ........................ 189

calculated value of grounding electrode and ground circuit
resistances from Althouse (1990) data in Table 21 ............... 190

Ground rod spacing, grounding electrode resistances, and ratio
of secondary to primary ground rod neutral-to-earth voltage at
the separated transformer location ............................ 196

I. INTRODUCTION

The most prevalent rural electric power distribution systems in the
United States are multi-grounded neutral systems. These are generally
single-phase two-wire or three-phase three-wire or three-phase
four-wire systems which are effectively grounded to provide for the
power delivery with maximum safety and reliability.

The multi-grounded distribution system under normal and
abnormal operating conditions will produce neutral-to-earth voltage all
along the distribution line. The neutral conductor is required by the
National Electrical Safety Code (IEEE, 1990) to be connected to the
earth by means of ground rods or similar means at four locations for
each mile of the line. When electrical current flows along this
grounded neutral conductor, a portion of the current will also flow
through the earth back to the substation or source transformer, resulting
in some neutral-to-earth voltage (NEV). Farm electrical systems are
also multi-grounded and also produce the neutral-to-earth voltage.

Under some abnormal operational condition (such as ground
fault), the power distribution system of a farm can produce a
substantially high level of neutral-to-earth voltage which bothers
animals. A loose terminal or damaged insulation can lead to an

on-farm electrical fault.

2

The neutral~to-earth voltage can affect animals in two different
ways. In the way of a "touch voltage" or in the way of a "step voltage".
The ”touch voltage" is a rise in potential of metal objects such as water
pipes, stanchions and feeders contacted by animals making contact to
the earth or through the floor to the earth. The "step voltage" is related
to the voltage gradient measured horizontally on the earth’s surface
particularly in close proximity to the ground electrode system. The "step
voltage" is actually a voltage experienced by an animal between hooves
making contact with the floor or earth.

Both "touch voltage" and "step voltage" could present an electric
shock to animals on farms when a complete current path is formed and
the voltage is of sufficient level. Some dairy farmers complain that their
cows’ milk production has been adversely affected by the electric shock
due to the presence of this kind of voltage, often called stray voltage.
Stray voltage is defined as (USDA, 1991) "A difference in voltage
measured between two surfaces that may be contacted simultaneously
by an animal.”

In past years, research on neutral-to-earth voltage on dairy farms
has made significant progress. However, the levels of neutral-to-earth
voltage along a distribution line are not clearly understood in relation to
grounding and ground resistance, loading demand, neutral resistance,
ground fault and primary / secondary neutral isolation. Guidelines have
not been available for efficient and accurate diagnosing and remedying
distribution line abnormalities which cause elevated levels of

neutral-to-earth voltage. The mechanisms of mitigation measures for

3

neutral-to-earth voltage on dairy farms have not been scientifically
elucidated for extensive acceptance. These are very important issues
discussed in this study.

In the study of this dissertation, computer simulation models are
developed and farm field tests were conducted to facilitate a thorough
investigation and analysis of neutral-to-earth voltage levels along the
rural electric power distribution systems.

Computer models of 4.26 kilometers (2.65 miles) long, 4.8 kV
rural power distribution systems simulate the neutral-to-earth voltage
changes arising from high resistance segments in the neutral of the
primary line, primary heavy loading at one point, substation and other
ground resistance changes, different levels of the neutral-to-earth
resistance, primary operating voltage change, secondary ground faults
and primary phase-to-earth faults.

Actual data from an operating distribution line is important to be
considered, but controlled experiments are difficult to conduct and in
some cases may create dangerous conditions. Computer simulation of a
primary distribution system provides the opportunity to study conditions
whose field data are not available or not practical or safe to obtain.

Based upon the analysis of simulation results, guidelines need to
be established for power suppliers to identify the sources of
neutral-to-earth voltage originating from the parameter changes of the
power distribution system. It is expected that the overall

neutral-to-earth voltage is substantially lower along the electric power

4

distribution system of a three-phase four-wire Yo/Y0 connection than
that of a single-phase line.

It is expected that the equi-potential contour lines of the voltage
gradient in the area of a single ground rod with a neutral-to-earth
voltage will be a series of circular rings. When there is more than one
ground rod the distribution of the voltage gradient is not exactly known.
When a primary and secondary neutral are separated, the secondary
ground rod at the transformer will likely be installed in the voltage
gradient created by the primary neutral-to-earth voltage. This research
will investigate the voltage gradient for a single ground rod and two
ground rods with a neutral-to-earth voltage.

It is expected that some of the primary neutral-to-earth voltage
will appear on the secondary ground rod when the neutrals are
separated. Past research indicates that the amount of neutral-to-earth
voltage appearing on the secondary is very small. It is also believed that
the farm grounding resistance-to-earth is an important parameter in
determining the amount of primary neutral-to-earth voltage which will
appear on the secondary neutral. A controlled study will be conducted
in the field to investigate the influence of farm grounding resistance on
the effectiveness of neutral separation as well as the influence of two

primary ground rods as compared to a single ground rod.

II. LITERATURE REVIEW

Since early this century, extensive research has been conducted
concerning the effect of electrical voltage and current on humans. The
lethal threshold value of continuous alternating current through the
human body was determined by several researchers. Based on the
results of Dalziel’s studies (IEEE, 1986), 99.5% of all persons can safely
withstand, without ventricular fibrillation, the passage of a current not
exceeding a magnitude 1B (in amperes) for duration ts (in seconds)
determined by the formula 13 = k ts-005. For so kg body weight,
k=0.116.

Since the 1960’s, extensive research has also found the effect of
the electric voltage and current on animals as well as general indications
of levels of voltage that bothered cows. Phillips (1963, 1969) and
Woolford (1971, 1972) studied the effects of voltage and current on milk
production (New Zealand). In past years, significant research
concerning farm stray voltage has also been conducted in the United
States.

Since the late 1970’s, the major concerns of stray voltage
researchers in the U. S. and other countries were focused on: (1) How,
and to what extent, does the neutral-to-earth voltage affect cows’ milk
production? (2) What is the source of the stray voltage? How is the
neutral-to-earth voltage produced? What is its changing trend under the

6

various operational conditions of an electrical system? (3) Is it possible
to mitigate or lower neutral-to-earth voltage to an insignificant level on
a farm? If yes, how? The literature will be reviewed with respect to these

issues.
2.1 Stray Voltage and Electric Current Effect on Dairy Cows

U. S. Department of Agriculture (USDA) summarized the Effects
of Electrical Voltage/Current on Farm Animals in Agricultural Handbook
No. 696 in 1991. This summary includes the following:

1. Today, stray voltage/current is a recognized phenomenon. Cows
are considered to be more susceptible to stray voltage / current
than humans because cows have lower body impedances and
lower contact impedances (cows are nearly always in contact with
moisture in the farm animal environment). There are several
pathways for current to pass through the cows and the lowest
impedance pathway is from the mouth to the hooves. Table 1
shows the resistances of various electrical pathways through the
cow (USDA, 1991).

2. The primary impact of stray voltages /currents on milk production
involves cow’s behavior changes rather than physiological changes.
Good milk yield can probably be maintained despite the presence

of moderate levels of stray voltage / current if farming practices are

good.

7

The effects of current (and voltage) on the cow’s

behavioral response and milk production are shown graphically in
Figure 1 (USDA, 1991).

 

Table 1. Resistance of various electrical pathways through the cow
Res-stances ct vanous electrical W W the cow‘
Rm Cm
‘ Mean Range My
hmm___ .n; mu m (Ha m—
Mouth to all hooves 70 350 323-393 60 cm a at. 1970
23 36l 244-52!»J so Naell et at. 1933
Mouth to te- hooves 23 as ass-776’ so Naell a at. m:
Mouthtolrunthooves 23 624 new so Naeuaettm
Front leg to re. leg 5 300 250-403 60 talcum. 1932
13 362 3024:: 60 wen-t a at. 1905
FM to M hooves 23 734 49641523 so Norell er .1. I903
Rumptoall homes 7 6” 411-130 50 Wall. 1975
Chainsaw 5 900 fill-1230 so Would”:
7 won 2 50 Mad. 1m
remom 23 433 29:41? so Naelleul. ma
Teutonllhoovu 28 594 402-933 60 Metal. I903
4 no ao-ttso 50 W a :1. I973
Testicular». 20 594 Mo”? 60 Naelleul. I903
Tantra-rm 20 014 593.1!“ 60 Noelle”. I903
mmnmm‘ 6 1320 too-1960 so Mud. 1973
7 two 7 so aniline et al. I963
vaccinating": :2 1700 mm to “mu-urns

 

‘mmwuomumr

“mam

’MgapmnfalO-mm.amdmmmmmmmu

‘Mmflaow.

(From USDA Agricultural Handbook No. 696)

 

 

 

 

 

 

 

 

 

 

W'm 5000M. 1.0mm
D I- 4 I- D
None Perception Moderate Seven 1.
u only ._.° I- I-
6 '1 1 ”P , n - 3 - 0
E . loamy .- .—
g ‘ - . V... ‘ ”MID - 2 p ‘ E
NW '0“ I" change In
5 d . .“l practicum m - - 3
2 " ...v'NeIouln I'M“ ? 1 I- 2
u ( production :ma _ _
O O 0
WWW
Figure 1. Behavioral and milk production responses to increasing current levels. Voltages

(right vertical axis) were estimated using a worst-case circuit impedance (500 ohms)
and a more realistic impedance (1,000 ohms).

(From USDA Agricultural Handbook No. 696)

The detailed literature review which some of this summary is based

upon follows next.

Appleman and Gustafson (1985) classified the observed effects of

stray voltage on dairy cows into four general areas: effect on milking

performance and behavior; effect on herd health; effect on nutritional

intake; and, effect on production. Nevertheless, it was reported that any

negative effects of electrical shock on milk production or mammary

gland health most likely were not directly related to shock, i.e.,

9

physiological responses to shock were minimal and milk yield was
generally maintained at normal levels during the shock period.
However, the severe behavioral responses to shock usually caused
management problems. In addition, the degree to which milk
production would be affected depended on how dairy producers would
deal with the abnormal behavior.

Gustafson et al. (1985) indicated that the effect of a stray voltage
on a dairy cow was influenced by many factors: (a) voltage magnitude
and waveform; (b) the resistance of a cow’s body pathway; (c) condition
of concrete, soil and metallic conductors affecting resistance to "true
earth”; (d) resistance of the cow’s contact points; (e) resistance of the
electrical pathway to the cow’s contact points; and (f) impedance of the
source. When combined, these factors determine the current flow
through the cow’s body.

It is known that a current above some level flowing through the
cow’s body directly affects the cow’s behavior or health. A voltage
drives the current in the circuit with the cow body pathways as
resistances.

Norell et al. (1982) reported the electrical resistance data for
eight pathways through a dairy cow. The mean resistances of a cow
ranged from 359 to 877 ohms, with the lowest for a mouth-all hooves
path. Some pathways were selected in order to model the common cow
conditions on farms. For example: the pathway of front-to-rear hooves
can be used to model the cows that stand or walk across an area of the

barn or milking parlor where a floor voltage gradient exists; the pathway

10

of mouth-to-all hooves was to model the cows that bridged the gaps
between metallic feeders or water bowls and ground; and the pathway
of body-to-all hooves was to model the cows that bridged the gaps
between metal pipeworks connected to the grounded neutral systems
and concrete floor or earth. They found that the hoof-to-earth contact
resistance was a large component of the pathway resistance but they did
not evaluate the hoof-to-earth contact resistance quantitatively. So far,
there has been little research reported on the quantitative analysis of
animal contact resistance.

Aneshansley and Czarniecki (1990) measured the complex
impedances in Holstein cows (lst through 4th lactation) between 10 and
100,000 Hz. They reported that a circuit model developed for humans
appears to be appropriate for cows. Currents delivered at frequencies
about 1000 Hz were well above perception levels at 60 Hz but caused no
behavioral response. ,

Gustafson et al. (1985) studied 3 dairy cow-body pathways’
behavioral sensitivity to DC and 30 seconds on-off AC current. They
reported that the pathway front-to-rear hooves response rate (ratio of
responding cows to total test cows) became statistically significant above
2.0 mA AC. Mouth-to-all hooves response rate became significant
above 2 mA AC and 4 mA DC. Response rate for a body-to-all hooves
pathway with currents from 0 to 7.5 mA AC and 0 to 9 mA DC were
inconclusive. Appleman and Gustafson (1985) reported for AC 60 Hz
source, the cow’s behavioral sensitivity threshold currents are different

with the different body pathways, ranging from 0.75 to 7.1 mA.

11

Gustafson et al. (1985) reported that the cow’s behavioral sensitivity
threshold to DC current was about 30 percent higher than AC.

Drenkard et al. (1985) did experiment which exposed six
multiparous non-pregnant Holstein cows to electric current to assess its
effects on cow’s behavior, health, milking performance, and endocrine
responses. Three treatments (0, 4 and 8 mA) were applied in a
changeover design over three consecutive one week periods. They
concluded that behavioral responses to regularly applied AC current
treatment decreased in frequency and intensity with time. Changes of
milking performance and milk composition were not significant.
Changes of milking related cortisol responses during 8 mA current
experiment were significant. Oxytocin release was delayed during 8 mA
treatments. Current treatments did not affect prolactin.

Gorewit et al. (1984) summarized that hormones play critical roles
in regulating milk synthesis and milk removal. Oxytocin is a hormone
stored and released from the posterior pituitary and is responsible for
milk ejection of cows. Oxytocin, prolactin (PRL) and cortisol are
normally released during milking in cows.

Aneshansley et al. (1987) reported their experiment which had
fifteen first-calf heifers and fifteen 2nd to 4th lactation cows exposed to
five voltages (0-4 volts) while drinking. Exposure was continuous for 21
days. They reported no significant difference in water consumption,
feed intake, milk production, or concentration of fat or protein in the
milk. Drinking behavior (number of drinks / day and time /drink) did

change significantly.

12

Aneshansley et al. (1990) and Gorewit et al. (1990) did the
controlled study to 40 Holstein cows with stray voltage. Three groups of
10 Holstein cows (2nd-5th lactation) were exposed to voltage (1, 2, or 4
V) at waterers over their full lactation. A fourth group received no
voltage. They reported that these voltages applied did not influence
animal health or reproductive performance. They also reported that no
significant reduction in milk production or milk quality was found when
comparing "no-voltage" cows to cows that received voltage. Water
intake was significantly higher for the cows exposed to 4 volts.

Over the past years, most studies of the cows’ behavioral
sensitivity response have been conducted with continuous, fixed and
steady state current /voltage levels and various duration. However, one
area that has not been adequately studied is dairy cow sensitivity to
short duration currents.

Currence et al. (1987) used three durations ( 1, 10 and 100 cycles
at 60Hz) of AC electrical current with the left front-left rear hoof
pathway to determine the magnitude of short bursts of 60 Hz current
that were required to cause a physical reaction in dairy cows. The
conclusion was that the magnitudes of 60-Hz AC currents required to
cause dairy cows to respond were essentially the same for current
duration of 1.67 and 0.17 seconds (100 and 10 cycles at 60 Hz) with a
value equal to about 3.5 mA (rms). The magnitude required for current
duration of 0.017 second (1 cycles at 60 Hz) was approximately 50%
higher. The similar tests were also given to twelve human volunteers to

determine the magnitude of short bursts of 60 Hz current that were

13

required to cause their "perception" and "equal level of discomfort".
The conclusion was that the difference in mean current magnitudes due
to current duration were statistically insignificant.

Gustafson et al. (1988) conducted the electric strength/duration
experiments to six Holstein cows by using the square current wave-form
(about 11 mA in height and 38 ms in width) to cows’ mouth-to-all
hooves pathways. They processed their experimental data statistically to
verify the model of Pearce et al. in 1982 and their own exponential
model. They concluded that the model of Pearce et al. did not fit their
experimental data as well as might be desired and their own statistical
model was better. Their model was: IS = 11.02 x (0°16. Where Is was
the current strength of the stimulation needed to evoke the response in
mA and t was duration of stimulation in mS. In the empirical formula
above, the time ranged from 0.1 to 300 m8 and the current strength
ranged within 3 to 14 mA.

Reinemann et al. (1994) conducted research to address the
concern over possible effects of transient voltages and magnetic fields
on dairy cows. Dairy cow behavioral response to transient voltages and
magnetic fields was observed. The wave forms of the transient voltages
applied were: 5 cycles of 60 Hz AC with a total pulse time of 83 m8; 1
cycle of 60 Hz AC with a total pulse time of 16 m8; and 1 cycle of an AC
square wave (spiking positive and negative) of 2 ms duration.
Alternating magnetic fields were produced by passing 60 Hz AC
fundamental frequency with 2nd and 3rd harmonic and random noise

components in metal structures around cows. The maximum magnetic

14

field associated with this current flow was in excess of 4 gauss. They
reported that a wide range of behavioral sensitivity to transient voltages
was observed among cows. Behavioral response levels from 24 cows to
each transient exposure were normally distributed. No behavioral

responses to magnetic fields were observed.

2.2 Possible Sources of the Farm Neutral-to-Earth Voltage

Surbrook and Reese (1981) defined the terminologies regarding
the farm stray voltage as follows: (1) Neutral-to-Earth Voltage -- A
voltage difference measured between the neutral of an electrical system
and the earth. Metallic structures and equipment bonded to the neutral
will also be at a difference in potential from the earth. (2) Transient
Voltage -- A voltage that is not constant. It can be a sudden voltage
spike or a gradual rise and fall of the voltage. This voltage is usually
measured between earth and the neutral. (3) Tingle Voltage -- A term
sometimes used to describe a very slight voltage that causes a slight
shock or tingle when encountered by a human. A 120 V supply may
only cause a tingle if the person is well insulated. (4) Stray Voltage -- A
general term often used to include all sources of voltages found on the
farm that may be encountered by humans and animals between a metal
object and the adjacent earth or floor.

Surbrook and Reese ( 1981) traced the origins of farm stray
voltage from two sources -- on-farm and off-farm. They indicated that

common on-farm stray voltage sources were: (1) ground faults on the

15

farm; (2) voltage gradient across the ground or floor arising from wires
faulted in the earth; (3) electric fencer wires shorting direct to
equipment or inducing a charge in pipes and equipment; (4) grounding
conductor intentionally used as a neutral and a neutral used as a
grounding conductor; and (5) voltage drop on the secondary neutrals.
Typical off-farm stray voltage sources were: (1) voltage drop on the
primary neutral; (2) a ground fault on a neighbor’s property; and (3) a
fault in primary equipment or a problem with primary grounding.

Bodman et al. (1981) surveyed approximately 100 dairy farms in
the State of Nebraska. They reported over 58% of the survey farms
have the stray voltages in excess of 0.5 V and found that the primary
causes of the stray voltages were poor electric system installation and
maintenance procedures.

Gustafson and Cloud (1982) indicated that in the field, several or
possibly all of these sources would interact. However, unless the
contribution from each source could be clearly distinguished and
analyzed, successful diagnosis was difficult. So, a good understanding of
the sources and their interaction, the electric nature of the problem, and
the effects of the electrical characteristics of the system were important
to proper diagnosis and solution.

Gustafson and Cloud (1982) further proposed the solutions to
stray voltage problems: (a) eliminate or minimize the voltage causing
the problem; (b) isolate the voltage from any equipment in the vicinity
of all potential animal contact points; or (c) install an equipotential

plane that will keep all possible animal contact points at the same

16

potential. The solution or solutions selected depends on (a) the source
or sources of the stray voltage; (b) the magnitude of the stray voltage;
(c) the cost of alternative solutions; (d) the physical facilities involved;
and (e) the policies of the power supplier. The solutions can be
relatively simple if the problem is clearly diagnosed and the alternatives
evaluated and explained to the farmer.

Several researchers have contributed to the identification of the
sources of the neutral-to-earth voltage in the past years based on
experimental circuit measurement and circuit calculation by computer.

Stetson et al. (1984) developed an analog model of the
neutral-to-earth voltage in a single-phase distribution system. They
used resistors, conducting wires, switches, and a DC power source to
build up the physical circuit experiment model of a single-phase
distribution system. Their analog used the DC power source to simulate
the substation transformer and resistors to simulate the load
transformers. The analog assumed that each connection between the
neutral conductor and earth ground could be represented by a
resistance between the conductor and true earth. True earth had zero
resistance, and zero potential difference existed between any two true
earth ground connections regardless of their physical spacing. By taking
physical measurement of their circuit model, they revealed the
neutral-to-earth voltage phenomenon associated with multi-grounded-
single-phase distribution systems.

Their analog could be used to demonstrate the effects on primary

lines of magnitude and location of loads with respect to the substation,

17

poor neutral connections, and poor grounds. On the secondary lines the
effects of poor connections, poor grounds, and undersized neutrals
could be illustrated with two farmstead loads. These loads could be
connected to illustrate the effects of loads in-phase and out-of-phase
with the primary line, and the influence of one farm on another. The
effect of separating primary and secondary neutrals could be illustrated.

They reported their six demonstrations: (1) The effect of primary
conductor resistance on the neutral-to-earth voltage demonstrated the
influence of neutral wire size. The voltage increased as neutral
conductor resistance increased. For the isolated line segment
represented by the analog, the voltage was a maximum at the ends of
the line and minimum at or near the center. (2) The effect of a high
resistance at some location along the primary or secondary neutral
conductor demonstrated the effect of a high-resistance splice or
connection. The voltage drop across the high-resistance splice or
connection was much higher than across any of the other neutral
segments. (3) The effect of different neutral-to-earth resistance on
neutral-to-earth voltage demonstrated the effect of good and poor
grounding. Ground resistance changes could cause the neutral-to-earth
voltage changes, but the effect was location-dependent. (4) The effect
of connecting a farmstead neutral at different locations along the
primary neutral demonstrated that the farm neutral-to-earth voltage
could be location-dependent as well as farm-load-dependent. With no
secondary neutral current, an off-farm source was responsible for the

farm neutral-to-earth voltage. With both on-farm neutral and an

18

off-farm source, the phase relationship between the primary and
secondary neutral currents and the line location affected the voltage on
farm. (5) With identical neutral current at each farm and farms at the
different taps, the interaction magnitude decreased as the distance
between farms increased. (6) The effect of connecting or disconnecting
the farmstead neutral from the primary neutral demonstrated
connecting the primary and secondary neutrals sometimes reduced and
sometime increased the measured neutral-to—earth voltage. They
summarized that to solve the problem of excessive neutral-to-earth
voltage, an orderly approach was essential because of the difficulties of
analysis necessary to identify its source.

Kehrle (1984) developed the DC circuit model of a 7.2 kV
single-phase distribution power system to facilitate the analyses of the
neutral-to-earth voltage on farm. The computer program of circuit
simulation was used to perform the theoretical calculations and analyses
of the neutral-to-earth voltage along the distribution line. Eight
conclusions were reached. (1) The results obtained from the theoretical
analyses showed many similarities to the measurement results of the DC
physical analog model of Stetson et al. in 1984. (2) The effect of a
neutral/ transformer grounding resistance was location-dependent. A
significant reduction of the grounding resistance at a specific point
resulted in a significant reduction of neutral-to-earth voltage at that
point, and to a less extent elsewhere in the system. Also, the resistance
changes made near the substation did not affect the general

neutral-to-earth voltage over the entire line. However, when changes

19

were made at the end of the line, net voltage changes at that point were
maximized. (3) The effects of a primary neutral conductor resistance
was also location-dependent. Increasing the resistance of the neutral
conductor in a segment located at the beginning of the line resulted in
an increase of the neutral-to-earth voltage at all points along the line.
Increasing the resistance in a segment located at the middle of the line
resulted in a decrease of the neutral-to-earth voltage at some location
ahead of that segment in the direction of the substation. An increase of
neutral-to-earth voltage resulted at all points behind that segment in the
direction toward the end of the line. However, an increase in the
conductor resistance in a segment located at the end of the line did not
have a significant effect along the line. (4) Changes made on the
secondary neutral conductor resistance did not affect the primary
neutral-to-earth voltage along the line. (5) Changing the load at one
location resulted in a change of neutral-to-earth voltage at that point,
but the change was much less significant as the distance increased away
from that point. (6) The effect of varying the resistance of the
substation grounding mat was more apparent at the substation and
adjacent locations. (7) The effect of a primary ground fault was found
to be location-independent. A sustained primary line to ground fault
near the substation, or at the middle, or at the end of the line had the
same effects on the neutral-to-earth voltage along the line. (8) The
effect of a secondary ground fault was found to be location-dependent.
The effect of a sustained secondary line to ground fault at a location was

more apparent at that location than at the adjacent locations.

20

Surbrook et al. (1986, 1987) reported that the stray voltage was
caused by voltage drop and ground faults and might have its origin in
certain parameter changes of the primary electrical distribution system
or in the customer’s secondary electrical system. The rms value of the
neutral-to-earth voltage along a primary distribution line might be at a
value of zero some distance from the substation, depending on the
condition of the conductor resistances and of the loads.

Gustafson (1985) used a computer program to find the
neutral-to-earth voltages in a DC circuit model. Some of his findings
revealed basic parameters that caused a rise in neutral-to-earth voltage
levels. Three of his findings are as follows: (1) Increased neutral wire
resistance resulted in the largest percentage effect (on the
neutral-to-earth voltage) in the central portion of the distribution line.
(2) Deterioration of splices in the primary neutral conductor can
dramatically change the apparent resistance of the conductor. When
the high resistance splice was near the substation, the increase of
neutral-to-earth voltage was largest on those farms on the substation
side of the high resistance. The highest value, at the end of the line, was
not changed significantly by a high resistance near the substation. When
the additional resistance was placed near the midpoint of the line, some
farms ahead of the poor connection had reduced levels of the
neutral-to-earth voltage, while those further along saw increased levels.
(3) As the substation grounding resistance was increased, the voltage

near the substation increased and the zero point along the line was

21

moved further down the line. The effect diminished with distance from
the substation.

Dick and Winter (1987) presented computer generated profiles of
primary neutral-to-earth voltage of a 7.2 kV electric power distribution
system under secondary heavy load condition. Their simulation outputs
showed secondary heavy load increased the local neutral-to-earth

voltage.

2.3 Field Test and Diagnosis of Farm Neutral-to-Earth Voltage

Beside theoretical analyses and computer simulations of the
neutral-to-earth voltage problem, several researches have made
contributions to the field testing and diagnosis of farm neutral-to-earth
voltage in the past years.

Sodcrholm (1982) discussed possible sources of stray voltage,
measurement techniques to determine their cause, and corrective
measures that could be applied. He indicated that proper choice of a
meter or recording device for measuring stray voltage was essential if
misleading indications were to be avoided. He suggested 6
specifications for voltmeters used for these measurements. (1) The
meter scale should be such that AC voltage levels of 0.1 to 1.0 V can be
observed. (2) The meter must be capable of separating AC and DC. (3)
A low input impedance equivalent to an animal mouth-to-hoof
resistance (approximately 300 ohms) should be used to evaluate

voltages with a low source impedance and avoid misleading

22

measurements due to stray pick-up. (4) The meter should be capable of
high impedance input (1 megohm or greater) for use in measuring
induced voltage. (5) The speed of response should be fast enough to
give an indication of transient voltages. (6) In addition to a voltmeter,
other measurement tools such as graphic recorder, oscilloscope or
clamp-on ammeter capable of measuring current levels from 1 mA to 20
A have been found valuable in monitoring and determining causes of
stray voltage.

Stetson et al. (1982) reported a test method developed to
electrically evaluate an operating primary neutral connector, or splice.
Their method required only a digital voltmeter and was not dependent
upon a constant or given line load. Connections could be evaluated
both before and after adjustment or replacement. They described
specific step by step measurement procedures as follows. (1) PrOper
safety procedures should be followed. (2) An insulated aerial lift was
the safest and quickest method to reach the connections in question.
(3) All connections and tests should be made using insulated electrical
safety gloves. (4) A digital voltmeter was preferred since precise,
low-level voltage readings were more quickly and easily obtained. (5) A
clamp—on ammeter could be used to determine the level and monitor
any changes in current. (6) Attach the voltmeter leads to the conductor
near the connection with leads approximately 1/ 3 meter apart and note
the readings. (7) Move to one side or the other of the splice and take a
voltage measurement across a 1 / 3 meter section of the wire without 3

splice. If the voltage reading across the splice was greater than the

23

voltage reading on the wire without the splice, then there is resistance in
the splice which should be eliminated. (8) In all cases prior to installing
any connector, dirt or corrosion on the conductors must be removed
with steel brush, steel wool or emery cloth. They reported their
procedure had been verified by field tests and worked well on primary
neutral or secondary neutral lines with bare conductors.

Surbrook et al. (1988) developed a stray voltage diagnostic
procedure with a minimum number of measurements. They described
the minimum instrumentation needed was voltmeter, reference ground
rod, wire and a 120 V load. Their procedure was: (1) Start at the main
meter location or transformer pole. Measure between the transformer
ground wire and a reference ground at least 15 feet away and not under
the primary right of way. (2) Make the voltage measurement near the
building or area on the farm where the voltage was suspected to be
affecting the animals. (3) Determine if there was an excessive voltage
drop on the neutral wire to any building on the property. (4) Check
ground faults on the farm. If other sources were not found in steps 1
and 3, then it was possible that a ground fault was present. (5) The
power supplier line crew would Open the bond between the primary and
secondary neutrals at the farm transformer. At this point, telephone
personnel and cable television personnel must check the grounding of
their lines at the farm to make sure they did not themselves form a bond
from the farm neutral to the electrical power supplier primary neutral.
This was because that if the previous tests did not lead to an

identification of the neutral-to-earth voltage, then it was possible that

24

the source may be due to a secondary ground fault some place other
than on the farm.

Prothero et al. (1988) studied the primary neutral-to-earth voltage
levels impacted by various wiring system treatments through taking field
measurements. The test circuit that they chose was a radial two phase
3-wire tap, from a three-phase 4-wire multi-grounded neutral feeder
beginning approximately 2.5 miles from the source substation. Through
analyzing the field measurement data, they reported that the present
preferred method of solidly bonding primary and secondary neutrals
was consistent with the goal of minimizing primary neutral-to-earth
voltage on rural feeders. And the large amounts of supplemental
primary neutral grounding, as well as load balancing and line
reconstruction accomplished during their investigation, could not

reduce the primary neutral-to-earth voltage to zero.
2.4 Mitigation of Farm Neutral-to-Earth Voltage Problem

Gustafson (1985) proposed three possible approaches to mitigate
neutral-to—earth voltage problem on farms. The first recommendation
was voltage reduction -- by either elimination or reduction of the voltage
source (e.g., by removing bad neutral connections, improving or
correcting wiring and load balancing), or by active suppression of the
voltage by a nulling device. (2) When the voltage could not be reduced
to acceptable levels, the suggestion was gradient control -- by use of

equipotential planes and transition zones to maintain the anirnal’s step

25

and touch potential at an acceptable level. (3) Isolation of a portion of
the grounding or grounded neutral system accessing the animals, so that
they will not be subjected to objectionable currents due to stray voltages
existing on the remainder of the grounded neutral system.

From the literature (USDA, 1991), the term "isolation" is used to
mean the electrical separation of all or a portion of the grounded
neutral system of a farmstead from the remainder of the primary
neutral system or the remainder of the farm electrical system. Isolation
of part of the grounded neutral system can prevent neutral-to-earth
voltage on the non-isolated portion of the system from accessing the
animals. Isolation can be accomplished on a conventional
multi-grounded system (1) ahead of the farm main service (whole farm
isolation); or (2) at the livestock building (single service isolation). The
whole farm isolation can be accomplished (1) by isolation at the
distribution transformer or (2) with an isolation transformer following
the distribution transformer. If a satisfactory solution can be obtained
by isolation of a single building service, an isolating transformer can be
used. Depending on farmstead load, the transformer for the single
service can be smaller and less expensive than a transformer for the
entire farmstead.

Cloud et al. (1987) summarized the available devices for neutral
isolation at the distribution transformer. They indicated that three
methods have been developed for isolating the primary and secondary
neutrals at the distribution transformers. They make use of (1)

conventional spark gap; (2) a saturable reactor; or (3) a solid state

26

switch. These methods provide a high impedance interconnect below a
specified threshold voltage, and a low impedance interconnect when the
voltage exceeds that threshold. The "high" impedance is relative to the
grounding impedance of the isolated secondary system. The "low"
impedance provides that, under any condition creating a primary to
secondary voltage above the threshold level, the device impedance will
be reduced to a value such that the neutrals are essentially bonded.
Surbrook et al. (1989) proposed a number of mitigation
techniques to reduce farm neutral-to-earth voltage. They indicated that
common mitigation techniques can be applied once the sources were
positively identified. On-farm sources usually responded to one or
more of the techniques described in their report. (1) Elimination of
resistance at splices and terminations of the neutral conductors on the
farm was necessary when this was found to be the source. (2) Increasing
neutral conductor size may reduce the voltage. (3) Reducing the length
of feeder conductors to a building was effective during initial layout of
the farm buildings. (4) Balancing the 120 V loads in a building to
maintain neutral current at or near zero was important. (5) A four-wire
feeder, separating the neutral from the equipment grounding conductor,
may be installed to a building where animals may be affected. (6)
Elimination of the interconnections between neutral conductors and
equipment grounding conductors in a building. (7) Providing all
electrical equipment with an equipment grounding conductor that was
continuous from the equipment to the grounding bus of the circuit

supply panel increased the safety and reduced the chances of

27

neutral-to-earth voltage. (8) Elimination of any fault in equipment or
wiring, or any wiring that potentially could cause a fault from an
ungrounded conductor to the equipment or the earth was extremely
important.

Surbrook et al. (1989) also indicated that off-farm source
mitigation normally handled by the power supplier included: (1)
separating the primary and secondary neutral conductors; (2) repairing
corroded neutral conductor splices; (3) increasing the neutral conductor
size; (4) increasing the primary line operating voltage; (5) reducing
neutral-to-earth resistance at one or more locations along a distribution
line; and (6) elimination of ground faults on the primary system, or at
another customer location.

They indicated that there were some mitigation techniques that
were an attempt to lower the cow contact voltage regardless of the
source. These mitigation approaches included: (1) Equipotential planes
installed in the floor of milking areas, at watering devices and feeding
areas were intended to put everything within reach of the cow at the
same electric potential so there would be no contact voltage. (2)
Installing an active suppression device counteracting the
neutral-to-earth voltage condition may be effective where the voltage
could not be eliminated.

Althouse et al. (1990) proposed a circuit model to evaluate the
effectiveness of neutral isolation at the distribution transformers on
neutral-to-earth voltage. Figure 2 shows this model. In Figure 2, RD is

the primary system source resistance, Rs’s are the resistances of the

28

primary ground rod at the power supply transformer, Rl’s are resistances

of the isolated secondary ground rod, RF’s are ground resistances of the

farm electrical system, and R A’s are resistance of electric pathway of the

animal body. When distance of neutral separation becomes farther and

farther, the position of line B-C in Figure 2 will move downward lower

and lower. The operations of this circuit model can be summarized into

three steps as follows (referring to Figure 2):

1.

With the farm grounding open, take the measurements of BS
(primary neutral-to-earth voltage), 135-] (the voltage between
separated primary-secondary neutrals), E1 (farm ground
neutral-to-earth voltage) and Rs (the primary source ground
resistance, RS = R31 + R52), and then use the voltage division
law to calculate the resistance R51 and R32 in the circuit model:
R51 = R5 x (E5111 / ES) (2.1)
and R82 = R8 X (E1 / ES) (2.2)

According to Kirchhoff’s Voltage Law (KVL), with farm
grounding closed:

Es=IrXRsl+IIXRl+El (2.3)
and R1=<Bs-E1-I.x Rso/Ii (2.4)
Where IT and I] are primary and secondary neutral-to-earth
currents, respectively; and R1 is effective secondary isolated

ground resistance.

29

3. Using resistance R51 and R1 calculated above, and taking
different levels of voltage measurements E1 and Es as well as
current measurements IT and 11 in this circuit model, to estimate

neutral-to-earth voltage on farm:
E1=ES-ITXRsl-IIXR1 (2.5)

DiStTibUfion Balatad

Neutral Es: Neutral
NEV
Source '7 l. I A

(.)HAD E

 

 

 

 

 

R31 8 RI FlFF RM
C
“o Es 51
R32 R" RH RM
H c. s
Earth

Figure 2 Circuit Model for Estimation of Efi'ect of Neutral Isolation on
Neutral-to-Earth Voltage (Althouse, 1990).

Althouse (1990) determined the circuit model for the connection
of an isolated farm electrical system with a grounded primary electrical

system, but he did not conduct experiments to quantify the effectiveness

30

of neutral separation as well as determination of the parameters which
influence the effectiveness of neutral separation.

Althouse and Surbrook (1990) installed and tested the different
transition designs for equipotential planes on farm. Figure 3 shows this
design. They reported that step potentials for a cow’s contact should be
reduced to below 1 volt. No transition design was needed up to 2 volts
on the equipotential plane. At higher voltages it was desirable to
extend bare copper wires out into the earth from the plane with the

wires deeper as the distance from the equipotential plane increases.

  
 

N

15"
at

_l.

AllwiresareAWG No.2
andareburiedintheeorth
ataslope oflftdOtt.

    
      

  

30mm

10ft.

Figure 3 Transition Design for Equipotential Planes on Farm [1, 2]

III. OBJECTIVE

The primary goal of this research was to develop effective
computer models to simulate and analyze the neutral-to-earth voltage
profile along a primary distribution line under some operational
conditions. The computer models developed will facilitate the power
suppliers’ identification of sources of neutral-to-earth voltage
originating from a power distribution system. The specific objectives

corresponding to the primary goal in this study were as follows:

1. To refine the network circuit models of both single-phase and
three-phase, neutral multi-grounded electric power distribution
systems with the electric parameters for the ground resistances

based upon actual field measurements.

2. To determine (theoretically with the refined circuit model) the
effect of the changes of neutral-to-earth voltage producing
parameters on the primary system ground impedance as observed

from a farm transformer location.
3. To simulate the refined distribution network models with the

computer program and analyze the computer numerical solutions

of the neutral-to-earth voltage changes arising from high

31

32

resistance segments in the neutral of the primary line, primary
heavy loading demand, substation and other ground resistance
changes, different density levels of neutral ground connection,
primary operating voltage change, secondary ground faults and

primary phase-to-earth faults.

4. Based upon the analysis of the computer simulation results, to
establish some guidelines for power suppliers to identify the
sources of neutral-to-earth voltage originating from the parameter

changes of the power distribution system.

The secondary goal of this study was to conduct the farm field
tests to investigate and analyze the effectiveness of isolation of
distribution primary neutrals from secondary neutrals on
neutral-to-earth voltage on a farm. Currently, isolation of neutrals is
one of the mitigation measures for neutral-to-earth voltage problems
when there is a significant off farm stray voltage source. The specific
objectives corresponding to the secondary goal in this study were as

follows:

1. To test and to investigate the primary neutral-to-earth voltage
gradient distribution near a single ground rod as compared with
the gradient altering due to two parallel ground rods acting as a

grounding system.

33

To investigate the extent of the electrical couplings of a secondary

ground system to the primary neutral ground system.

To investigate the impact of farm ground resistance on neutral

isolation effectiveness.

To propose an equivalent circuit of the isolated secondary ground
system to simplify and generalize the estimation of the levels of
farm neutral-to-earth voltage according to a minimum set of

electric parameters available from an isolated system.

IV. METHODOLOGY

4.1 Computer Simulations of Primary Neutral-to-Earth Voltages

In order to implement the primary goal of this study, that is, to
develop effective computer models to simulate and analyze the
neutral-to-earth voltage profile along a primary distribution line under
some operational conditions, the electrical network simulation program
SPICE running on the Michigan State University mainframe IBM-3090
computer was employed. Three kinds of power distribution circuit
models, simplified single-phase DC model, single-phase AC model and
three-phase AC model, were established and simulated by the SPICE
program. To facilitate the visualization and analysis of the simulation
results, the simulation outputs from the SPICE program were used as
the input data to the personal computer software SUPERCALC to
draw the corresponding neutral-to-earth voltage profiles of all the

models and case studies.

4.1.1 The DC Single-Phase Electric Power Distribution Model

A DC base model has been developed for simulation of a

neutral-to-earth voltage profiles along a distribution line. Figure 4

34

35

:83.
99:75 595 Euro? SEE—Ea... 28:89.0 351-9359. 5.. .052: .328 OD

 

 

 

m n s a n n. m m .. m m

G 6 s n

e a m m m m m m

“W

u 8101» . <10P1 . 100l1 - 10011 11 ‘Ht 1 i tidal i 1’11“" Oils” 1

. owe. me. as. as. News... a»... 88 ~ 52. a as.

n 3...; m~4¢¢ 03.3 :3
1111 l' ‘ 000“ 51.941 '

.4 292m

 

 

 

 

a. as. 3.. :2. a... 22. 3..

 

36

shows the circuit network for this single-phase electrical primary
distribution line system model. The parameter values in the DC base

model are shown in Table 2. The following scenario was modeled:

1. The distribution line to be modeled had a length of 4,267 meters
(14,000 feet). The line was 4,800 volts multi-grounded
single-phase distribution system and fed fourteen load loops
evenly spaced about 304.8 m (1,000 feet) apart. Half of the loads
were assumed to be residential and half of the loads were
assumed to be farms. Each residential load (RRL) was followed
by a farm load (RFL). Ground electrodes were positioned at each
farm load (RFG), at each residential load (RRG), and between
transformers including the substation, node 1 (RG). Between the
substation and the first grounding electrode there was one
additional grounding electrode (RG). In one of the simulations,
extra ground electrodes (REG) were added uniformly along the
primary neutral line to simulate the effect of increasing ground
rod number.

2. Along the primary neutral conductor line, from the substation to
the last ground rod, 57 equally spaced nodes were selected. The

node numbers are shown in Figure 4.

37

Table 2. Parameters for Base DC Circuit Model in Figure 4.

 

 

 

Ungrounded Neutral Earth Load
Wire Wire Ground Resistance
Symbol Ohms Symbol Ohms Symbol Ohms Symbol Ohms
RS 0.5
RU2 0.344 RDI 0.086
RGZ 25.0
RD2 0.&6
R03 25.0
RD3 0.086
REG4 IN K
R04 0.086
RRGS 10.0 RRLZ 3,200
RU4 0.344 9.05 0.086
REG6 1N K
RD6 0.086
RG7 25.0
RD7 0.&6
REGB 100 K
RD8 0.&6
RFG9 1.5 Rm 1,6N
RU6 0.344 RD9 0.086
REGIO IN K
RDIO 0.&6
RGII 25.0
RDII 0.&6
REGIZ IN K
RDIZ 0.&6
RRGI3 10.0 RRL6 3,200
RUB 0.344 RDI3 0.&6
REGI4 IN K
- RDI4 0.&6
RGI5 25.0
RD15 0.&6
REG 16 IN K
RDI6 0.&6
RFGI7 1.5 RFLB 1,6N
RU 10 0.344 RDI7 0.&6
REGIB IN K
RDIB 0.&6
R619 25.0
RDI9 0.&6
REGZO IN K
RD20 0.&6
RRGZI 10.0 RRLIO 3,2N
RU 12 0.344 RDZI 0.&6
REGZZ IN K
RD22 0.&6
R023 25.0
RD23 0.&6
RE024 IN K
RD24 0.&6
RFGZS 1.5 RFLIZ 1,6N
RUI4 0.344 R025 0.&6
REGZ6 IN K
RD26 0.&6
R627 25.0
RD27 0.&6
REGZS IN K
m 0.&6
RRGZ9 10.0 RRLI4 3,200
RU16 0.344 RD29 0.&6
REG30 IN K
RD30 0.&6
R631 25.0
RD31 0.&6
REG32 IN K
R032 0.&6

RFG33 1.5 RFLI6 1,6N

 

38

 

 

 

Table 2 (cont’d)
Ungrounded Neutral Earth Load
Wire WIIC Ground Resistance
Symbol Ohms Symbol Ohms Symbol Ohms Symbol Ohms
RS 0.5
RUl8 0.344 R033 0.&6
R5634 1N K
RD34 0.086
R635 25.0
R035 0.&6
R5636 1N K
R036 0.&6
RR637 10.0 RRL18 3,2N
RUZO 0344 R037 0.086
R5638 1N K
R038 0.&6
R639 25 0

R039 0.&6
R040 0.&6

R5641 1.5 RFLZO 1,6N
RU22 0.344 RD41 0.&6
R5642 IN K
R042 0.&6
R643 2.5.0
R043 0.&6
R5644 1N K
R044 0.&6
RR645 10.0 RRL22 3,200
RU24 0344 R045 0.&6
R5646 IN K
R046 0.&6
R647 2.5.0
R047 0.&6
R5648 1N K
R048 0.&6
R5649 1.5 RFL24 1,6N
RU26 0.344 R049 0.&6
R5650 1N K
R050 0.&6
R651 25.0
R051 0.&6
R5652 1N K
R052 0.&6
RR653 10.0 RRL26 3,2N
RU28 0.344 R053 0.&6
R5654 1N K
R054 0.&6
R655 25.0
R055 0.&6
R5656 IN K
R056 0.&6

R5657 1.5 R5128 1.6N

 

39

3. Primary conductors employed for the model were number 2 AWG,
ACSR. The ungrounded conductor (phase line) resistance (RU)
between every two loads was 0.344 ohm (this was the actual
impedance Z value of 304.8 m of 2 ACSR and line inductance
effect was considered here). The neutral conductor resistance
(RD) between every two adjacent nodes analyzed was 0.086 ohm
(76.2 m of ACSR, line inductance effect included as just

mentioned).

4. The substation (node 1) ground mat resistance (RS) was 05 ohm;
each residential load neutral grounding electrode resistance
(RRG) was 10 ohm; and each farm load grounding resistance
(RFG) was 1.5 ohm. The additional neutral grounding electrode
resistances (R6) were 25 ohm. It was assumed that all grounding
electrodes were connected to the ”true earth” whose resistance was

2610.

5. At each residential transformer, the load resistance (RRL) was set
at 3,200 ohm to produce a 15 ampere primary load. At each farm
transformer, the load resistance (RFL) was set at. 1,600 ohm to give

a 3 ampere primary load.

In this simplified single-phase DC model, the distribution line system
was supplied by a DC voltage source equivalent to an actual AC source;

and the resistance load was equivalent to the transformer load. The

4O

neutral-to-earth voltage values shown were equivalent to rms values.
The negative DC values represented phase reversals in the AC network
model.

The previous research work in modeling the behavior of the
neutral-to-earth voltage along the power distribution systems frequently
has overestimated the values of the ground resistances for the
transformer ground connections. This has resulted in that the simulated
neutral-to-earth voltage values were always higher than the practical
field measured values. In order to overcome this problem, the electric
parameters for the ground resistances in this model presented in Table
2 were approximated from resistance values measured in a field
investigation conducted by Consumers Power Company. Table 3 lists
some typical values of ground resistances for farms and residences in

the service area of the Consumers Power Company.
4.1.2 The AC Single-Phase Electric Power Distribution Model

In verifying the effectiveness of the DC model, a single-phase AC
model with the 150/75 kVA, 4,800/ 120/120 V distribution transformer
model and the 60 Hz, 4,800 V (rms) sinusoidal voltage source as well as
the transmission line inductance effect was developed. Figure 5 shows
the AC network model. Compared with the single-phase DC model in
Figure 4 of the last section, in this AC model, the 60 Hz, 4,800 V (rms)
sinusoidal voltage source replaced the 4,800 DC source; the 150/ 75
kVA, 4,800/120/ 120 V distribution transformer model replaced each

41

 

 

 

 

:8»...

092.9, 595 829?. 5.55.56 28:80.0 3.23-0359. .8 .258 .323 U< .n 039m

ans angel]

1...... 3:3

’RFGS'I

R655

' REESE

K554

 

 

 

 

:3: .7

umv

 

ll

 

 

own; 823

amaz—

omv

 

 

 

 

 

 

 

 

 

 

  
 

 

~25 Nana] :3
mac... nan—8 Na.— NS—
5 ...q. a m 1 fl .- 5 fl 3
m m m m a a m m m a a
II . a. .
_1 u I: .I 35
£0.00

23.5.. a: 23 22 .2 .3 5. 8. .2... 8,.

 

42

Table 3. Some typical values of the ground resistances for the farm and
residential distribution transformers on the field in the service area of
the Consumers Power Company

TYPICAL FARM AND RESIDENTIAL GROUND RESISTANCE,
CONSUMER POWER COMPANY.

 

AJ .80 5.30
AIZ 1.60 4.70
AK .80 4.20
831 2.60 3.00
BW3 .90 .71
AE2 4.40 11.54
CW1 1.00 .61
CW2 1.20 7.70
CW3 1.10 7.87
NE1 1.10 9.90
NE3 .20 2.80
NES 1.00 1.53
NE6 1.30 8.10
N28 1.80 8.01
NW5 5.60 11.76
NW9 1.90 1.92
NW10 1.00 19.56
COUNT 3 17.00 17.00
AVERAGE I 1.66 6.42

STDV - 1.34 4.81

43

simulated load resistance; and simulated inductance was inserted along
each segment of both ungrounded phase and neutral conductors (XL/ R
= 0.394 on each segment). On the secondary side of each residential
transformer, a 4 ohm resistor was connected to draw 3 30A load. On
the secondary of each farm transformer, a 2 ohm resistor was connected

to create a 60A load.
4.1.3 The AC Three-Phase Electric Power Distribution Model

Some proportion of dairy farms are supplied by three-phase four-wire
wye connection power distribution systems. In order to simulate and
analyze the primary neutral-to-earth voltage behaviors in this situation,
the AC single phase electric power distribution model in Figure 5 of the
last section was expanded into the three-phase four-wire Yo/Yo
connection model. This is implemented by connecting the current
single-phase AC Model with two additional single-phase AC models.
The three single-phase models used here were structurally symmetrical
and identical to each other, except that: (1) they all shared one
common grounded neutral line; and (2) the phase angle differences of

substation source voltages of any two phases were 120 °. See Figure 6.

With the base models developed in the preceding sections, eight cases
were designed to simulate effects of some different operational
conditions on the neutral-to-earth voltage along the power distribution

line.

"I

0.:

Imaging;

.0!

.< 03:; E 0.0 >03 0:80 05 5 00.00580 3.0300 0.8 5053 -- 58:0.
8: 08 8088080: 855038 05 .8 O 08.8 0.8 m 00.0.8 08
0.020. b09808 0:. 6000.808 830800.08 05 @000. 8. .Aommxg
308099-038 £000 08 > .5va 80:»... 8:99.58 50:00:80
c>\o>v 8:00:80 0.9. 00.05.88 08590005 08 .0008 0800.0 U<

I:

   

0.585.. O

8.3. 8.3 0 02.:

g .m. w,

an 3.! .84 .3. .84?!

 

 

 

<33. <33 < 09...... < unit <23 3...! <2. <23. <23 < it

.0 88E

 

 

45

In the field, several operational conditions are usually superimposed
making the source identification of neutral-to-earth voltage difficult.
However, a clear and thorough understandings of these effects requires
an understanding of the individual contribution of each operational
condition. Only one parameter at a time in each of eight cases was

changed, and all other parameters remained unchanged.

4.2 Simulation of Abnormal Conditions along the Distribution Line

The base models of power distribution lines as described in preceding
sections were used for comparison with each of the abnormal line
conditions studied. The deviation of the neutral-to-earth voltage of the
test case as compared with the base model was considered as significant
criteria for evaluation rather than the actual magnitude of the voltage

per se.

4.2.1 Neutral Conductor Resistance Change

In order to study the neutral-to-earth voltage changes arising from
corroded and other poor connection (high resistance) conditions on the
primary neutral conductors, the neutral resistances or impedances in the
following individual segments of the DC circuit model in Figure 4 and
the single-phase AC circuit model in Figure 5 were changed from the
value 0.086 (normal, base case value) ohm to 1.0, 5.0, 20.0 and 100K

(open circuit condition) ohms:

1. The segment between node 9 and 10, RD9, near the substation of

each circuit model.

2. The segment between node 33 and 34, RD33, near the mid-point of

each circuit model.

3. The segment between node 56 and 57, RD56, at the end of each

circuit model.

4.2.2 Heavily Loaded Power Line on Farm

In order to study the neutral-to—earth voltage changes arising from
heavily loaded power line on farm, the primary load demands of 3 A
(normal at base case), 6 A, 9 A, 12 A and 15 A were simulated in the
following individual branch transformer locations of the DC circuit

model in Figure 4 and the single-phase AC circuit model in Figure 5:

1. The branch between node 404 and 9, near the substation of each

circuit model.

2. The branch between node 416 and 33, near the mid-point of each

circuit model.

47

3. The branch between node 428 and 57, at the end of each circuit
model.

The load demand changes in the DC circuit model of Figure 4 were
implemented by altering the corresponding primary branch load
resistance from 1,600 ohms (normal at base case), to 800, 533.3, 400 and
320 ohms, sequentially.

The load demand changes in the single-phase AC circuit model of
Figure 5 were implemented by altering the corresponding branch
secondary load resistances from 2 ohms (normal at base case), to 1, 0.67,
0.5 and 0.4 ohm, sequentially. This gave a secondary current of 60 A
(normal at base case), 120, 180, 240 and 300 A in sequence. The
primary / secondary current radio for 4800/ 120/ 120 transformer was
1:20, producing the corresponding primary load demands.

4.2.3 The Distribution System Ground Resistance Calculation

In order to evaluate the neutral-to-earth voltage changes due to the
ground resistance changes at a particular location, the system ground
resistance along the 4,800 V distribution line model in Figure 4 were
calculated through the following procedures:

1. Set the substation source voltage VI to zero.

48

2. Replace ground resistance at node 1 (substation), node 9, 33 and

57 (end of the line) with test voltage source, respectively.

3. Two different levels of test voltage were selected, 5 V and 10 V.
The corresponding test currents were obtained through the SPICE

simulation output.

4. The test voltage divided by the corresponding test current gives the

system ground resistance (Thevenin Equivalent Resistance).

The corresponding Thevenin Equivalent Open Circuit Voltages at
location of node 1 to earth (substation) and node 33 to earth were also
calculated in the DC model of Figure 4 through removing the ground
resistance RS and RF G33 and obtaining the neutral-to-earth voltage
values of node 1 and node 33, respectively from the SPICE simulation

output.

In order to determine the effect of the changes of neutral-to-earth
voltage producing parameters on the primary system ground impedance
as observed from a particular farm transformer location, the following

procedures were performed:

1. Each time respectively, increase the neutral line resistances from
0.086 ohm (base case) to 100 ohms (bad neutral connection) in the

immediate left segment next to simulated transformer load of the

49

circuit model, i.e., in the circuit model of Figure 4, the resistances
were increased from 0.086 ohm to 100 ohms in the neutral line
segments RD4, RD8, R012, R016, R020 R056, respectively.

2. Calculate primary system ground resistance (Thevenin Equivalent
Resistance) and Thevenin Equivalent Open Circuit Voltages at the
location of node 33 to earth as well as neutral-to-earth voltage at
node 33 in each of corresponding neutral segment RD settings
above, respectively.

4.2.4 Substation Grounding Resistance Change

The normal grounding resistance for the substation in this model is
0.5 ohm. The substation grounding resistance change effect on the
neutral-to—earth voltage was studied by increasing substation grounding
resistance RS from 05 ohm respectively to 1 ohm, 5 ohms, 10 ohms, 20
ohms and 40 ohms at the substation, node 1 for the DC model. in Figure
4 and the single-phase AC model in Figure 5. This was to model the
effect of a poor grounding connection at the substation on

neutral-to-earth voltage.
4.2.5 Neutral-to-Earth Resistance Reduction

The normal grounding resistance in both DC and single-phase AC

models for each residential transformer is 10 ohm and for each farm

50

transformer is 1.5 ohm. The neutral-to-earth resistance reduction effect
on neutral-to-earth voltage was studied by reducing the grounding
resistance RFG33 from 1.5 ohms to 1.2, 0.9, 0.6 ohm and then to 0.3
ohm. Grounding electrode RF G33 was at the middle of the distribution
neutral line (node 33) which was the fourth farm transformer location
from the substation shown in Figure 4 and Figure 5. This exercise was
to model how well a low resistance-to-earth electrode could lower a
neutral-to-earth voltage at a farm.

In the base case, the normal grounding resistance for each residential
transformer was 10 ohm, and for each farm transformer was 1.5 ohm.
There was one grounding electrode at 25 ohm between each
neighboring transformers. The effect of lowering the grounding
electrode resistance all along the distribution line was studied by
changing the extra grounding electrode resistance, REG, from 100
k-ohm to 25 ohm. This resulted in a simulation with a grounding
electrode of not more than 25 ohms at every pole along the distribution
line. The distribution line models are shown in Figure 4 for the DC
model, Figure 5 for the single-phase AC model and the base case values
of grounding electrode resistance are listed in Table 2.

4.2.6 Primary Operating Voltage Level Change
In the base case, the primary operating voltage level was 4.8 kV. The

effect of the change of primary operating voltage level on the

neutral-to—earth voltage was studied by simulating the other primary

51

operating voltage levels of 2.4 kV, 7.2 kV and 26.4 kV in the DC model

of Figure 4, respectively. The primary currents were set so that the
A secondary load currents for all these cases were maintained the same as
in the base simulation of 30 amperes for residential transformer
secondaries and 60 amperes for farm transformer secondaries. The
power consumed on the circuit remained the same as the base case.
The resistances RRL and RFL were changed to maintain the desired
secondary currents. Values of primary voltage, VI, load resistances RFL

and RRL, and primary current are listed in Table 4.

Table 4. Primary Current and Load Resistance, RRL and RFL, for Different

 

 

Primary Voltage Simulations.
Distribution Residential Load Farm Load
Voltage Pri. A See. A RRL Pri. A See. A RFL
2.4kV 3.0 30 SN 6 0 60 4N
4.8kV 1.5 30 3,2N 3.0 60 1,6N
7.2kV 1.0 30 7,211) 2.0 60 3,6N
26.4kV 0.27 30 96,8N 0.54 60 48,400

 

52

4.2.7 Secondary Earth Fault

The secondary circuit simulation for the DC model in Figure 4 is
shown in Figure 7. This was a 120/ 120 volt, single-phase, three-wire
grounded system with a balanced load. The secondary earth fault effect
on neutral-to-earth voltage was studied by introducing a 3, 6, 9 and 12 A
secondary ground fault through the resistor RF3 set at 40, 20 13.3 and
10 ohms from the ungrounded conductor of the secondary circuit to
earth. The faults of in-phase (only SW3 and SW4 closed in Figure 7)
and 180 ° out-of-phase (only SW3 and SW5 closed in Figure 7) on the
secondary side were respectively attached to: (1) node 9 which was the
farm transformer location nearest the substation; (2) node 33 which was
the fourth farm transformer located at the middle of the primary
distribution line; and (3) node 57 which was the farm transformer
location farthest from the substation.

Every transformer simulation in the single-phase AC model of Figure
5 has its own secondary circuit. The 3, 6, 9 and 12 A secondary earth
faults (in-phase and 180° out-of-phase at three different locations as
described above) were also introduced by connecting the corresponding
fault simulation resistance directly to the corresponding position of the

secondary circuit of the single-phase AC model.

53

.9 008.351.. :38 5.80
.8800... 5.3 50.0.? 8:25:36 30:80.0 08;..-0320. .2 3.88 .5020 Un—

 

       

5n:
an E a HE! 522:... :—

 

 

‘7‘."
s

o 9
3 2
r.- ..B
u m
. ’

s 050E

54

4.2.8 Primary Phase-to-Earth Fault

In order to analyze the neutral-to-earth voltage changes arising from
phase line to earth fault, first, the effect of the primary phase line to
earth fault size on the neutral-to-earth voltage was studied by simulating
a fault in the middle of the distribution line. This was implemented by
introducing a resistor RF2 from the ungrounded phase conductor node
416 to earth in Figure 8 for the DC model and the similar connections
for the AC single-phase model. The similar connection for the
three-phase AC model was shown in Figure 9 (RF2 connected from
node 416A of phase A to earth). RF2 was set at the values of 1,600,
800, 533.33 and 400 ohms to model the fault levels of 3A, 6A, 9A and 12
A, respectively.

The effect of this fault’s location on the neutral-to-earth voltage was
studied by moving the resistor RF2 above from node 416 (or node 416A
for three-phase model) to node 404 and to node 428 (or node 404A and
node 428A for three-phase model), respectively. Node 404 was the farm
transformer location nearest the substation and node 428 was the farm
transformer location farthest from the substation. At this time, RF2
was only set at 400 ohms to model a 12 A fault, the most serious fault

situation simulated.

55

00.0—38.0. .3... 580-208....
.008...— 53 80.0.»... 828:3... 80:80.0 03:90.35... 8.. .28... .3030 06 .w 0.89m

 

 

 

‘ n In
0 m s m a
...- u 0 0
mm 3. 8 8 .n

.2... 8.... ..- 030 .2...
6
39.300..- m W couuflunmam
0e.4-.e-ue no eo..ou .

u L<<<T l.<<<,l u

2.... 0.. 2....

RRGZS
R769

R8656

 

 

 

 

Nags «macs c.458 v-aac

4.3 N

 

’9’? 1' ‘ 'P’Pl {F
‘155 4 1‘1 ‘4‘5

 

”we. a. mam

-<
V
O
i
v
_
v
-
0
v

8.... a. 0:... 0

<9 19 3
a m m m s
"M n" "N w" 9“
<00 H 000fl <000fl 01F ..‘prdm <00p< thilfi’flpd WNWM <000 0<<<Wo0if 1w: 4 <1-M-901-ani4 -~m»fl-Ml~vw;
l... 0...... 42.. 0...... ~30 a... 8.... ~20 . 0.. . 9...... 2... m... 0.... Mtg . 2... ~
'I

56

.< 0.0.... 0. .0 >03 00.0. 0... 0. 00.000000
3.00.00 0.0 00.03 - 030... .00 0.0 ..00..0..00.. 00.05....6 00. .0
U 0.0.... 000 m 0.000 .0. .00.. b00080. 0... 60.0.0000: 0000.00.00.
0... 000.. 0 H 60.0.00... .30.. 5.00 0. < 0.000 5.3 Aow0._0>
00000-0780.... 0000 .0. > 8w... 0.0.... 00.05....6 302000000
0 .
>\0>. 000000000 0.? 0.5-.00. 0.0.3-003. .0. .0000. .3000 O< a 0.035

..v I I.“

“
‘4 ll:

'5 '23 ..3 ..a ..a ..3
a;

-
E" DDDDD '

ii‘i‘

an .09. .03 .3. .3. WIS-.1!

 

   

on:
0003. 005... 0

     

 

 

 

008500303 in!

IN

ONE 0.!

3... .r
<- . <05. .35.. c 02:. c 02.... 3.3 .35. Sam-3...... < 092.1

 

F

I put»

 

..w

 

57

4.3 Field Tests of Neutral-to-Earth Voltage Gradient

In order to implement the secondary goal of this study, that is, to
conduct the farm field tests to investigate and analyze the effectiveness
of isolation of distribution primary neutrals from secondary neutrals on
neutral-to-earth voltage on a farm, a series of tests were devised. These
tests included: (1) neutral-to-earth voltage gradient distribution near a
single primary ground rod; (2) neutral-to-earth voltage gradient
distribution near two parallel primary ground rods acting as a grounding
system; (3) neutral-to-earth voltage gradient distribution near one
primary ground rod and one isolated secondary ground rod when the
farm ground system was not connected (farm ground open); (4)
neutral-to-earth voltage gradient distribution near one primary ground
rod and one isolated secondary ground rod when the farm ground
system was connected with 5.24 ohms ground resistance; and (5)
discussion of the neutral isolation circuit model from literature
(Althouse, 1990). The test site was an actual farm located in
Williamston, NH and was supplied by 4,800/120/ 120 V electric power
distribution system.

To facilitate the analysis and visualization of the test data, the
personal computer program software SURFER was employed to draw
3-dimensional distribution and equi-potential contour lines of the tested
neutral-to-earth voltage gradients. The personal computer software
SUPERCALC and its spreadsheet were also employed to draw the
tested neutral-to-earth voltage gradient along some particular directions

and perform some data analysis.

58

4.3.0 Data Collection of Voltage and Resistance of Ground Rod

A test primary ground rod was located in a farm field at a distance
of not less than 200 feet from any known metal in contact with the earth
and connected to an electrical grounding system. A site in Williamston,
MI was chosen which was served by an ungrounded 4,800 volts
single-phase distribution system. This site was chosen so that earth
voltage gradients from other electrical systems would be negligible.

A voltage was established on the test primary ground rod by
passing a 60 Hz AC current from the test ground rod to a low resistance
return ground electrode system located 400 feet away from the test
primary ground rod. An isolation type variable transformer was used to
create this earth current at the test primary ground rod. This is
illustrated in Figure 10. The test ground rod was placed in sandy loam
soil with a depth to bed rock greater than 50 feet.

All measurements of earth voltage were taken from a point on the
test site to a reference ground which was located 400 feet south of the
primary test ground rod.

A digital Fluke 87, multimeter, with an internal impedance of 10
mega-ohms was used for all neutral-to-earth voltage gradient data
measurements. The AC Volt scale was selected. It has been verified
that the meter did not read DC quantities on the AC scale. A copper
probe for voltage measurement consisted of a AWG number 8 solid
copper conductor fastened to a wooden handle. The conductor

extended one inch below the bottom of the handle and the insulation

59

   
 

Test Forn N
Grounding
Systen

 

  

Residence

 

Drive

 

 

     
 
    

 

 

 

Test
Site

' Tronsforner
ole

Test Current
Return
Grounding Systen

60 Hz 120 v
Isolation Type
Vorioble Tronsforner

-* Reference

Si 1 ~Ph
Ground no e use Unqrounded

4000 Volt Prinory

 

 

0

Figure 10. A variable transformer was used to drive electric current into the earth

at the test primary ground rod.

 

 

 

 

 

0 5 10 15 20 25 30 35 4O
‘0 l i T l i r T ‘0
.35 r- _- 35
30 L. I O _- 30
DUO-30) COO-30)
A .
25 - 3.00.26) - 25
20 ‘- . - 20
RAN-20)
N 15 r- - 15
10 .— O O ..- lO
AUG-10) Baa-m)
5 ~ ‘ 5
O 1 l 1 L 1 1 1 O
O 5 10 15 20 25 30 35 40

Figill 11. The 40 x 40 Feet Square Shaped Test Plot for NEV Gradient Testing,
Primary One and Two Ground Rods

61

was removed from that portion. When taking the voltage measurement,
the probe was firmly placed against the soil and the meter was allowed
to stabilize before any voltage measurement was recorded. For some
earth voltage measurements, a solid AWG number 12 copper wire was
inserted approximately 0.5 cm into the soil. This technique was used for
all earth voltage measurements where the distance between
measurements was less than six inches.

A Megger Earth Tester (Null Balance, Battery Operated, James
G. Biddle Co.) was used for all ground electrode resistance data
measurements. The principle of this instrument was based upon The
fall-ofipotential method or 3-point method (Tag, 1964 and IEEE, 1983).
The distance between the test rod and the potential probe of the Tester
was 65 feet (19.8 m), with another additional 35 feet (10.7 m) distance
to the current probe of the Tester.

4.3.1 Test of NEV Gradient Distribution, Primary One Ground Rod

The 40 x 40 feet square shaped test plot is shown in Figure 11 in
which the direction expressed was consistent with conventional map
(upper - north, lower - south, left - west and right - east). The
coordinate system of the plot was established with origin at low-left
corner (south-west corner), positive X axis toward the horizontal
direction (east direction), positive Y axis toward the vertical direction
(north direction) and the length unit as foot. One primary electrode
rod Rs (connected with neutral-to-earth voltage source) was located in
the center (20, 20) of the plot and was driven 8 feet deep into earth.

62.

Within the inner square A (10, 10), B (30, 10), C (30, 30) and D
(10, 30), the earth voltage measurements were taken at every foot, both
horizontally and vertically. Outside the inner square ABCD, the earth
voltage measurements were taken at every 5 feet, both horizontally and
vertically.

From the center (20, 20) of the plot, along the four radial
directions of straight north, straight south, straight west and straight '
east, the earth voltage measurements were taken up to 100 feet
distance. The measurement incremental intervals were: every inch for
the distance from the center to 1 foot, every 3 inches for distance up to 2
feet, every 6 inches for distance up to 4 feet, every foot for distance up
to 10 feet, every 5 feet for distance up to 20 feet, and every 10 feet for
distance up to 100 feet.

4.3.2 Test of NEV Gradient Distribution, Primary Two Ground Rods

The layout configuration of this test was same as the previous test

in section 4.3.1, except:

1. One more primary ground rod (driven 8 feet deep into earth) was
located at 6 feet (1.83 m) away straight north from the original
one at center. The coordinate for this additional primary ground
rod was Rsa (20, 26). Both primary rods were connected with the

same neutral-to-earth voltage source.

63

2. The earth voltage measurements along the radial direction of
straight north were taken with the incremental intervals as: every
inch for the distance from the center to 1 foot, every 3 inches for
distance up to 5 feet, every inch for distance up to 7 feet, every
foot for distance up to 10 feet, every 5 feet for distance up to 20

feet, and every 10 feet for distance up to 100 feet.

4.3.3 Test of NEV Gradient Distribution, Primary One Ground Rod,
Secondary Isolated One Ground Rod Without Farm Ground

Connection

The 4 x 10 feet rectangular shaped test plot is shown in Figure 12
in which the direction expressed was consistent with conventional map
as described in section 4.3.1. The coordinate system of the plot was also
established with origin at low-left corner (south-west corner), positive X
axis toward the horizontal direction (east direction), positive Y axis
toward the vertical direction (north direction) and the length unit as
foot. One primary electrode rod Rs was located in the position (2, 2)
and one secondary isolated rod R1 was located in the position (2, 8).
The distance between two rods was 6 feet and both rods were driven 8
feet deep into earth.

Within this plot, the earth voltage measurements were taken at
every half foot, both horizontally and vertically.

More detailed earth voltage measurements were taken along the
connection line from the south side center (2, 0) to north side center (2,

10) of this plot. The measurement incremental intervals were: every 3

 

 

 

O 1 2 3 4
9— -9
a— 0 —8
size)
7. "7
+ 6" ..5
51- ‘5
4— -“
N .- --
0 2— R. -2
R_(-4,2) -(12)
1- -1
O l 2 3 4

 

 

Figure 12. The 4 x 10 Feet Square Shaped Test Plot for NEV Gradient Testing,
Isolated Primary and Secondary Ground Systems

65

inches for the distance from the south side to 1 foot, every inch for
distance up to 3 feet, every 3 inches for distance up to 7 feet, every inch
for distance up to 9 feet, and every 3 inches for distance up to 10 feet --
north side center (2, 10).

In this test, the primary rod was connected with a neutral-to-earth
voltage source, the secondary isolated rod was not connected with a
neutral-to-earth voltage source, and the farm ground system was not

connected.

4.3.4 Test of NEV Gradient Distribution, Primary One Ground Rod,
Secondary Isolated One Ground Rod With Farm Ground

Connection

From the layout configuration of the test in the section 4.3.3, an
isolated secondary ground rod was driven 6 feet north away from the
test primary ground rod. To create an actual earth circuit as exists on a
farm, seven ground rods were driven in an area 600 feet (183 m) to the
north of the test site. This farm grounding system was connected to the
secondary ground rod at the test site. By connecting different
combinations of ground rods, several levels of ,farm grounding
resistances could be achieved from infinite to as low as 5 .24 ohms. The

test procedure is shown in Figure 13.

 

 

 

 

 

 

 

 

60 H2
120 V Isolation Type
llllll, Variable Transforner
Tl ll]
? v v v 7 Test Prlnory Test Secondary 7 '
Test Current Ground R°d 5'0“"6 R°d Fara Grounding
Return Systen
Grounding Systen
400 Ft. 600 Ft.

 

 

Figure 13. A secondary isolated ground rod and farm grounding system were
established to an area away from influence of earth voltage gradients
from other electrical systems.

67

4.3.5 Verification of the Neutral Isolation Circuit Model

This circuit model was described in detail in Figure 2 and in
section 2.4, Chapter 2. The schematic diagram of the circuit model,
Figure 2, also showed the locations of the instrumentation to measure
the voltages and currents during the testing. The test plot was the same
as in Figure 12, described as in section 4.3.3 and 4.3.4, except that one
more primary ground rod RSS was driven 8 feet deep into earth at the
location (-4, 2), 6 feet away straight west from the other primary ground
rod Rs- The combinations of different farm ground electrode
connection schemes gave the farm ground resistance (R) levels at open
circuit, 23.7, 10.3 and 5.24 ohms, respectively. The animal simulation
resistors, RM and R,u were not connected.

The resistances of the original primary ground Rs, additional
primary ground rod Rss» isolated secondary ground rod R1 + R11 and
primary NEV source ground RD were measured. Their values were
18.6, 11.4, 22.0 and 6.2, respectively.

Sixteen sets of the tests were performed: (1) primary one rod,
primary NEV 10 V and farm ground open; (2) primary one rod, primary
NEV 10 V and farm ground resistance 23.7 ohms; (3) primary one rod,
primary NEV 10 V and farm ground resistance 10.3 ohms; (4) primary
one rod, primary NEV 10 V and farm ground resistance 5.24 ohms; (5)
primary two rods, primary NEV 10 V and farm ground open; (6)
primary two rods, primary NEV 10 V and farm ground resistance 23.7
ohms; (7) primary two rods, primary NEV 10 V and farm ground
resistance 10.3 ohms; (8) primary two rods, primary NEV 10 V and

68

farm ground resistance 5.24 ohms; (9) primary two rods, primary NEV 5
V and farm ground open; (10) primary two rods, primary NEV 5 V and
farm ground resistance 23.7 ohms; (11) primary two rods, primary NEV
5 V and farm ground resistance 10.3 ohms; (12) primary two rods,
primary NEV 5 V and farm ground resistance 5.24 ohms; (13) primary
one rod, primary NEV 5 V and farm ground open; (14) primary one
rod, primary NEV 5 V and farm ground resistance 23.7 ohms; (15)
primary one rod, primary NEV 5 V and farm ground resistance 10.3
ohms; and (16) primary one rod, primary NEV 5 V and farm ground
resistance 5.24 ohms.

In each set of the test above, the following physical quantities

were measured as shown in Figure 14:

(1) Es (primary NEV);

(2) I, (the earth current on primary side);

(3) II (the earth current on isolated secondary side);

(4) E51 (the voltage between separated primary-secondary neutral);
(5) E1 (isolated secondary NEV);

(6) ED (primary source NEV);

(7) E,- (isolated farm NEV).

Figure 14 is actually the same circuit as proposed by Althouse
(1990) in Figure 2. i

69

Isolation Type
Variable Transforner

 

 

 

 

 

 

 

120 V 60 HI
I ITIT I
b ED
(
Ru :5
Earth Earth
P""°'Y Systen Prinary Secondary Forn Systen
Resistance Ground Rod Ground Rod Resistance
at Transforner at Tronsforner

Figure 14. The physical quantities measured in the neutral separation circuit
model.

V. RESULTS and DISCUSSION

5.1 Results and Analysis of NEV Computer Simulation

The computer simulation results of the distribution systems were
analyzed to determine the level of the neutral-to-earth voltage
produced under various operational conditions.

Although the neutral-to-earth voltage profiles were presented as
if continuous, only the discrete voltage values at each of the 57 node

locations were meaningful.

5.1.0 Normal Operational Condition

The profile of the neutral-to-earth voltage along the distribution
line for the base case of the single-phase DC model is shown in Figure
15. The profile shows that the magnitude of the neutral-to-earth
voltage at the substation (9.2 V) was much higher than at the end of the
primary neutral line (4.9 V). Note that the magnitude of the
neutral-to-earth voltage decreased as the distance away from the
substation increased, reaching zero and experiencing a 180 degree

phase angle change at node 12. Then the neutral-to-earth voltage

70

71

BASE CASE NEV, 4.8K 1—PHASE LINE
FARM 8: RESIDENTIAL GROUND 1.5 8: 10 OHM, POLE 25 OHM

 

VOLTAGE (V)

 

 

 

 

IiiitiITTii—rriiiriitiiiirfliimiiitittrritTitiiriiiiri
'- in O in O in O in 3 in O iOI‘
v- v- N N t') 1') 1' in min

NODE # (1 -57) 9405271:-

Figure 15, Profile of the neutral-to—earth voltage along the primary distribution

line from substation (node 1) to 57th node for the DC single-phase
base model

72

increased as the distance from the substation became greater, leveling
off toward the end of the neutral line.

Along the multi-grounded neutral power distribution systems
which are commonly used in rural areas of the United States, some level
of neutral-to-earth voltage is always present; it is an inherent
phenomenon. Kehrle (1984), Gustafson (1985) and Surbrook et al.
(1986) also obtained theoretical profiles of neutral-to-earth voltage
similar to Figure 15 .

The neutral-to-earth voltage profile along the distribution line for
the base case of the single-phase AC model is shown in Figure 16. This
profile is about the same as the profile in Figure 15 obtained from the
DC model if only the magnitudes of neutral-to-earth voltages are
considered. Figure 17 shows the comparisons of two neutral-to-earth
voltage magnitude profiles obtained from the DC based model and the
AC based model. To make the DC model profile more comparable
with that of the AC model, the minus signs of the voltage values near
the substation in the DC profile were removed in Figure 17. Although
the curve position of the AC model was slightly higher than that of the
DC model, the differences between these two profiles were very small
and the shapes of these two profiles were almost identical. The profile
of the AC model did not experience the ”zero voltage value" point like
the DC model did,‘ the reason being that the neutral return currents
from the earth could not cancel each other out completely when the AC

model distribution line series inductance effect was taken into account.

73

AC BASE CASE NEV, 4.8KV LINE

 

 

 

 

 

 

10
8-
>
v 6‘
Lu
0
<
+—
_J
O 44
>
2-1
0 IIIIIIIII—ITIITITIITTIIIIIIITTITIIIII[III—ITVIWTIITITFY
-- In 0 In 0 10 o in o In o

105
v- s" N N n I") V’ V’ W 101.0

NODE # (1-57) 940526b

Figure 16. Profile of the neutral-to-earth voltage along the primary distribution

line from substation (node 1) to 57th node for the AC single-phase
base model

74

COMPARISON OF AC & DC BASE CASE NEV
4.8 W I—PHASE DISTRIBUTION SYSTEM

 

 

 

 

10,
Ii
80: DC
|.
A
> 6‘ H
LLJ
2 “ -‘-':-:'._- u -- - -l
.— "".":"""."'_;"
...l | ‘--
O 4‘ a.
> I;
‘ 4
|i
2.- p
l\ /,‘
\. .l
H n
0‘ -[t[lTTTIIIIIIIFITIIIIITIIITFITITIIIIIrrTI'II
'- In C n C If) C In C n o ‘0"
v- e- N N n l’) v * m an

NODE # (I -57) 940527d

Figure 17 . Neutral-to-earth voltage profile comparison between the simulations
of single-phase AC and DC models

75

In fact, under other operational conditions that have been
studied, the neutral-to-earth voltage profiles obtained from single-phase
DC model simulations were about the same as the profiles obtained
from the corresponding single-phase AC model simulations when only
the magnitudes (rms values) of the voltages were considered.

In the following sections, the effects of the different operational
conditions on the neutral-to-earth voltage were studied primarily
through the DC model. The corresponding single-phase AC simulation

profiles are provided in the the appendix for reference.
5.1.1 Neutral Conductor Resistance Change

Figures 18 through 20 present the neutral-to-earth voltage
simulation curves resulting from changing the neutral conductor
resistance RD9, RD33 and RD56 (referring to Figure 4) in three
different representative locations at normal load demand condition (3 A
load), as shown in Figure 21.

Figure 18 shows the the primary neutral-to—earth voltage profiles
with the variable neutral conductor resistance RD9 between the nodes 9
and 10 near the substation. When RD9 increased, the magnitudes of
neutral-to-earth voltages also increased accordingly in most segments
along the distribution line. Note the most significant increases of the
neutral-to-earth voltage occurred at node 12 which is near the high
neutral resistance segment and near the last customer load cut off from

the substation by the high resistance segment in the neutral. At this

76

RD9 CHANCE IMPACT ON NEV, 4.8K I—PHASE
RD9 0.086, 1, 5, 20 AND IOOK OHM, LOAD 3 A

 

 

 

 

 

 

6
w.’.1-l-l-l,l A . . , . - - .=.=‘:7_’Y:1...Y-
44 II 5=::==="’"JPL
2« 'r“
A - -------------------------------------------
>
V
LtJ
(é) —-i—- 0.086 OHM
5 -a—- 1.00 OHM
O
> —*— 5.00 OHM
+ 20.0 OHM
+ TOOK OHM
-12 r—rfilllTTTlflllllTlllIllllllllllllllllerTTTllllTllTllll
— In 0 in o to o in o 1.0 o In I\
.— .— N m n n v- v to In In

NDDE # (1 —57) 9405280

Figure 18. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 5 7th node with variable neutral conductor
resistance RD9 between node 9 and node 10 and a normal customer
load of 3A at node 9 (from single-phase DC model simulation)

77

POSS CHANCE IMPACT ON NEV, 4.8K I—PI—IASE
RD33 0.086, I, 5, 20 AND 100K OHM, LOAD 3 A

 

 

—+— 0.086 OHM
-a— 1.00 OHM
_6 q —~— 5.00 OHM
—v— 20.0 OHM
-a . + 100K OHM

VOLTAGE (V)

 

 

 

-10lrlllrllIrlIlTrlllllIlllTlIllllllrlllllllllllllrllIlllll

'- 1.0 O in O in O 10 O in

NODE # (1-57) 9405270

Figure 19. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with variable" neutral conductor
resistance RD33 between node 33 and node 34 and a normal customer
load of 3A at node 33 (from single-phase DC model simulation)

78

node, when neutral resistance RD9 was 0.086 Ohm (normal base case),
the magnitude of the neutral-tO-earth voltage was zero. When RD9
increased to 1.0 Ohm, the voltage was increased to 1.9 V. When RD9
increased 5.0 Ohm, the voltage was increased by 2.2 V, to 4.1 V. When
RD9 increased to 20 Ohm, the voltage was increased another 1.0 V, to
5.1 V. When RD9 increased tO 100 k-Ohm (open circuit simulated), the
voltage was only slightly increased by 3 additional 0.5 V, to 5.6 V.

Near the substation, a high resistance segment in the neutral will
usually increase the neutral-tO-earth voltage for all customers in the
local area.

Figure 19 shows the the primary neutral-tO-earth voltage profile
with the variable neutral conductor resistance RD33 between node 33
and 34 near the middle of the distribution line. The neutral-to-earth
voltage became lower at nodes toward the substation from the location
of the abnormal neutral resistance, and higher at nodes toward the end
Of the line. As the abnormal resistance RD33 was increased, the voltage
drop across this segment also increased. Note in the base case, when
normal neutral resistance RD33 was 0.086 Ohm, the voltage drop across
this segment was only 0.1 V. (45 V neutral-tO-earth voltage at node 33,
4.6 V at node 34). When RD33 was increased to 1.0 Ohm to simulate an
abnormal resistance in the neutral, 0.6 V voltage drop occurred between
nodes 33 and 34. The neutral-to-earth voltage decreased by 0.2 V at
node 33 toward the substation side, while it increased by 0.3 V at node
34 toward the end Of the line. When neutral resistance RD33 was
increased to 20 Ohm, there was a 1.4 V voltage drop between node 33

79

PBS-6 CHANCE IMPACT ON NEV, 4.8K I—PHASE
ROSS 0.086, I, S, 20 AND TOOK OHM, LOAD 3 A

 

E
Lu
0
<(
5 _+_
o 0.086 OHM
> ‘4 ‘ _9_
1.00 OHM
-6 q -*— 5.00 OHM
+ 20.0 OHM
-8 -I + TOOK OHM

 

 

llrllllllrllllIIIIIITIIITIIIIIITTIITITITTITIITIITIITI

— in o In o n o in o In
.— .— N N :0 n v- v I?) 8

NODE # (1-57) 940528b

!\
t!)

Figure 20. Profile of the neutral-tO-earth voltage along the primary distribution
line from substation to 57th node with variable neutral conductor
resistance RD56 between node 56 and node 57 and a normal customer
load of 3A at node 57 (single-phase DC model simulation)

80

and node 34. The neutral-to-earth voltage decreased by 0.5 V at node
33 and it increased by 0.8 V at node 34 as compared with the normal
RD33 (0.086 Ohm) condition. Note that the effect on neutral-tO-earth
voltage of an abnormal neutral conductor resistance diminished as the
distance from the neutral resistance RD33 increased. In the extreme
case, when RD33 increased to 100 k-Ohm (open circuit simulated), the
voltage drop across this segment reached 1.5 V. The neutral-tO-earth
voltage decreased by 0.5 V at node 33 and increased by 0.9 V at node
34. However, the neutral-tO-earth voltage had no change at the
substation and was only changed 0.2 V at the end of the line, as
compared with the normal RD33 (0.086 Ohm) situation (base case).
Figure 20 shows the the primary neutral-tO-earth voltage profile
with the variable neutral conductor resistance RD56 between node 56
and 57 near the end Of the distribution line. The neutral-tO-earth
voltage became slightly higher at nodes toward the substation from the
location Of the abnormal neutral resistance, and slightly lower at nodes
toward the end of the line. As the abnormal resistance RD56 was
increased, the voltage drop across this segment also slightly increased.
But this effect was very small. In the extreme case, when RDS6
increased to 100 k-Ohm (Open circuit simulated), the voltage drop across
this segment was only 0.9 V. The neutral-tO-earth voltage decreased by
0.5 V at node 57 and increased by 0.4 V at node 56. However, the
neutral-tO—earth voltage had almost no change from the node 50 Of the
line all toward the substation as compared with the normal RD56 (0.086

Ohm) situation (base case).

81

.008. < m. 80.2.8 08.8.. .08. .808
.0 00: 00. m00_0 .000000. 0>_.0.00.0.00. 50.0.2. 00.... 0. 0mg 000
30.. find 0000....0. 8.000000 0.0.00 00. M03005 0.0.. m0.._...0.
.0m0000 0m0=0> :..00-0.-_0....00 0... .30.. 0. 02.000000 000.00 00... AN 039%

 

02. mm... on.

.. 00oz mm .002 a 00.2 . ~00.

82

With the line segment RD56 open the current at node 57 is 2.94
amperes which is the farm load at that point. All of that current must
flow to earth at the farm which in the simulation has a ground resistance
of 1.5 ohms. The neutral-to-earth voltage at node 57 is therefore 4.41
volts. But with normal resistance in the neutral segment RD56, even
though a portion of the 2.94 amperes at node 57 flows along the line to
other grounding electrodes, more current is flowing to the ground at
node 57 from the adjacent line, resulting in a current at node 57 of 3.27
amperes. This results in neutral-to-earth voltage at node 57 of 4.9 volts.
Therefore, for this situation, the neutral-to-earth voltage at the end of
the line is lower when the line segment RD56 is open. This will not
necessarily be true in all situations.

Figure 18 through 20 indicate that high neutral resistances (poor
neutral connections) will cause the neutral-to-earth voltage changes
(increase or decrease) along the distribution line. These changes are
localized near the places where bad neutral connections occur. The bad
neutral connection near the substation will cause more significant
neutral-to-earth voltage changes than if near the middle or toward the
end of the distribution line. The reason for this is that the neutral line
segments near the substation collect more returned currents from the
network than the segments toward end of the line.

The profiles obtained from the single phase AC model
corresponding to Figure 18 through 20 are listed as Figure A1 through
A3 in the appendix.

83

5.1.2 Heavily Loaded Power Line on Farm

Figures 22 through 24 present the neutral-to-earth voltage
simulation curves resulting from the different levels of the heavy load
demand in three different representative locations along the power
distribution line.

Figure 22 shows the neutral-to-earth voltage simulation profiles
under the condition of the different levels of the heavy load demand
occurred at the transformer branch between node 404 and node 9 near
the substation (2000 feet from substation) of the single phase DC circuit
model.

Compared with the base case, when the load demand current was
increased from 3 A (base case) respectively to 6 A, 9 A, 12 A and 15 A,
for some locations near the substation, the neutral-to-earth voltage
actually decreased as the load was increased. And neutral-to-earth
voltage increased on other segments along the distribution line as the
load was increased. Note the most significant increases of the
neutral-to-earth voltage occurred at node 12 near the heavy load
demand took place. At this node, in the base case, when the load
demand current was 3 A, the value of the neutral-to-earth voltage was 0
V. When the load demand was increased to 6 A, the voltage increased
to 0.6 V. When the load demand was increased to 9 A, the voltage
increased by another 0.5 V, to 1.1 V. When the load demand was
increased to 12 A, the voltage again increased another 0.6 V, to 1.7 V.

When the load demand was increased to 15 A, the voltage reached 2.3

HEAVY LOAD IMPACT ON NEV, 4.8K 1—PHASE LINE
LOAD 3, 6, 9, 12 a: 15 A AT 404—9

 

10 ¢ 3A
—B~—6A

9A

12A

15A

VOLTAGE (V)

 

 

 

 

-157IIIIIIIITIIIIIlIllllIlllI—ITIITTIIIITIIIITIIITIIIIITIIII
.— n o n o in o In 0 II)"
n e — N N n n v- V' In 1010

NODE # (1—57) 9405ng '

Figure 22. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with variable load demand
conditions at node 404-9 (single-phase DC model simulation)

85

V. The neutral-to-earth voltage increased almost linearly with the load
demand current at a rate of 0.2 V/A near the high load demand node.

As the distance toward the end of the line node 57 increased, this
neutral-to-earth voltage increase was less and less significant. From the
middle toward the end of the line, these neutral-to-earth voltage
profiles were nearly unchanged as load was added.

’ Toward the direction of the substation, the neutral-to-earth
voltages decreased at some segments and increased at some other
segments, the increases near the substation were small.

Figure 23 shows the neutral-to-earth voltage simulation curves
resulting from the different levels of heavy load demand at the
transformer branch between node 416 and 33, near the mid-point of the
circuit model.

Compared with the base case, when load demand current was
increased from 3 A (base case) respectively to 6 A, 9 A, 12 A and 15 A,
the neutral-to-earth voltage also increased accordingly along the
distribution line. Note the most significant increases of the
neutral-to-earth voltage occurred at node 33 where the high load
demand took place. At this node, in the base case, when the load
demand current was 3 A, the value of the neutral-to-earth voltage was
4.5 V. When the load demand was increased to 6 A, the voltage
increased by 1.2 V, to 5.7 V. When the load demand was increased to 9
A, the voltage increased by another 1.2 V, to 6.9 V. When the load
demand was increased to 12 A, the voltage again increased another 1.2
V, to 8.1 V. When the load demand was increased to 15 A, the voltage

HEAVY LOAD IMPACT ON NEV, 4.8K l-PHASE LINE
LOAD 3.6.9.12 81:15 A AT 416-33

VOLTAGE (V)

 

 

3A
6A
9A
12A
15A

 

 

 

TITIrrTIrrIrIIIIIIIIIIIIIIIIIIIr1IIIIIIrrIITIIIIIIIII
.— in O m o to 0 1.0 O In 0 ION
v- v- N N n n <- V’ In 1010

NODE # (1 -57) 9405290

Figul'e 23. Profile of the neutral-to-earth voltage along the primary distribution

line from substation to 57th node with variable load demand
conditions at node 416-33 (single-phase DC model simulation)

87

reached 9.3 V (increased by 1.07 times from the base case of 4.5 V).
The neutral-to-earth voltage increased linearly with the load demand
current at a rate of 0.4 V/A at the high load demand node 33.

In the direction toward the end of the line, node 57, the
neutral-to-earth voltage increase was less and less significant. Even
when the load demand was 15 A, the voltage at nOde 57 only increased
by 0.7 V, to 5.6 V, only increased by 14.3% of the base case value of 4.9
V. '

Toward the direction of the substation, the neutral-to-earth
voltage increases first diminished, approaching the values close to those
of base case near node 16 (1.7 V). From node 16 to the substation node
1, as the distance increased, the voltages also increased slightly. Near
the substation, when the load demand was 15 A, the magnitude of the
voltage increased to 13.0 V, 1.41 times of the base case value of 9.2 V.

Figure 24 shows the neutral-to-earth voltage simulation curves
resulting from the different levels of the heavy load demand occurred at
the transformer branch between node 428 and 57, at the end of the
circuit model.

Compared with the base case, when load demand current was
increased from 3 A (base case) respectively to 6 A, 9 A, 12 A and 15 A,
the neutral-to-earth voltage also increased accordingly along the
distribution line. Note the most significant increase of the
neutral-to-earth voltage occurred at node 57 where the high load
demand took place. This voltage increase was similar to that of the high

load demand at node 33 except that the significant voltage increase

HEAVY LOAD IMPACT ON NEV, 4.8K 1-PHASE LINE

VOLTAGE (V)

LOAD 3, 6, 9, 12 8c 15 A AT 428-57

 

 

 

 

 

15 —+— 3 A
6 A
10 _ 9 A
12A
15A
5 .4
O - ---------- ”5?; -------------------------------------
4.31:0}1/
5f}?
'1"!
431',"
-5 D 933",
.1?"
:12”
{gal
“1014';
l
V
1
-15 ITTTTTIIIIIITIIIIIIIIIIIITTTTIIFIWI[ITIIIIIIIIITIIIIII
—~0 2 2 a a .9, a 9 “.2 a :21;
NODE # (1 -S7) 9405290
Figure 24. Profile of the neutral-to-earth voltage along the primary distribution

line from substation to 57th node with variable load demand
conditions at node 428-57 (single-phase DC model simulation)

89

occurred at the different location. At node 57, in the base case, when
the load demand current was 3 A, the value of the neutral-to-earth
voltage was 4.9 V. When the load demand was increased to 6 A, the
voltage increased by 1.9 V, to 6.8 V. When the load demand was
increased to 9 A, the voltage increased by another 1.9 V, to 8.7 V.
When the load demand was increased to 12 A, the voltage again
increased one more 1.9 V, to 10.6 V. When the load demand was
increased to 15 A, the voltage reached 12.5 V (increased by 1.55 times
from base case 4.9 V). The neutral-to-earth voltage increased linearly
with the load demand current at rate 0.63 V/A at the high load demand
node 57.

Toward the direction of the substation, the neutral-to-earth
voltage increases first diminished, approaching the values close to those
of the base case near node 25 (3.7 V). From node 25 to the substation
node 1, as the distance increased, the voltages also increased slightly.
Near the substation, when the load demand was 15 A, the magnitude of
the voltage increased to 13.1 V, 1.42 times of the base case value 9.2 V.

From Figure 22 through Figure 24, it can be seen that when a
primary high load demand occurs somewhere in the distribution line,
the neutral-to—earth voltages increased in most segments along the
distribution line. These voltage changes are load demand magnitude
and load demand location dependent. The voltages increased as the "
load demand currents increased. The most significant voltage increase
occurs near the high load demand location. This is due to the large

neutral current injection from high local load demand current. The

90

same conditions would be created when the phase line to neutral fault
occurred at a specific location along the distribution line.

The effect of high load demand on neutral-to-earth voltage is
equivalent to having a phase line to neutral fault occurred near the high
demand location. The faults in equipment on a primary distribution
line have been reported to have resulted in a lowering of the
neutral-to-earth voltage in a local area when the fault situation was
corrected. Previous research did not explain how this phenomenon
could occur.

The profiles obtained from the single phase AC model
corresponding to Figure 22 through 24 are listed as Figure A4 through
A6 in the appendix.

5.1.3 The Distribution System Ground Resistance Calculation

Table 5 presents the calculation results of the distribution system
ground resistance from four different locations along the power
distribution line model of Figure 4.

In Table 5, the first row discriminates the different node locations
and the first column discriminates the different electric parameters. In
the first column, the meaning of Thev_R5‘ and TheleO" is the
Thevenin equivalent system ground resistances calculated from 5 V and
10 V test voltages, respectively.

As described in Section 4.2.3, the grounding electrode resistance

at the test location was removed and replaced with a test voltage of 5

91

mm.._ .00. Nmm. 31.. "*O~m>0£.—.

00... .00. 000. .v... .xmm >om»
000.0 000.0. 000.0. «00.0 .0.auovu.
000.0. 000.0. 000.0. 000.0. "0.0.0010
.0..v v.0.0 .00.0 ~00.v .0 «nevi.
000.0 000.0 000.0 000.0 .0 0.0010

000.0 0n..00 . coao 0

.0000. 00: 000000.000
0Q 0.0000300. 00. m00_0 0000000. 03.808030. 00.0.00 .00.. 00.
0.0.. 0000.50. 0000.» 0.0.0.? 000000.000 00. .0 3.0.0. 0000—00—00 .w 030,—.

92

volts and 10 volts. The 4,800 volt substation supply voltage was replaced
with a short circuit. Figure 25 shows the test circuit to determine the
system resistance from the farm at node 33. The same technique can be
used to determine the system resistance from any location along the
distribution line.

At node 1 (substation), for the test neutral-to-earth voltages 5 and
10 V, respectively, the simulation program output the current 4.382 A
and 8.764 A, respectively. Using Ohm’s Law, R = V/I, both levels of
the test voltage gave the same results, i.e., the system resistance at node
1 (substation) was 1.141 ohm.

At node 9, the first farm transformer location, for the test
neutral-to-earth voltages 5 and 10 V, respectively, the simulation
program output the current 8.891 A and 17.780 A, respectively. Using
Ohm’s Law, R = V / I, both levels of the test voltage gave the same
results, i.e., the system resistance at node 9 was 0.562 ohm.

At node 33, near the middle of the distribution line, for the test
neutral-to-earth voltages 5 and 10 V, respectively, the simulation
program output the current 8.314 A and 16.630 A, respectively. Using
Ohm’s Law, R = V/I, both levels of the test voltage gave the same
results, i.e., the system resistance at node 33 was 0.601 ohm.

At node 57, at the end of the distribution line, for the test
neutral-to-earth voltages 5 and 10 V, respectively, the simulation
program output the current 4.191 A and 8.382 A, respectively. Using
Ohm’s Law, R = V/I, both levels of the test voltage gave the same

results, i.e., the system resistance at node 57 was 1.193 ohm.

93

mm 0002

.mm 0000
.0 0:0. 0... 0.00 0000.50. 0.0.0». 00. 00.8.0.0“. 0. 0000.0 .00. 00,—. ..nN 0.09m

 

mm 0002 m 0002 _ 0002

94

From the above description, at each node, the two levels of the
test voltage, 5 V and 10 V, gave the exact the same result of the

distribution system resistance at that node. This has been expected for

the linear circuit model.

The distribution system ground resistance is an important
indicator for some of the abnormalities in the electrical parameters
producing neutral-to-earth voltage along the power distribution system.
This is illustrated by Figure 26 and Table 6.

Figure 26 and Table 6 present the changes of the distribution
system ground resistance (Thevenin Equivalent Resistance) and
Thevenin Equivalent Open Circuit Voltages at the location of node 33
to earth as well as neutral-to-earth voltage at node 33 with the
resistances increased from 0.086 ohm to 100 ohms in the neutral line
segments RD4, RD8, RD12, RD16, RD20 RD56, respectively as
shown in the circuit model connection of Figure 27. From Figure 26
and Table 6, it can be seen that generally speaking, the primary ground
system resistance at the location node 33 to earth was higher with a bad
neutral connection on any segment of the neutral conductor than
without the bad neutral connection compared to the base case system
(all RDs were 0.086 ohm). The increase of this ground system
resistance to earth was very significant when the bad neutral connection
occurred near the neighborhood of node 33 where the ground systems
resistance was observed. In this location, the ground system resistance
for the normal base case was 0.601 ohm. With a bad neutral connection
of 100 ohms at RD28, near the left neighbor transformer of the

95

PD4, PDB PD56 : IOO OHM, PESPECTIVELY
PPIMAPY SYSTEM GROUND RESISTANCE, VOLTACES, NODE 33—O

 

 

 

 

 

 

 

IO [ -
i —+—- Voc33 :
—0— NEVIS} ;
8 1 —)(—- Lsys '
m —'?- RF033 '
1.... .
__J ‘ W
O 6— .
> I
m 49%;]
E
I
O
2 -1
O I I I I I I I I I I I I I I :IT
10 o o o o o o o o o o o o o 0
(<12 0 o o o o o o o o o o o o o
O II II II II II II II n , II II II II II II
m m to o 0- m N (a o sr co 01 co
< <2- 00 .— .— m N (\l n r) v 0- v to n
m D O Q a o o o o c: o o o c: o
0: Ct at o: o: c: o: a: o: 0: cr 0: o: m

Figure 26. The simulation plot of changes of the distribution system ground
resistance (Thevenin Equivalent Resistance) and Thevenin Equivalent
Open Circuit Voltages at the location of node 33 to earth as well as
neutral-to-earth voltage at node 33 with the resistances increased from
0.086 ohm to 100 ohms in the neutral line segments RD4, RD8, RD12,
RD16, RD20 RD56, respectively in the DC circuit model.

0.9.08.2 .080 .000 .050 0.00 .000 .000 088.9.
00.. .0..000 0... 0. 000.0 8. 0. 00.0 0.8.0 0.0.0 00.00.00. «0000.000.
0... 003 mm 0000 .0 0m0..0> 0..00.0.-.0..000 00 ..03 ...0 0.0.0 0.
mm 0000 .0 000000. 0... .0 m0M0..0> 000.5 0000 00.0203. 0.0050...
000 3000.003. .00.0>.0..m. 0.0050,... 0000.30. 0000.» 80.0...
000000.00 0... .0 00m0000 0... 00.00.00 0. 000000000 00000 00... SN 0.00....

 

   
   
  
 
     

     

     

     

    

     

   

  

0mg. ~20. 0:... 3.... 0:... 03.. ~83 swam 02... 8:3 0:... ~33 2.8 2...

mm 0002
0:0..000. 00.00.00.
.0 00000.00. 00: 00:0.m_00.

.000000 00.. 00 :0_.000. 0.0.
.0 0.0.0.0.00 00000000. 00..

97

Table 6. The simulation data of changes of the distribution system ground
resistance (Thevenin Equivalent Resistance) and Thevenin Equivalent
Open Circuit Voltages at the location of node 33 to earth as well as
neutral-to-earth voltage at node 33 with the resistances increased from
0.086 ohm to 100 ohms in the neutral line segments RD4, RD8, RD12,

RD16, RD20 RD56, respectively in the DC circuit model.

13le SYSTEH BRUNO RESISTME, m 33-0, R0__ : 1m [I'll 94-9-22

 

RD8:0Hfl lV_opon Vchloso Vings33_0§RF833 0H:*NEV_CflLthost V: 1_tost R:

 

885 8958 6.283 4.486 .6013 1.50:!) 4.4850 10 16.63
R04 :100 7.3052 5.2041 .6057 1.5000 5.2039 10 16.51
R08 :100 7.2454 5.1586 .6068 1.5013 5.1586 10 16.48
120123121] 7.4034 5.2378 .6192 1.511!) 5.2381 10 16.15
120167.111) 7. 2792 5.1391 .6246 1.5111] 5.1392 10 16.01
9020:1113 7.7194 5.3026 .6835 1.5% 5.3029 10 14.63
120242111] 7.5011 5.0922 .7097 1.51!!! 5.0919 10 14.09
1:028:11!) 9.1786 5.437 1.0323 1.5000 5.4369 10 9.687
$203221“) 8.976 4.9887 1.1989 1.5000 4.%87 10 8.341
R03621m 6. 7569 3.8607 1.1252 1.5000 3.8607 10 8.887
RD40=1CIJ 7.2473 4.3819 .9814 1.5!!) 4.3811 10 10.19
R0443!” 6.2766 4.2%9 .6940 1.5111] 4.2913 10 14.41
R0483“) 6.5343 4.5151 .6707 1.5% 4.5154 10 14.91
9052000 6.2452 4.4245 .6173 1.5011 4.4244 10 16.2
12056211!) 6.3721 4.5248 .6124 1.5000 4.5248 10 16.33

 

mm svsren mm Resrsrmce, vumsss, none 33-0
RD4, ma, R012

arms a was

Ro_u (940922) R0933; : [tut/Lust '

vapon W50. : Vopon'I-RFS33/(RFB334'R39533_O)
$933

a

-‘9‘
5633

R056 : too am, RESPECTIVELY

98

observation, the ground system resistance was 1.032 ohms, 71.7% higher
than the normal. With a bad neutral connection of 100 ohms at RD32,
near the transformer of the observation, the ground system resistance
was 1.199 ohms, 99.5% higher than the normal. With a bad neutral
connection of 100 ohms at RD36, near the right neighbor transformer
of the observation, the ground system resistance was 1.125 ohms, 87.2%
higher than the normal. With a bad neutral connection of 100 ohms at
RD40, near the second right neighbor transformer of the observation,
the ground system resistance was 0.981 ohm, 63.2% higher than the
normal. This implies that a bad neutral connection along the line can
be detected by the measurement of the local primary ground system
resistance.

The distribution system ground resistance is very important
information for the evaluation of neutral-to-earth voltage changes due
to ground resistance changes at a particular location. This can be seen

in next sections.

5.1.4 Substation Grounding Resistance Change

Figure 28 presents the neutral-to-earth voltage simulation curves
resulting from changing the substation grounding resistance. Such a
change can occur due to corrosion, nearby earthwork construction,
water-table fluctuation, soil surface layer freezing or other natural
season effect. Low substation resistance to earth may be difficult to

achieve due to soil conditions.

CHANGE IMPACT ON NEV, 4.8KV LINE
SUBSTATION R61 0.5, 1, 5, 10, 20 8c 40 OHM, DC CASE

 

 

 

 

10
(55?? ‘ 2 ‘ E '57“ -
O-T ........ 555‘. 'h .............................
55‘ 1'"
V _O'
A ,1 .0"
> n ‘0'"
v '1 4'
Lu n v; —+-—
$2 -103 n 2"; 0.5 OHM
,_ h' I“ —a- 1.0 OHM
.J '0"
g E .i’ + 5.0 OHM
/; + 10 OHM
-20fi y.” + 20 OHM
_7' + 40 OHM
1;
,m.
//
-30 fiTTITIIIIT—TrTllllIIIITITIIIIIWIIITTITIIIIITIWIIIlllllr
— n o m o n o n o n o m rx
'- v- N N 7’) F) V’ 1' n to W)

NODE # (1-57) 940531 b

Figure 28. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. variable substation
resistance RS (single-phase DC model simulation)

100

Compared with base case, when substation resistance was
increased from 0.5 ohm (base case) respectively to 1.0, 5 .0, 10, 20 and 40
ohms, the neutral-to-earth voltages increased accordingly in some
segments near the substation and decreased some distance away from
the substation toward the end of the line. Note the most significant
changes of the neutral-to-earth voltage occurred near node 1 where the
substation was located. At the substation, in the base case, when the
substation resistance was 05 ohm, the value of neutral-to-earth voltage
was 9.2 V. When the substation resistance was increased to 1.0 ohm,
the voltage also increased by 4.9 V, to 14.1 V. When the substation
resistance was increased to 5 .0 ohms, the voltage increased by another
105 V, to 24.6 V. When the substation resistance was increased to 10
ohms, the voltage increased another 2.5 V, to 27.1 V. When the
substation resistance was increased to 20 ohms, the voltage was only
increased by 1.4 V, to 28.5 V. When the substation resistance was
increased to 40 ohms, the voltage was only increased another 0.8 V, to
29.3 V. Generally, the voltage change was increased near the substation
as the substation resistance was increased; the effect was significant in
the resistance range from 0.5 to 5 ohms.

This effect on the neutral-to-earth voltage diminished rapidly as
the distance increased from the substation (node 1) toward the end
(node 57) of the line. With 40 ohms substation resistance, at the
substation node 1, the neutral-to-earth voltage was 29.3 V, 3.18 times of
the base case value of 9.2 V; at the middle node 33, the voltage was 35

101

V, only 77.8% of the base case value of 45 V; at the end node 57, the
voltage was 4.7 V, 96% of the base case value of 4.9 V.

From Figure 28, it can be seen that when the substation grounding
resistance is increased, the neutral-to-earth voltages increased in some
segments near the substation and decreased some distance away from
the substation toward the end of the line. These voltage changes are
related to the distance from and the resistance-value of the substation.
The neutral-to-earth voltages are increased significantly near the
substation and changed less and less significantly as the distance from
the substation increases. Similar results were reported by Kehrle (1984)
and Gustafson (1985) in their simulation.

Table 7 lists some of the numerical calculation results in this case.
Looking inside from node 1 and earth (referring to Figure 4, Chapter 4),
the Thevenin equivalent circuit (when RS removed) resistance
(Thev_R) was 1.141 ohm and open circuit voltage (V_open) was 30.176
V. The second row lists the calculated values of the substation
neutral-to-earth voltages corresponding to different levels of the
substation resistances. For example, when substation resistance RS was
1 ohm, the corresponding substation neutral-to-earth voltage can be
calculated as

NEV_Cal‘ = V_open x RS/(ThevR + R8)
= 30.176 x 1000/(1.141 + 1.000)
= 14.094 (V)

102

8m + EVE-» u 26-»!

«8.8 26.8 98.8. ownéu 8.: 31¢ .28-»!
8.9 8.8 8.2 806 8°; 8. 3.... £18 _ «3-8
:8 8-9.8 8.9.8 2.2.3... ataxia 185.832: «.92: 58-2 5.33.
canon... 39 n5 8:325. 3.3.6 «3.8 83338

.358 cc... 8:22:86 OD 823-2»:7. on. we ems—g 38367353:
new cog—8&8: 2:55 .omng 55992830: :383 some
62.538 258m 829?. 8:933. mo $.82 cone—=28 38.8952 .5 035.

103

From this, it can be seen that the substation Thevenin equivalent open
circuit voltage was divided in series by the substation ground resistance
and the system Thevenin equivalent resistance at this location as shown
in Figure 29. The level of neutral-to-earth voltage is determined by the
open voltage and the proportional relation between the substation
ground resistance value and the system Thevenin equivalent resistance
value at this location. A lowered substation grounding mat resistance
results in a lowered substation neutral-to-earth voltage.

The profile obtained from the single-phase AC model
corresponding to Figure 28 is listed as Figure A7 in the appendix.

5.1.5 Neutral-to-Earth Resistance Reduction

Figure 30 presents the neutral-to—earth voltage simulation curves
resulting from reducing neutral-to-earth grounding electrode resistance
RFG33 at the middle of the distribution neutral line.

Compared with the base case, when the grounding electrode
resistance RFGB3 was reduced from 1.5 ohm to 1.2, 0.9, 0.6 and then to
0.3 ohm at node 33, at node 33 and nearby along the distribution line,
the neutral-to-earth voltage decreased as the grounding resistance was
reduced. However, this reduction of neutral-to—earth voltage was
localized. The most significant neutral-to-earth voltage reduction
occurred at node 33 itself. At node 33, in the base case, when RFG33
was 1.5 ohms, the value of neutral-to-earth voltage was 4.5 V. When the
RFG33 was decreased to 1.2 ohms, the voltage also decreased by 0.3 V,

104

NUDE l NUDE l

l l
if e [7 e

RS Thev-R RS Thev-R
.S ohn J. .5 ohn

Figure 29. The substation Thevenin equivalent open circuit voltage is divided in
series by the substation { ground resistance and the system Thevenin
equivalent resistance at this location.

105

8033 CHANCE IMPACT ON NEV, 4.8KV LINE
GROUND R033 1.5, 1.2, 0.9, 0.6 8: 0.3 OHM, DC CASE

 

 

 

 

 

A
>
V
Lu
0
<
..—
5‘
+0.90HM
-6‘ +0.60HM
+0.3OHM
-8.
“'10IIIIIIITIITIIIIIlllIlIllllFIIITIIITTlllllllITIIITIIITI
.-— Ln O In C in 0 t0 0
.- ._ a. N n n 4 9 8 :25

NODE # (1-57) 940530-1

Figure 30. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. ground resistance
RFG33 change at node 33 (single-phase DC model simulation)

106

to 4.2 V. When RFG33 was decreased to 0.9 ohm, the voltage
decreased by another 0.4 V, to 3.8 V. When RFG33 was decreased to
0.6 ohm, the voltage decreased still by another 0.7 V, to 3.1 V. When
RFG33 was decreased to 0.3 ohm, the voltage decreased again by
another 1.0 V, to 2.1 V. The neutral-to-earth voltage decreased more
drastically at this node when RFG33 was reduced from 0.9 ohm to 0.3
ohm than from 1.5 ohms to 0.9 ohm.

This voltage reduction effect was less and less significant as the
distance increased from node 33 toward both the substation and the end
of the distribution line. When RFG33 was 0.3 ohm, at node 33, the
voltage was 2.1 V, 46.7% of the base case value of 4.5 V; at the end
node 57, the voltage was 4.5 V, 91.8% of the base case value of 4.9 V;
however, at the substation node 1, a very small neutral-to—earth voltage
change was experienced with the voltage at 9.3 V, slightly changing in
the contrary directions to those of nodes 33 and 57, 101.1% of the base
case value 9.2 V.

From Figure 30, it can be seen that when the neutral-to-earth
resistance is decreased somewhere in the middle location of the
distribution line, the voltage reduction effect is localized. The most
significant voltage reduction occurs at the location where the grounding
resistance is reduced. The voltage reduction is less and less significant
as the distance increases from that location in either direction. Similar
results were also reported by Kehrle (1984) and Surbrook et al., (1988)

in their simulation.

107

Table 8 lists some of the numerical calculation results in this case.
Looking inside from node 33 and earth (referring to Figure 4, Chapter
4), the Thevenin equivalent circuit (when RFG33 removed) resistance
(Thev_R) was 0.601 ohm and open circuit voltage (V_open) was 6.283
V. The second row lists the calculated values of the neutral-to-earth
voltages corresponding to different levels of the ground resistances at
this location. For example, when ground resistance RFG33 was 0.9

ohm, the corresponding neutral-to-earth voltage can be calculated as

NEV_Cal" = V_open x RFG33/(ThevR + RFG33)
= 6.283 x 0.9/(0.601 + 0.9)
= 3.767 (V)

From this, it can be seen that actually at any ground connection location
of the distribution line, the system ground Thevenin equivalent open
circuit voltage was divided in series by the ground resistance and the
Thevenin equivalent resistance (system ground resistance) at this
location. Generally speaking, a lowered ground resistance at a
particular location along the line will results in a lowered local
neutral-to-earth voltage.

Prothero et al. (1988) also reported from their field measurement
data that their tests showed a net resistance of less than 0.5 ohm for the
primary neutral network. The measured resistance of farm ground
systems ranged from 2.0 to 5 .0 ohms per farm. The connection of these

low resistance farm grounds to the primary neutral network was found

108

Snob. + Eekmohi-a .. .84.!

 

~8.~ 8....” Sad 3.... 84... Eco->9.

 

8m. 8. 8o. 8.... 8a.. .8. 8...... ...3-8

 

“no mmoufiv. mmouafi. mmoH—SN. nab—:0 finch!— ¢t>OF= coco-.3. 08—308

 

K-o-u a! ..5 ouioui... 9.3.6 80.8 .89....

.002: 0:: coasts... OD omega-0&5... on. we ESQ-BE .a owes?
580-03830: new 35.3.8. 958» dug—9. stun-9-1.330:
.583 some .3536... 959% 88»? ..o 3.82 norm—=28 ...Otofizz .m nigh.

109

to have a larger influence over primary neutral-to-earth voltage levels
than the additional ground electrodes or counterpoise. This is
consistent with the theoretical analysis in this research.

Figure 31 presents the neutral-to-earth voltage simulation curves
resulting from improving neutral-to-earth grounding by adding an extra
grounding electrode of 25 ohms at every node on the neutral conductor
where the resistance was previously set at 100 k-ohms. See Figure 4 and
Table 2. For this simulation, there is a grounding electrode of not more
than 25 ohms at each pole along the distribution line.

Compared with the base case, when the grounding electrode
number was doubled, the neutral-to-earth voltage was lowered at most
nodes (from node 13 to node 57) along the distribution line from the
middle to the end. This voltage reduction effect was significant from
the end to the middle location of the distribution line. At the end (node
57), the value of neutral-to-earth voltage decreased by 0.6 V, to 4.3 V,
87.8% of the base case value 4.9 V; at the middle node 33, voltage was
3.9 V, 86.7% of the base case value 45 V; at the substation node 1, very
small neutral-to-earth voltage decrease (0.1 V) was experienced with the
voltage 9.1 V, compared with 9.2 V of the base case.

From Figure 31, it can be seen that when the grounding electrode
number was doubled uniformly along the neutral line, the
neutral-to-earth voltage was lowered a small amount from the middle to 5
the end of the line. This voltage change was insignificant near the

substation.

110

INCREASE GROUND ROD NUMBER IMPACT ON NEV
ALL GROUND RESISTANCES 100K CHANGED TO 25 OHMS

 

VOLTAGE (V)

 

 

 

 

'10‘rrlllIIIIIIIrITrTrTrIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
.— In 0 In 0 in O in O in O ION
.— .— N N r’) n 1' 1' in min

NODE # (1 -57) 940530-15

Figure 31. Profile of the neutral-to-earth voltage along the primary distribution
line fi'om substation to 57th node with base case vs. ground rod

number increasing along the entire system (single-phase DC model
simulation)

111

Reduction of the grounding electrode resistance at node 33 in the
middle of the line decreased the local neutral-to-earth voltage.
Addition of extra grounding electrodes was equivalent to increasing the
parallel resistance number and decreasing the actual overall resultant
ground resistance.

Prothero et al. (1988) also reported from their field measurement
that the addition of numerous supplemental grounding electrodes, to
the extent that every distribution pole was grounded, had only a slight
effect on reducing primary neutral-to-earth voltage.

Although increasing the number of grounding electrodes can
decrease the neutral-to-earth voltage to some extent, a large number of
ground rods will be costly and the effect may not be as good as
expected. For this simulation, doubling the number of grounding
electrodes along the primary line only resulted in approximately a 12.2
percent reduction in neutral-to-earth voltage at the end of the line.

The profiles obtained from the single-phase AC model
corresponding to Figure 30 and 31 are listed as Figure A8 and A9 in the
appendix.

5.1.6 Primary Operating Voltage Level Change

Figure 32 presents the neutral-to-earth voltage simulation curves
resulted from changing the primary operating voltage levels. The base
case simulation assumes a primary ungrounded conductor operating at
4.8 kV.

112

SUBSTATION SOURCE VOLTAGE CHANCE IMPACT
v SOURCE VOLTAGE 2.4, 4.8, 7.2 s- 26.4 KV

 

 

 

 

 

 

10
5“
01 ______ -. ..::_
A ,-" .
> 1 a
v
-5.
< I
.—
_‘ +2.4KV
O
>-10‘ +4.8KV
+7.2KV
+26.4KV
-15.
-20 llllImIIlTTrllIIITIIIIIIFTIITIIIIITIIIIIITTIITIIIITT
v- In 0 In O tn 0 In 0 n
v— v- N N '0 n 1' V 8 iii-'3

NODE # (1 -57) 940531 c

Figure 32. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case 4.8 kV operating
voltage vs. 2.4, 7.2 and 26.4 kV operating voltages (single-phase DC
model simulation)

113

When the primary operating voltage levels were increased from
2.4 kV respectively to 4.8 kV, 7.2 kV and 26.4 kV with power
unchanged, the neutral-to-earth voltages in all these four distribution
systems decreased correspondingly along the distribution line. These
voltage profiles had the similar shapes with the different proportion
scales. All four voltage profiles approached convergence near node 12.

The neutral-to-earth voltages of all four systems decreased as the
distance away from the substation increased, reaching zero and
changing 180 degree in phase angle near node 12. Then the voltages
increased as the distance toward the end of the line increased. The
neutral-to-earth voltage profiles changed sharply at first, then leveled
off toward the end of the neutral line. At node 12, the neutral-to-earth
voltages corresponding to the systems of 2.4 kV, 4.8 kV, 7.2 kV and 26.4
kV were 0.1 V, 0 V, 0 V and 0 V, respectively. At the substation node 1,
the neutral-to-earth voltages corresponding to the systems of 2.4 kV, 4,8
kV, 7.2 kV and 26.4 kV were 17.6 V, 9.2 V, 6.2 V and 1.7 V, respectively.
At the end (node 57), the neutral-to-earth voltages corresponding to the
systems of 2.4 kV, 4.8 kV, 7.2 kV and 26.4 kV were 9.3 V, 4.9 V, 3.3 V
and 0.9 V, respectively.

Generally, the higher the operating voltage, the lower the
neutral-to-earth voltage. This is due to the constant power situation:
the higher the operating voltage, the lower the current on the primary
side of the transformer. Lower transformer primary current leads to

lower current injection to the neutral line and parallel ground path.

114

This neutral-to-earth voltage change effect due to the primary
operating voltage change was much more significant in the operating
voltage level range from 2.4 -- 4.8 kV than the range from 7.2 -- 26.4 kV.
The neutral-to-earth voltage decreased 20% per kV on the average in
the operating voltage level range from 2.4 -- 4.8 kV, 13.6% per kV on
the average in the range 4.8 -- 7.2 kV, and only 3.8% per kV on the
average in the range 7.2 -- 26.4 kV. This implies that striving for high
operating voltage levels may not reduce the neutral-to-earth voltage as

much as expected.
5.1.7 Secondary Earth Fault

Figures 33 through 38 present the neutral-to-earth voltage
simulation curves resulting from normal line loading plus a ground fault
on the secondary side of one of the transformers. The simulation
network used in this case is shown in Figure 7.

When the out-of-phase secondary ground faults of 3, 6, 9 and 12
amperes were introduced at the transformer between nodes 416 and 33,
at the middle of the distribution line, from Figure 33, compared with the
base case, some changes in the neutral-to-earth voltage were observed.
The most significant voltage changes occurred near node 33 where the
faulted secondary network was attached. At node 33, the voltage
increased significantly as the fault current increased. In the base case,
when the fault current was zero, the value of neutral-to-earth voltage

was 4.5 V. When the fault current was increased to 3 A, the voltage

115

Secondary Ground Fault, Out-of-Phase

IMPACT ON NEV, FAULT 3, 6, 9 8c 12 AT 416

 

VOLTAGE (V)

 

 

 

 

-1OTT—TITTIIITTIITTTITITTTTTITTITTITTITTIII-TIjTTTT'IT'TTTr
-n o In C n o 10 0 ton
'— ‘n 53 .— N N n n 1' wt in mm

NODE # (1 --57) 940504—4

Figure 33. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. variable secondary

ground fault attached to node 33 (out-of-phase, single-phase DC
model simulation)

116

increased by 1.2 V. to 5.7 V. When the fault current was increased to 6
A, the voltage increased by another 1.2 V, to 6.9 V. When the fault
current increased to 9 A, the voltage increased an additional 1.1 V, to
8.0 V. When the fault current increased to 12 A, the voltage increased
still an additional 1.2 V, to 9.2 V. The neutral-to-earth voltage
increased linearly with the fault current at a rate about 0.4 V/A at the
faulted location node 33.

This voltage change effect was less and less significant as the
distance increased from node 33 toward both the substation and the end
along the distribution line. When the fault was 12 A, at node 33, the
voltage was 9.2 V, 2.04 times of the base case value 4.5 V; at the end
node 57, the voltage was 5.7 V, 116% of the base case value 4.9 V;
however, at the substation node 1, very small neutral-to-earth voltage
change was experienced with the voltage 9.0 V, slightly changing in the
contrary directions, 97.8% of the base case value 9.2 V.

When the in-phase secondary ground faults of 3, 6, 9 and 12
amperes were introduced at the transformer between node 416 -- 33, at
the middle of the distribution line, from Figure 34, compared with the
base case, it can be seen that neutral-to-earth voltage changed in a
similar way to the out-of-phase secondary ground fault situation except
the voltage changes were in the contrary direction. Note in Figure 34
that a 12 ampere secondary in-phase ground fault is subtractive with the
primary load which produced base neutral-to-earth voltage resulting in

a net neutral-to-earth voltage value near zero at node 33.

117

Secondary Ground Fault, In-Phase

IMPACT ON NEV, FAULT 3, 6, 9 8c 12 AT 416
10

 

 

 

 

5-1
A
>
V
8
0d ---------
<
'5'
-B-—SA
-54 +6A
+9A
+12A
...10 I
v- 10 0 1n 0 tn 0 In C In O ION
'- v- N N I") n V Q' In IOU-D

NODE # (1 -57) 940604-3

Figure 34. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. variable secondary

ground fault attached to node 33 (in-phase, single-phase DC model
simulation)

118

When the out-of-phase secondary ground faults of 3, 6, 9 and 12
amperes were introduced at the transformer between node 404 -- 9, the
first farm transformer location along the distribution line, from Figure
35, compared with the base case, it also can be seen that there were
changes in the neutral-to-earth voltage. The significant voltage changes
occurred at the node 12 near the faulted location node 9. At node 12,
the voltage was increased significantly as the fault current increased. In
the base case, when the fault current was zero, the value of
neutral-to-earth voltage was 0 V. When the fault current was increased
to 3 A, the voltage also increased by 0.9 V, to 0.9 V. When the fault
current was increased to 6 A, the voltage increased by another 1.0 V, to
1.9 V. When the fault current was increased to 9 A, the voltage
increased an additional 0.9 V, to 2.8 V. When the fault current was
increased to 12 A, the voltage increased still an additional 0.9 V, to 3.7
V. The neutral-to-earth voltage increased linearly with the fault current
at a rate about 0.3 V/ A near the faulted location.

This voltage change effect was less significant as the distance
increased from node 12 in either direction of the distribution line.
When the fault was 12 A, at node 12, the voltage was 3.7 V, much higher
than the base case value 0 V; at the end node 57, the voltage was 5 .0 V,
102% of the base case value 4.9 V; however, at the substation node 1,
the voltage was 7.3 V, changing in the contrary directions to those of
nodes 33 and 57, 79.3% of the base case value 9.2 V.

When the in-phase secondary ground faults of 3, 6, 9 and 12

amperes were introduced at the transformer between node 404 -- 9, the

119

Secondary Ground Fault, Out-of-Phase

IMPACT ON NEV, FAULT .3, 6, 9 8c 12 AT 404

10

 

-+—BASE
+3A

VOLTAGE (V)

-x—-BA

 

+9A

 

+12A

 

 

-10 IIITTTTVTTIVTTTIIIITTTTTTIIIIIIIITTTIIIIIIIIIIITITTIIIT

'- 10 C 1n 0 IO 0 1n 0 1n 0 ION
v— v- N N I") F) 1’ V’ In 1010

NODE # (1 -57) 940604-2

Figure 35. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. variable secondary

ground fault attached to node 9 (out-of-phase, single-phase DC model
simulation)

120

the first farm transformer location of the distribution line, from Figure
36, compared with the base case, it can be seen that the neutral-to-earth
voltage changed in a similar way to the out-of-phase ground fault
situation except the voltage changes went to the contrary direction.

With an out-of-phase secondary ground fault at node 33 in the
middle of the line, most customers experienced an increase in
neutral-to-earth voltage as shown in Figure 33. When there was an
in-phase secondary ground fault at node 33 in the middle of the line,
most customers experienced a decrease in the neutral-to-earth voltage
(Figure 34). But this was not true when the secondary ground fault was
at a location near the substation. Examination of Figures 35 and 36
shows that some customers experience a lowering of the
neutral-to-earth voltage while others experience an increase whether
the secondary ground fault was in-phase or out-of-phase with the
primary load which produced the base neutral-to-earth voltage.

When the out-of-phase secondary ground faults of 3, 6, 9 and 12
amperes were introduced at the transformer between node 428 -- 57, the
last farm transformer location of the distribution line, from Figure 37,
compared with the base case, it still can be seen that there was some
changes in the neutral-to-earth voltage. The most significant voltage
changes occurred near the end node 57 where the faulted secondary
network was attached. At node 57, the voltage was increased
significantly as the fault current increased. In the base case, when the
fault current was zero, the value of the neutral-to-earth voltage was 4.9

V. When the fault current was increased to 3 A, the voltage also

10

VOLTAGE (v)

121

Secondary Ground Fault, In-Phase

IMPACT ON NEV, FAULT 3, 6, 9 8c 12 AT 404

 

 

 

 

 

-15IIITI[ITIIIIITIITTTITIFTIIIITIIIITIITIIITIIIIlllllrill
v- 10 O In 0 in O n o 10 0
v- v- N N n n v 1' In 31'?-

Figure 36.

NODE # (1 -57) 940504-1

Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. variable secondary

ground fault attached to node 9 (in-phase, single-phase DC model
simulation) -

122

Secondary Ground Fault, Out-of-Phase

IMPACT ON NEV, FAULT 3, 6, 9 8c 12 AT 428

 

 

 

 

15
F
10«
' t
1"{':'- '—:::‘J
A l't'f,.-+:’-:::===‘-
\>_/ 5 q .. f- ~.‘ : ;;:r=-="; F
LIJ
0
<1
1.—
_l
O 011- -----------------------------------------------
>
+BASE
+12A
-10 IIIIIIIIIIIIIIIITTII1IIIIIIIIIIIIIIIIIIITIIIIITIITIIlllr
'- 1.0 O n O n o m o m o n"
v- v- N N n n 1’ 1' In IDIO

NODE # (1 -57) 940604-6

Figure 37. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. variable secondary
ground fault attached to node 57 (out-of-phase, single-phase DC
model simulation)

123

increased by 1.9 V, to 6.8 V. When the fault current was increased to 6
A, the voltage increased by another 1.7 V, to 8.5 V. When the fault
current was increased to 9 A, the voltage increased by an additional 1.8
V, to 10.3 V. When the fault current was increased to 12 A, the voltage
increased by still an additional 1.6 V, to 11.9 V. The neutral-to-earth
voltage increased linearly with the fault current at a rate about 0.53 --
0.63 V/A at the faulted location node 57.

This voltage change effect was less significant as the distance
increased from the end node 57 toward the substation node 1 along the
distribution line. When the fault was 12 A, at the end node 57, the
voltage was 11.9 V, 2.43 times of the base case value 4.9 V; at node 33,
the voltage was 5 .2 V, 116% of the base case value 4.5 V; however, from
the node 10 to substation node 1, the voltage had almost no change.

When the in-phase secondary ground faults of 3, 6, 9 and 12
amperes were introduced at the last farm transformer location node 428
-- 57, from Figure 38, compared with the base case, it can be seen that
the neutral-to-earth voltage changed in a similar way to the
out-of-phase ground fault situation except the voltage changes went to
the contrary direction. The net effect of the in-phase secondary ground
fault was to reduce the neutral-to-earth voltage.

From Figure 33 through Figure 38, it can be seen that for a
ground fault on the secondary side of a transformer, the
neutral-to—earth voltages along the distribution line were changed to
some extent. Generally, this kind of change is fault-size, fault-phase and

fault-location dependent. The heavier the fault currents, the more

124

Secondary Ground Fault, In-Phase

IMPACT ON NEV, FAULT 3, 6, 9 8c 12 AT 428

 

 

 

 

10
51 , ...... .. _..._ 1
.' OQ“<;:“ ‘-::::===‘
A --a
>
V
8 K
0.1 ------------------------------------------- - -
<(
1— 7
_.|
O +BASE
>
+315-
...5-1 +6A
+12A
-1OTTTTTITTTTTTIITTIIITIITITIIIIlllllll‘lllIIIIIIIITTITTIrl
'— ID 0 In 0 In C in O ID O on
v- P N N l’) n V V 10 1011')

NODE # (1 -57) 940604-5

Figure 38. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. variable secondary
ground fault attached to node 57 (in-phase, single-phase DC model
simulation) .

125

significantly the neutral-to-earth voltages change. The voltages change
in contrary directions when the in-phase or out-of-phase fault occurs.
Much more significant changes in neutral-to-earth voltage occur at and
near the node at which the faulted secondary circuit is attached. The
farther the distance along the line away from the fault node, the less
significant the neutral-to-earth voltages change. A similar result was
also reported by Kehrle (1984) and Dick et al. (1987) in their
simulation. And Bodman et al. (1981) reported from their field survey
and measurement process that in one installation, a short circuit within
the milk pump resulted in a constant 21.3 V voltage between the milk
pump and the milk house floor.

In this simulation, a significant neutral-to-earth voltage change
due to a secondary ground fault was confined to the distribution system
within a limited distance of approximately one-half mile of the fault
location. As the level of fault current increased, the neutral-to-earth
voltage change on the distribution line increased to approximately one
mile for a heavy ground fault.

The change in level of neutral-to-earth voltage due to a secondary
ground fault is greater when the fault occurred at the end of the line
than at the middle of the line as can be seen from Figures 35 through
38. Near the substation the amount of change was less, but whether a
particular location would experience a net increase or decrease in
neutral-to-earth voltage was difficult to predict. When a secondary fault
occurred at the end of the line as was the case at the middle of the line,

an out-of—phase secondary ground fault caused an increase in

126

neutral-to-earth voltage for all customers. For the in-phase secondary
ground fault, customers experienced a reduction of neutral-to-earth
voltage unless the amount of reduction was great enough to result in a
phase angle change in the neutral-to-earth voltage. In this latter
situation, it is possible for the net rms value of the voltage to actually
become higher.

It is important to note that a secondary ground fault produced
neutral-to-earth voltage can be additive or subtractive to the primary
line load produced neutral-to-earth voltage. The elimination of a
secondary ground fault at one location resulting in a lowering of
neutral-to-earth voltage at that location may indeed cause a significant
increase in the neutral-to-earth voltage at other locations. When a
secondary ground fault is discovered, it is recommended that voltage
measurements be taken along the distribution line in the area after the
elimination of the secondary ground fault to determine if any customers
have experienced an increase in neutral-to-earth voltage.

The out-of-phase and in-phase secondary ground fault condition
at the end of the distribution line is illustrated in Figures 39 and 40.
Figure 39 shows the end of the distribution line with the simulated
secondary electrical system attached. The out-of-phase fault condition
with respect to the primary line is defined by observing the phase
relationship of the phase angle of the fault current of the secondary
ground fault circuit with respect to the primary current. It is important
to note that the fault current and the normal line load current flowing

through the grounding electrodes is in-phase thus resulting in a net

0::
6:22.36 2: ..O can 2.. .a 5n one: .a vogue... .38 2:55 3.29.9.6
.3662... .53 30:3 2.8.00.0 9:05.03 0:... 5.26:8 .8502 an 9.33

 

page 3.. .II'
page p.53. l-I'.

 

127

 

 

..-

 

      

 

 

 

3
. -II A" 5
5;: m u. m
m as. mmam ¢mfi _ an: N m
f
2.... 52 Mm
‘III. ‘III
omv am?— omv muse

128

0:: 5:353.» 0.: go 0:0 05 .0 5. 0.3: .0 @0583 .38 9:5...
00.23;: b09500» 55 30:30 03:00.0 9.6.50.3 can 3.03:3 .8302 .3 0:85

 

 

 

 

mi m 39 an. 28.

RV II' ..l' .ll m w II D“ —m

TI

I Name .22 ‘III was. mmac 33. «max when .33
m m .0 >8. lulu“ In... I!
‘ 1111'—1—d1”1¢* m ‘ aunt ’
..l 02.! a m“.

;

 

 

129

increase in grounding electrode current at the end of the line. Figure 37
shows the increase in neutral-to-earth voltage at the end of the line as a
result of the increase in grounding electrode current.

Figure 40 shows the end of the distribution line with simulated
secondary network attached at node 57. In this case the secondary
ground fault is in-phase with respect to the primary load. Note in
Figure 40 that the normal line load produced grounding electrode
current is out-of-phase with the secondary ground fault current flowing
on the grounding electrodes. There will be a net reduction of the
grounding electrode current at the end of the line when there is an
in-phase secondary ground fault and the neutral-to-earth voltage will be
reduced as shown in Figure 38.

The previous analysis of secondary ground faults assumes a 0
degree or a 180 degree phase difference between the current flowing in
the neutral, grounding electrodes and earth as a result of normal line
loading and secondary ground fault current. This will only occur if the
power factor of the primary distribution line and of the secondary
ground fault circuit are unity. This analysis shows the maximum
changes possible as a result of a secondary ground fault. If the primary
neutral grounding electrode current and the secondary fault current
flowing on the grounding electrode are at some phase angle difference
other than 0 degrees or 180 degrees, the resultant current flowing in the
grounding electrode will be at some magnitude less than the maximum

shown here and more than the minimum values. It would, therefore, be

130

expected that less extreme results would occur for a secondary ground
fault on an actual operating single-phase distribution line.
The profiles obtained from the single-phase AC model

corresponding to Figure 33 through Figure 38 are listed as Figure All
through Figure A16 in the appendix.

5.1.8 Primary Phase-to-Earth Fault

Distribution line faults can occur because the lines are exposed to
the natural elements. Strong winds, heavy rains, lightening and
tornado-like weather are facts of life in most parts of the United States.

Figure 41 presents the neutral-to-earth voltage simulation curves
resulting from fault current through a resistance between the primary
ungrounded line conductor at node 416 and the earth in the single
phase distribution line model (circuit connection in Figure 8).

Compared with the base case, when fault current was increased
from 0 A (base case) respectively to 3 A, 6 A, 9 A and 12 A, the
neutral-to-earth voltages increased accordingly in some segments near
the substation and decreased some distance away from the substation
toward the end of the line. This is shown in Figure 41. Note the most
significant changes of the neutral-to-earth voltage occurred near node 1
where the substation was located. Near the substation, in the base case,
when fault current was zero, the value of neutral-to-earth voltage was
9.2 V. When the fault was increased to 3 A, the voltage also increased
by 1.0 V, to 10.2 V. When the fault was increased to 6 A, the voltage

131

PRIMARY-GROUND FAULT IMPACT ON NEV

10

VOLTAGE (V)

Figure 41.

DC MODEL, FAULT 3, 6, 9 8c 12 A AT 416-O

 

 

 

 

 

TjTrT—TllTIITIITTTTWIITTTTITTTTITIITTTTTTTIIITTTTTIITVTT
.— n o in o n o in o n o tnrx
.— -- N N n n vr v. in non

NODE # (1 -57) 9405300

Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with an ungrounded phase conductor

to earth fault at node 416 compared with the normal line (single-phase
DC model simulation)

132

increased by another 1.0 V, to 11.2 V. When the fault was increased to
9 A, the voltage increased again by 1.0 V, to a level of 12.2 V. When the
fault was increased to 12 A, the voltage reached 13.2 V (increased by
43.5% from the base case of 9.2 V). The neutral-to-earth voltage
increased linearly with the fault current at a rate of 0.33 V/A near the
substation node 1.

This effect of the change of the neutral-to-earth voltage
diminished rapidly as the distance increased from substation node 1
toward the end (node 57) of the line. In a 12 A fault, at the middle
(node 33) of the line, the voltage was 4.3 V, 95.6% of base case value 4.5
V; at the end node 57, the voltage was 4.8 V, 98% of the base case value
of 4.9 V.

Figure 42 presents the comparison of the neutral-to-earth voltage
simulation curves resulting from a 12 A simulated fault from the
primary line conductor to earth introduced at three locations -- node
404, node 416, and at node 428. It can be seen that the neutral-to-earth
voltage profiles along the distribution line were identical for all three
fault cases. An observed abnormally high neutral-to-earth voltage near
the substation may be caused by a primary ungrounded line conductor
to earth fault at any point along the whole single phase distribution
system.

It is important to note from Figure 41 and 42 that the
neutral-to-earth voltage decreased from node 15 to the end of the line
when a primary ungrounded line conductor to earth fault occurred at

any location along the single phase distribution line. ‘Even when the

133

PRIMARY-GROUND FAULT IMPACT ON NEV
DC MODEL, FAULT 12 A AT 404-0, 416-O & 428-O

 

 

 

 

lO
5-
A
2, 0+ -------- - ------------------------------------
LJJ
O
<
’—
_J
o -5.
> —+— BASE
—a— 404
. . -I— 416
_10-1 .’
+ 423
'15 FTFTIIIITIITIIIlIIIFIITIIIIIIIIIIIIIIIIIIIIIITIIIITTTTTT
'- W O In C tn 0 in O I!) O n l\
v— r- N N f") n V V’ L0 L01")

NODE # (1 -57) 94053013

Figure 42. Profile of the neutral-to—earth voltage along the primary distribution
line from substation to 5 7th node with 12 A phase conductor to earth

fault at node 404, 416, or 428 compared with normal line (single-phase
DC model simulation)

134

primary fault to earth occurred near the end of the line (node 428) the
neutral-to-earth voltage was still decreased at the fault node as
compared to the base case without a fault condition. From Figure 41 it
also can be seen that it is not possible to determine that a primary phase
line to earth fault is actually present (node 416) by observing a
neutral-to-earth voltage abnormalities in the area of the fault.

Figure 43 presents the neutral-to-earth voltage simulation curves
resulting from fault current through a resistance between the one
primary phase line conductor (phase A, referring to Figure 9) at node
416A and the earth in the three-phase four-wire distribution line model
(Yo/Y0 connection).

From Figure 43, it can be seen that under normal balanced and
no fault operational condition (base case), the neutral-to-earth voltages
in the three-phase four-wire (Yo/Yo connection) distribution system
were zero all along the neutral line. Compared with the base case, when
fault current was increased from O A (base case) respectively to 3 A, 6
A, 9 A and 12 A, the neutral-to-earth voltages increased accordingly all
along the distribution line. Note the most significant changes of the
neutral-to-earth voltage occurred near node 1 where the substation was
located. Near the substation, in the base case, when fault current was
zero, the value of neutral-to-earth voltage was also zero. When the fault
was increased to 3 A, the voltage increased by 1.0 V, to 1.0 V. When the
fault was increased to 6 A, the voltage increased by another 1.0 V, to 2.0
V. When the fault was increased to 9 A, the voltage increased again by
1.0 V, to a level of 3.0 V. When the fault was increased to 12 A, the

135

SINGLE LINE TO GROUND FAULT ON NEV, 4.8KV
3-PHASE MODEL, FAULT 3, 6, 9 8c 12 A AT 416-O

 

 

 

 

 

5
—~+—BASE
4 —a—:5A
+6A
34 +9A
A -
> +12A
v
8
2
<
1.—
.1
O
>
-1 111[IITITITTTTIIIIIITTTIIIIIIIIIITITTIIITITIFTTTTITIIII
'- In O 10 O in O to O in O 1.05
.- — N N n n v- v- 1:1 mm

NODE # (1 —57) 9405301d

Figure 43. Profile of the neutral-to-earth voltage along the three-phase primary
distribution line from substation to 57th node with an ungrounded
phase conductor (phase A) to earth fault at node 416A compared with
the normal line (three-phase AC model simulation, also referring to
Figure 6 and Figure 9 for the circuit connection)

136

voltage still increased by 1.0 V, to 4.0 V. The neutral-to-earth voltage
increased linearly with the fault current at a rate of 0.33 V/A near the
substation node 1.

This effect of the change of the neutral-to-earth voltage
diminished rapidly as the distance increased from substation node 1
toward the end (node 57) of the line. In a 12 A fault, at the middle
(node 33) of the line, the voltage was 0.2 V, a very slight increase
compared with base case value of 0 V. From the node 40 toward the
end node 57, the voltages were 0.1 V, very close to the base case value
of O V.

Figure 44 presents the comparison of the neutral-to-earth voltage
simulation curves resulting from a 12 A simulated fault from the
primary single phase-line conductor to earth in the three-phase
distribution system introduced at three locations -- nodes 404A, 416A,
and 428A. It can be seen that the neutral-to-earth voltage profiles
along the distribution line were almost identical for all three fault cases.
An observed abnormally high neutral-to-earth voltage near the
substation may be caused by a primary single line conductor to earth
fault at any point along the whole three-phase distribution system.

From Figure 41 through Figure 44, it can be seen that primary
phase line to earth fault occurred somewhere in the distribution line
systems (single-phase or three-phase system) will cause the
neutral-to-earth voltage changes all along the lines. These voltage
changes are fault-size dependent and fault-location independent. The

voltage changes increase as the fault currents increase. The most

137

SINGLE LINE T0 GROUND FAULT ON NEV, 4.8KV
S-PHASE MODEL, FAULT 12 A AT 404, 416 a 428 -0

 

 

 

 

5
n —+—BASE
4 1
\‘
+404
\\‘_ +416
3-4 '3“
A +428
>
V \‘
LL]
3 2‘
.—.
..J
O
>
1-1
0. '- ++++++++++++++++++
-1 Ill[ITTrTrTlTTTTTTI[IT]TTIIITIlTjTUIIIIII'TTTTWUIIIIF
'- In C In C In C In C In 0 In
'- v— N N n n V V In 1011':

NODE # (1 -57) 940531 e

Figure 44. Profile of the neutral-to-earth voltage along the three-phase primary
distribution line from substation to 5 7th node with 12 A phase
conductor (phase A) to earth fault at node 404A, 416A, or 428A
compared with normal line (three-phase AC model simulation, also
referring to Figure 6 and Figure 9 for the circuit connection)

138

significant changes of the neutral-to-earth voltage occur near the
substation. The voltage profiles of the same faults occurred at the three
different locations (the middle and the both ends of the line) are almost
identical for the corresponding single- or three- phase models.

Between the single-phase and three-phase distribution system, the
neutral-to-earth voltage changes caused by a primary phase line to earth

fault are different on the following two aspects:

1. Along the single-phase distribution system, the voltages increase
in some segments near the substation and decrease some distance
away from the substation toward the end of the line compared
with the corresponding base case; while along the three-phase
distribution system, the voltages increase all along the distribution

line compared with the corresponding base case.

2. The overall neutral-to-earth voltages were substantially lower
along the electric power distribution system of the three-phase
four-wire Yo/Yo connection than that of the single-phase. This
also should be true for other case studies in preceding sections.
The reason is that along the neutral line of the three-phase
system, the neutral current consists of three-phase components
with different phase angles 120 degrees out-of-phase and the sum.
of these three vectors (or phasors) under the balanced operational

condition should be zero. This greatly reduces the base

139

neutral-toearth voltages or neutral-to-earth voltages in the base

cases.

The behaviors of neutral-to-earth voltages along the three-phase
distribution systems under different operational conditions have not
been well documented prior to this study.

The profiles obtained from the single-phase AC model
corresponding to Figure 41 and Figure 42 are listed as Figure A17 and
Figure A18 in the appendix.

5.2 Guidelines to Identify NEV along the Distribution Systems

Based upon the research analysis of the computer simulation
results on neutral-to-earth voltage in the preceding sections, the
following guidelines can be established to identify the sources of
neutral-to-earth voltage originating from the parameter changes of the

power distribution systems.

A. Excessive Neutral-to-Earth Voltage Occurs near Substation

If the field investigation reveals the excessive neutral-to-earth

voltage near the substation, the possible reasons are:

1. load demand imbalance in primary three-phases.

140

2. high resistance primary neutral connection or undersized neutral

conductor near the substation.

3. secondary in-phase ground fault near the substation.

4. primary ground faults at any point along the distribution line.

5. high resistance substation grounding (e.g., possible ground mat

corrosion).

6. Heavy load demands (or line to neutral faults) at any point along

the distribution line.

B. Excessive Neutral-to-Earth Voltage Occurs away from the
Substation

If the field investigation reveals the excessive neutral-to-earth

voltage away from the substation, the possible reasons are:

1. load demand imbalance in primary three phases.

2. high resistance primary neutral connection or undersized neutral

conductor near that location.

141

3. high resistance grounding or inadequate grounding (e.g., ground

rod corrosion) near that location.

4. secondary out-of-phase ground fault near that location, either at

the sight or at a neighbor location in the immediate area.

5. heavy load demand (or primary line to neutral fault) near that

location.

Standard neutral-to-earth voltage measurement and diagnostic
techniques are appropriate and should be followed at the location
where neutral-to-earth voltage has an elevated level. Procedures
published by Surbrook et al. (1988) are good procedures to follow. If
these procedures fail to find the source of the elevated neutral-to-earth
voltage at the farm, then these guidelines can be used to more
efficiently identify an off-farm cause for the elevated neutral-to-earth

voltage.

5.3 Field Tests and Analysis of Neutral-to-Earth Voltage Gradient

Farm field tests were conducted to study the neutral-to-earth
voltage gradient distribution near grounding electrodes on the primary
and secondary sides of a neutral isolated distribution transformer.

Because the secondary ground rod is most likely placed in the

gradient of the primary ground rod, and that the earth has a finite

142

resistance, a purpose of this study was to determine how much
secondary neutral-to-earth voltage would arise from the primary
electrical system. And finally, what are parameters which influence the
amount of primary neutral-to-earth voltage which will appear on the
isolated secondary electrical system neutral. The equivalent circuit of
the isolated primary and secondary electrical system neutrals and
grounds as proposed by Althouse (1990) was studied to determine if it

was accurate, and if modifications were necessary.
5.3.1 Test of NEV Gradient Distribution, Primary One Ground Rod

A specific test voltage was applied to the ground rod utilized as
the primary transformer ground rod. This was done using a variable
transformer between the test ground rod and a 6.2 ohm grounding
electrode system located 400 feet from the test ground rod. The
distance was made large so that the earth voltage gradient would not
influence each other. The primary ground rod resistance measured by
the three point fall of potential method was 18.6 ohms.

Voltages in the following tables and figures are ground rod to
reference ground voltages. The reference ground representing true
earth was located 400 feet from the test ground rod and more than 400
feet from any other grounding electrode. The utility distribution system
in the area was an ungrounded delta system and there were no

underground pipelines or overhead transmission lines in the area.

143

Table 9 presents the test data of the neutral-to-earth voltage
gradient distribution resulting from one earth electrode rod as the
primary ground rod at a distribution transformer. Figure 45 shows the
corresponding three dimensional graphic plot of the voltage gradient
distribution and Figure 46 shows the equi-potential lines of the voltage
gradient distribution in this case. For detailed analysis, the voltage
gradient distribution in the smaller inner square ABCD of the test plot
(referring to Section 4.3.1) is presented here.

Table 9, Figure 45 and Figure 46 show that the ground rod
neutral-to-earth voltage created the highest peak at the very immediate
vicinity of the ground rod and then diminished very rapidly along all
directions. Nearly 60% of the voltage dropped within the distance of 1
foot (0.3 m) from the ground rod and nearly 80% of the voltage
dropped within the distance 6 feet (1.83 m) from the ground rod. The
voltage distribution was symmetric along all directions.

In order to confirm the symmetric distribution of the voltage
gradient of the single rod, from the center (20, 20) of the test plot, along
the four radial directions of straight north, straight south, straight west
and straight east, the earth voltage measurements were taken up to 100
feet (30.48 m) distance. Table 10 shows the measurement data and
Figure 42 shows the graphic plot of the data. From Table 10 and Figure
47, it can seen that the voltage gradient distributions along four
directions are almost identical and symmetrical.

144

 

 

N¢.~ NO.— so.— so.“ no.“ 08.“ ms.— mo.“ o.~ on.— an.“ ad
vo.N OO.N NO.N Na." ha.— am." «0.. hh.~ oo.— to.~ on.~ ~—
w~.~ NN.N -.N v«.N oo.~ a0.— '0._ no.“ as.“ an." no." N—
.v.~ vv.~ om.~ m.N MN.“ ou.N nO.N 00.— 50.. o.— 00.« mg
no.N vo.~ 0.N 0v.N on.“ 0N.N 0—.N vo.~ 06.— v0.— nh.— Va
oa.~ NO.N no.~ uh.N on.~ NQ.N ~m.~ o—.N 0O.— v¢.~ o." n“
NM.M 0N.m uo.m nO.N o.N on.N Nv.N vN.N 0°.N no.“ '0.“ o—
No.m v¢.m Nv.m c—.m so.“ on.“ Nfl.~ NM.N vu.N OO.N oo.~ Nu
v.v ~N.v mo.m on.m «u.m NO.N mo.N Nv.N N.N oo.~ «0.— on
mv.n No.n hfl.v kh.m 0N.m OO.N wo.N ov.~ MN.N oo.~ n5.~ on
vs.~— on.» n.v n0.m om.n mo.n vs.“ ov.~ VN.N oO.N 06.“ ON
nm.n «o.n v.1 o.m an.m «o.m oo.N nv.~ QN.N NO.N na.¢ —N
n. v.v to.v va.m o—.m O.N 0.N hm.N «N.N mo.~ no.“ «N
oo.m 0&.m m¢.m NN.m no.m nu.~ on.N mm.N v~.N 00.« a.— MN
~v.m om.m VN.m on.~ on.~ ~0.N nm.~ v~.N to.“ ‘0." no.~ vfl
~6.m 00.N ~0.N «N.N on.~ nv.~ NM.N n—.N 00.. 00.— ON.— nu
-.~ oo.~ vo.~ ~n.~ ~v.~ m.~ o~.N ~0.N do.“ N¢.~ MN.— 0N
o¢.N nv.N om.N nm.~ VN.N n—.N vo.~ vo.~ do.“ mn.~ 00.— NN
VN.N MN.N 1N.N o—.~ co.“ «O.N o.« mo.~ vs.— oo.— o.— ON
oo.~ no.“ vo.N «o.— to.“ n¢.— vo.— as.” no.« kn.“ nn.~ 0N
oo._ ao.— oo.~ oo.~ no.— on.“ «5.. v0.— mn.~ mn.— ov.~ on
ON a“ o— s— 0‘ n— Cu mu N— «a on nasv x .Aasv >

 

«Immhova .068 azaoam uzo >¢¢t~¢t .CFCO haw» h2fi~a¢¢a >mz

Coo“ & :5. Ease: no. 035020 .836 03:7. Bo: cozantfio
2.060% 093.9 Eager—830: 05 .3 San 20805308 amok. d 0305.

145

 

 

vv.. .0.. 00.. 00.. 00.. ms.. 0.. N0.. 00.. 00.. 0.
.0.. 00.. 00.. N5.. 05.. 00.. .0.. 00.. '0.N 00.N ..
00.. 00.. Nn.. 05.. 00.. N0.. ~0.N m..N v..N .N.N N.
00.. .~.. N0.. .0.. 00.. ...N N.N 0N.N 00.N 0v.N m.
00.. 0h.. 00.. 00.. 00.N 0N.N 0N.N 0v.N v0.N h0.N v.
05.. .0.. .0.. 00.N NN.N v.N v0.N 00.N m0.N v0.N 0.
0b.. .0.. 00.N m..N m. 00.N .N.N .0.N no.0 NN.m 0.
.0.. 00.. 00.N N.N Nv.N 00.N N0.n NN.m 00.0 .h.m N.
0.. .0.N ...N MN.N '0.N v&.N m..m .v.m 00.0 MN.v 0.
N0.. N0.N N.N v.N N0.N 00.N VN.m v0.0 .m.v N..0 0.
00.. N0.N 0..N mn.N .0.N 0.N 5N.m .~.m 00.0 mv.0 ON
00.. N 0..N NM.N 0.N N0.N .N.m N0.m N.v v0.v .N
00.. .0.N 0..N 00.N 00.N MN.N 00.0 Nv.m 00.0 0N.v NN
00.. 00.. ...N 0N.N 0v.N v0.N .0.N 0..m v0.0 mn.m MN
08.. 00.. N0.N 0..N 0M.N N0.N N&.N N0.N v..0 .m.m 0N
05.. 05.. 00.. ~0.N 0N.N Nm.N '0.N N0.N 00.N ~0.N 0N
.h.. 00.. 00.. 00.. '0.N NN.N 00.N 0Q.N s0.N v0.N 0N
N0.. 00.. 0.. 00.. 00.0 00.N N.N 0N.N vm.N v.N NN
00.. 00.. .h.. 00.. 00.. 00.. 00.N ...N 0..N .N.N 0N
.0.. 00.. N0.. 05.. 0b.. «0.. 0.. n0.. 00.. '0.N 0N
Nv.. 09.. 00.. 00.. 00.. mu.. 0.. .0.. 00.. 00.. 00
00 0N 0N sN 0N 0N vN 0N NN .N “at. x ..ubv >

 

.imNhov0 .000 020000 020 >¢¢t.¢0 .0h00 hmmb h20.0¢¢0 :02

0.233 a .2:

146

 

AEUTRAL—TO—EAPTH VOL TAGE ( l/J

 

 

Figure 45- Three dimensional graphic plot of the neutral-to-earth voltage

gradient distribution from single driven electrode rod (length unit is
foot)

147

EQUl—ROTENTIAL LINES, ONE ROD, DETAIL, 6—9—94

 

 

 

 

 

 

10 1s 20 25 30
30 r I r 30
25 *- - 25

W

10

n]
N 20 - 20
1s - ~ 15

e
10 1 1 1 1o
10 1s 20 25 30

Figure 46. Equi-potential lines of the neutral-to-earth voltage gradient
distribution from single driven electrode rod (length unit is foot)

148

Table 10. Test measurement data of neutral-to-earth voltage gradient
distributions from single driven electrode rod along four directions

The NEV Gradiont with 5109!. Rod along 4 diroctions 940808-1

 

 

 

DistancozNorth :East :South :Uost 3
(inch) 0 12.64 12.8 12.72 13.09
1 8.94 9.03 8.75 8.96
2 7.99 7.86 7.76 8.4
3 7.51 7.37 7.41 7.74
4 7.08 7.05 7.09 7.41
5 6.75 6.74 6.85 7.15
6 6.53 6.47 6.6 6.72
7 6.4 6.22 6.21 6.5
8 6.3 6.01 5.82 5.96
9 6.24 5.83 5.76 5.72
10 5.56 5.63 5.68 5.62
11 5.42 5.52 5.63 5.54
12 5.33 5.44 5.5 5.47
15 5.05 5.15 5.27 5.24
18 4.88 4.84 4.92 5.04
21 4.71 4.58 4.67 4.74
24 4.47 4.4 4.52 4.6
30 4.14 4.04 4.15 4.25
36 3.9 3.82 3.84 3.96
42 3.6 3.47 3.57 3.68
(Foot) 4 3.39 3.25 3.36 3.47
S 3.02 2.91 3.03 3.11
6 2.7 2.63 2.73 2.82
7 2.43 2.39 2.46 2.51
8 2.22 2.17 2.31 2.31
9 2.04 1.98 2.08 2.15
10 1.89 1.88 1.96 1.99
15 1.35 1.38 1.45 1.43
20 1.07 1.11 1.15 1.15
30 .79 .78 .81 .79
40 .58 .59 .61 .62
50 .47 .46 .48 .49
60 .39 .36 .4 .42
70 .31 .29 .33 .33
80 .27 .24 .27 .28
90 .22 .21 .23 .24

100 .21 .18 .2 .21

 

149

The NEV Gradient along 4 directions
SINGLE ROD ELECTRODE

 

 

 

 

151
“‘B—- N
+ E
10*
A “F S
>
v + w
LiJ
O
<
}_..
__1
O
>
5..
O T T T T
O 20 4O 60 80 100
DISTANCE FROM CENTER (FEET) 940606-1

Figure 47. Neutral-to-earth voltage gradient distributions from single driven
electrode rod along four directions (length unit is foot)

150

Theoretically speaking, (IEEE, 1983; IEEE, 1991; Tag, 1964 and
Kraus, 1953) for the single-driven ground rod electrode in a uniform

soil, its earth resistance is given by the formula:

 

 

 

 

 

R = 1’ Ln 4 x L (5.1)
2 x «XL D
and its ground electrical potential away from the rod is
v: ”P (1.114"L _Ln X2 _Ln X3 ...) (5.2)
2 x 1r x L D X1 X2

where p is resistivity, L is the length of the rod, D is diameter of the rod,
X1, X2, X3, are horizontal distances from the rod, I is the current, R
is ground resistance, and V is the ground potential away from the rod.

Taking the voltage at the ground rod as reference base, the
equation (5.2) divided by I x R will yield the per unit normalized
neutral-to-earth voltage:

U (per unit) = 1 - [Ln(X2/X1) - Ln(X3/X2) - ]/Ln(4xL/D) (5.3)

The test rod length is 8 feet (2.44 m) and the diameter D is 5 / 8 inch
(0.0159 m). Plotting out the voltage profile up to 15 feet (4.6 111)
defined by (5.3) with the comparison to the test measurement data
along the east direction, the result is shown in Figure 48 and the
corresponding numeric values is shown in Table 11. The test data from
Table 10 was normalized by dividing all values by the "voltage at the

Table 11.

151

Data comparison between theoretical value calculated from equation
(5.3) and field measured value of neutral-to-earth voltage gradient
distributions from single driven electrode rod along east direction
(length unit is inch)

Comparison 0? NEU Grodiont to tho Colculotod with 1 Rod 6-19-94-1

 

Diot.-inlEootTootholcu1ot:Toot-Col1

 

 

.3125 1.0000 1.0000 .0000 LN(4§L/d) = 6.4206
1 .7055 .8189 -.1134 d/2 (inch) = .3125
2 .6141 .7109 -.0968 L (Foot) = 8.0000
3 .5758 .6478 -.0720
4 .5508 .6030 -.0522
5 .5266 .5682 -.O416
6 .5055 .5398 -.0343
7 .4859 .5158 -.0299
8 .4695 .4950 -.0255
9 .4555 .4767 -.0212
10 .4398 .4602 -.0204
11 .4313 .4454 -.0142
12 .4250 .4319 -.0069
15 .4023 .3971 .0052
18 .3781 .3687 .0094
21 .3578 .3447 .0131
24 .3438 .3239 .0198
30 .3156 .2892 .0265
36 .2984 .2608 .0377
42 .2711 .2368 .0343
48 .2539 .2160 .0380
60 .2273 .1812 .0461
72 .2055 .1528 .0527
84 .1867 .1286 .0579
96 .1695 .1080 .0615
106 .1547 .0897 .0650
120 .1469 .0733 .0736
180 .1078 .0101 .0977
TEST 0 U_:o1 = 1-(LN(X2/X1)-LN(X3/X2)-...)/LN(4§L/d)
CHLC 180

Cooporioon of NEV Grodiont to tho Colou1otod
SINGLE ROD ELECTRODE
For Unit Noroolixd

DISTflNCE FROM CENTER (INCH)

940619-1

152

Comparison of NEV Gradient to the Calculated
SINCLE ROD ELECTRODE, TO EAST DIRECTION

 

1 0

 

 

 

I --8— TEST

.8 ‘ + CALC
>
LL] I
Z 1.
o I
N .1
'3 .6 d '.
o 11
g “I.
0 ii;
2 .
f: .41
C
:3 "5 .2
6 F = =
C1. :

.2 " = = ‘

- a
O T T T r T T I T 1
O 20 40 50 80 100 120 I40 160 180
DISTANCE FROM CENTER (INCH) 940619-1

Figure 48. Comparison between theoretical value calculated from equation (5.3)
and field measured value of neutral-to-earth voltage gradient
distributions from single driven electrode rod along east direction
(length unit is inch)

153

ground rod. In Table 11, the first column is the distance (inch) from the
center of the rod, the second column is the per unit voltage from the
test measurement, third column is the per unit voltage calculated from

(5.3) and third column is the difference of the two values (test value -

calculated value).

From Figure 48 and Table 11, it can be seen that within the first
20 inch (0.51 m) from the ground rod, the test value and the theoretical
calculated value of the earth voltage were very close. After this point,
the discrepancy increased as the distance from the center of the rod
increased.

The test results indicate that gradients extend out further
distances than determined by the theoretical calculation. For example,
at 15 feet (180 inches), the theoretical calculation indicates that the
voltage is only 1% of the ground rod voltage while the test data
indicates the voltage is 10%.

Althouse (1990) conducted similar measurements of the voltage
gradient in the immediate area of a single ground rod. That research
only shows the gradient out to a distance of 84 inches from the ground
rod. For a ground rod with a resistance to earth of 20.6 ohms, the
voltage at 84 inches was 18% of the voltage at the ground rod. When
the ground rod to earth resistance was 118 ohms, the voltage at 84
inches was only 6% of the voltage at the ground rod. The results from
this research (Table 11) with a ground rod resistance to earth of 18.6
ohms at 84 inches was 19% of the voltage at the ground rod. The

calculated value from the theoretical formula showed that at 84 inches

154

the voltage was 13% of the voltage at the ground rod. The formula is
approximately true, but does not take into consideration the influence
of earth resistivity at distance away from the ground rod. In low
resistance soils, the voltage gradient extends out from the ground rod a

greater distance than for high resistance soils.
5.3.2 Test of NEV Gradient Distribution, Primary Two Ground Rods

Table 12 presents the test data of the neutral-to-earth voltage
gradient distribution resulting from two earth electrode rods acting as
the primary isolated ground at the transformer. Figure 49 shows the
corresponding three dimensional graphic plot of the voltage gradient
distribution and Figure 50 shows the equi-potential lines of the voltage
gradient distribution in this case. For detailed analysis, only the voltage
gradient distribution in the smaller inner square ABCD of the test plot
(referring to Section 4.3.1 and 4.3.2) is presented.

Table 12, Figure 49 and Figure 50 show that the neutral-to-earth
voltage created two peaks at the very immediate vicinity of the two
ground rods and then diminished rapidly along all directions. The
resistance to earth of the two ground rods as measured by three point
fall of potential method was 11.4 ohm. More than 55% of the voltage
dropped within the distance of 3 feet (0.91 m) from each of two ground
rods.

In order to detect the altering of the voltage gradient distribution
compared to that of the single rod, from the center (20, 20) of the test

155

 

0v.N Nv .N v.N 0.N 0.N .0.N VN.N N.N .. .N 8N N0.N 0.

 

00.N .0.N 00.N .0.N 00.N 0v.N 8N 0.N .N.N 5. .N ..N ..
00.N 55 .N 05.N 5.N N0.N 00.N 59.0. 00.N 00.N 0N.N 0. .N N.
0 .0.0 50 .N 00.N 05.N 00 .N .0.N 00.N 00.N 8N 0N.N 0.
0N.0 MN .0 0. .0 8.0 50 .N 50 .N 55.N 00.N 00.N 0v.N N0.N v.
00.0 N0.0 3.0 0N.0 0. .0 50.0 N0.N 05 .N N0.N 00.N 0.N 0.
00 .0 0.0 .5 .0 00.0 00.0 NN.0 8.0 00.N '5 .N N0.N 00.N 0.
Nv... 00... 000 N0.0 N0.0 0v.0 .N.0 v0.0 .40.... 05.N 0.N 5.
N0... .0.v 09.0 .N.v N0.0 00.0 8.0 v. .0 N0.N 05.N 0.N 0.
00.0 .0.0 00.0 0.0 No.0 '5 .0 00.0 0N.0 8N .0.N 00.N 0.
00.N. .0.0 0. .0 5.0 0. .v 00.0 00.0 .0 .0 00.0 8N 00.N ON
00.0 N0.0 0N.0 '5 .v vN.v N0.0 .0.0 0.0 .. .0 00.N N5.N .N
00.0 00.0 N. .0 .5.v 0N.0 00.0 N0.0 .0.0 N. .0 00.N 05.N NN
5v.0 00.0 50.0 00.0 0N0 00.0 N0.0 00.0 .. .0 8N 05.N 0N
v0.0 3.0 8.0 00.? NN.v 50.0 00.0 00.0 6.0 8N .5.N 0N
00.0 05 .0 8.0 N0; 0. .v 00.0 00 .0 0 .0 00.N .0.N 50.N 0N
50.N. 5. .0 .0.0 00.0 8.? 05 .0 Nv.0 0. .0 RN .0.N N0.N 0N
50.0 .0.0 N0.v 006 00.0 00.0 00.0 v. .0 00.N N5.N 00.N 5N
.0.v N0... :6 0. .v '5 .0 50.0 N.0 No.0 0.N 00.N 0.N 0N
VN.v NN.v '0; 00.0 50 .0 .N.0 8.0 50 .N 5.N 00.N :.N 0N
0.0 55.0 0.0 3.0 0N.0 N. .0 00.N 55 .N .0.N 00.N 00.N 8
ON 0. 0. 5. 0. 0. v. 0. N. .. 0. 350 xzvuv >

 

N10N5000 .080 2600 g $.00 :38 #00... 5.20.9.6 0U.

Coo. a. 5.2: 50:00 00:80.0 6080000
.00. 0 5.3 60: 0:090 0.008020 a»; 806.00 80:. 60.5505
60580 008.9 5000-00-70.50: 05 .o 0.00 6060:3008 80... .N. 030....

156

 

 

00.. N0.N 50.N 0. .N NN.N 0N.N 00.N *.N 00.N 00.N 0.
00.N N. .N 0. .N 5N .N N0.N 00.N 00 .N 00.N .0.N N0.N ..
.. .N 0. .N .0 .N *.N N0.N 00.N 00.N 5.N 05.N 05.N N.
N.N 0N.N 00.N 00 .N 50.N 5.N .0.N 50 .N 50.N 00.0 0.
0N.N 00.N 50 .N 0.N N5.N 00.N 00 .0 50 .0 0. .0 0N.0 0.
00.N 00.N 00.N 5.N 00.N 00.0 0. .0 0.0 00 .0 00.0 0.
.0.N 50.N N5.N 0.N 3N 0. .0 0.0 00.0 05.0 00.0 0.
00.N 00.N 05.N 0.N 0. .0 8.0 05.0 00.0 0. .0 0N.0 5.
00 .N 05 .N 00.N N0.N .0 .0 0.0 .0 .0 N. .0 00.0 .0.0 0.
00.N 05.N 00.N NN.0 00.0 00.0 8.0 N0 .0 00.0 0.0 0.
00.N .0.N 00.0 NN.0 00.0 0.0 .N.0 0.0 0. .0 8.0 0N
00.N 00.N 00.0 0N.0 N0.0 0.0 0N.0 00.0 N.0 05.0 .N
00 .N 00.N 0. .0 8.0 00.0 00.0 5N.0 00.0 N. .0 .0 .0 N
.5.N 50.N ..0 5N.0 .0.0 50.0 0.0 00.0 00.0 0N.0 0N
00.N 00.N 50.0 .0 .0 00.0 00.0 0. .0 00.0 00.0 0.0 0N
00.N 05.N .0.0 N.0 N0.0 0.0 0. .0 N0.0 8.0 00 .0 0N
00.N 05 .N 00.N 0. .0 00.0 00.0 00.0 00.0 8.0 00.0 0N
00.N 00.N 0.N .. .0 50 .0 00.0 00.0 0. .0 55.0 00.0 5N
00.N 00.N .0.N 0.N 0N.0 0.0 55.0 00.0 00.0 05.0 0N
00 .N .0 .N 00.N 00.N . .0 N.0 0.0 05 .0 50 .0 0. .0 0N
.0.N 00.N 0.N 05.N 00 .N 50.0 0N.0 0.0 N0.0 05 .0 8
a 0N 0N 5.N 0N 0N 0N 0N NN .N A»: x. no: >

 

NI0N5000 .088 2g 9.... $.80 .02! #00... 5.98 0!

5.83 ,2 «a...

157

 

 

 

3 l
‘6‘
E 9 ~
9i
i \
CE \.
é “ ‘ /4,‘::3i) ET?“
/ \‘EEE:
é! 7251”‘I:“II ‘ k§§$i§§3 \\
E ‘ \ '555’1ll’ 'I’ '0 Q Q ‘\\\§§§§&:§
6': I to .0 w 0
‘fi '242'5:"$:'o'0.'o’o’&:s“§‘i‘a 383%
‘ 0
O

    

.2 I. .‘0 <)
NEV GRADIENT. TLJO RODS. TEST B. 6-1—94

Figure 49. Three dimensional graphic plot of the neutral-to-earth voltage

gradient distnbution from two electrode ground rods with 6 feet
separation distance (length unit is foot)

158

EQUl—POTENTIAL LINES WITH TWO RODS (DETAIL), 6—9—

10
30

 

30

25*-

20 - 20

 

 

 

to P 1 l 10

 

Figure 50. Equi-potential lines of the neutral-to-earth voltage gradient
distribution from two electrode ground rods with 6 feet separation
distance (length unit is foot)

159

plot, along the four radial directions of straight north, straight south,
straight west and straight east, the earth voltage measurements were
taken up to a 100 foot (30.48 m) distance. Table 13 shows the voltage
profile data of the two-rod primary ground system along the north
direction and the Figure 51 shows the graphic plot of the voltage profile
data comparison (per unit normalized) between one rod and two rods of
the primary ground systems along the north direction. Table 14 and
Figure 52 show the voltage comparison (per unit normalized) along the
other three directions.

From Table 12, Table 13, Table 14, Figure 49, Figure 50, Figure
51 and Figure 52, compared with that of one primary ground rod
connection, it can seen that the most significant altering of the
neutral-to-earth voltage gradient by the primary two-rod ground system
occurred within the 30 foot distance along the north direction which is
the direction of straight connection line of two rods. The second peak
created by the additional primary ground rod not only elevated the
neutral-to-earth voltage along the north direction, but also elevated
neutral-to—earth voltage along other three directions to some extent.
This elevation diminished as the distance increased away from the both
rods.

At a distance of 100 feet the voltage as measured to the reference
ground was approximately the same when there was a two ground rod
system as compared to a single ground rod. At closer distance the
voltage was higher than for a single ground rod. At six feet from either

of the two ground rods in any direction, the voltage was approximately

160

Table 13. Neutral-to—earth voltage test data (per unit normalized) from primary
ground systems of two electrode rods along the north direction

The lEV Gradient with Two Rods along N directions (Per Wit) 940609-3

 

 

Distance! North 1 Distance! North 1
(inch) 0 1.000 76 ~57“
1 .721 77 5“
2 .610 7° "'32
3 .578 79 e521
4 .560 80 .5‘3
5 .544 81 .491
6 .535 82 . .493
7 _532 83 .464
a .530 94 “‘75
9 .527 (Feet) 8 .403
10 .509 9 -3“7
11 .501 10 .302
12 .493 13 .1”
15 .486 2° 45°
19 .430 30 .107
21 .472 4° -°79
24 .464 5° -°52
27 .459 6° -°52
30 .453 7° -°42
33 .452 80 -°35
36 .450 90 -°3°
39 .450 10° -°27
42 .452
45 .454
4a .456
51 .461
54 .472
57 .465
60 .500
61 .501
62 .514
63 .517
64 .525
65 .534
66 .549
67 .593
6a .618
69 .643
70 .686
71 .766
72 .990
73 .736
74 .675
75 .625

 

161

The NEV Grodient OIOng N direction
COMPARISON OF TWO RODS 8c 1 ROD, NORTH DIRECTION

 

 

 

 

 

 

 

 

1% .
> .a 3
L1J
Z + 2RD
8 + 1RD
.2 .6 «
O
E H I]
o 1..., t'
z
._. .4« ..
'E ..
3 v
a
O- .2 '1 u
o . r H 9 fl
0 20 4o 60 80 _ 100
DISTANCE FROM CENTER (FEET) 940714-1

Figure 51. Graphic plot of the neutral-to-earth voltage profile data comparison
(per unit normalized) between one rod and two rods of primary
ground systems along the north direction (length unit is foot)

162

Table 14. Test measurement data of the neutral-to-earth voltage profile data
comparison (per unit normalized) between one rod and two rods of
primary ground systems along the other three directions

The NEV Gradient Cooparison along different directions 940615-4

 

Dist.-inlEast 2Rdl$outh2RdIHost 2RdIHorth1RdiEast 1RdlSouth1Rleost lel

 

(inch) 0 1.000 1.” 1.” 1.000 1.!!!) I.” 1.11!)

 

1 .641 .648 .654 .707 .705 .688 .684
2 .602 .601 .585 .632 .614 .610 .642
3 .564 .557 .562 .594 .576 .583 .591
4 .555 .546 .553 .560 .551 .557 .566
5 .544 .536 .544 .534 .527 .539 .546
6 .533 .523 .531 .517 .505 .519 .513
7 .523 .509 .522 .506 .486 .488 .497
8 .517 .494 .501 .498 .470 .458 .455
9 .508 .491 .491 .494 .455 .453 .437
10 .498 .488 .488 .440 .440 .447 .429
11 .492 .480 .487 .429 .431 .443 .423
12 .488 .477 .483 .422 .425 .432 .418
15 .468 .464 .470 .400 .402 .414 .400
18 .454 .438 .459 .386 .378 .387 .385
21 .425 .424 .439 .373 .358 .367 .362
24 .425 .410 .431 .354 .344 .355 .351
30 .398 .382 .404 .328 .316 .326 .325
36 .376 .360 .384 .309 .298 .302 .303
42 .358 .337 .364 .285 .271 .281 .281
(foot) 4 .341 .321 .348 .268 .254 .264 .265
5 .313 .294 .322 .239 .227 .238 .238
6 .290 .268 .296 .214 .205 .215 .215
7 .268 .246 .273 .192 .187 .193 .192
8 .249 .233 .251 .176 .170 .182 .176
9 .230 .213 .237 .161 .155 .164 .164
10 .219 .202 .221 .150 .147 .154 .152
15 .165 .155 .164 .107 .108 .114 .109
20 .134 .125 .134 .085 .087 .090 .088
30 .095 .089 .094 .063 .061 .064 .060
40 .072 .069 .073 .046 .046 .048 .047
50 .056 .05! .059 .037 .036 .038 .037
60 .046 .045 .049 .031 .028 .031 .032
70 .037 .039 .041 .025 .023 .026 .025
80 .030 .032 .034 .021 .019 .021 .021
90 .024 .026 .030 .017 .016 .018 .018
100 .022 .024 .026 .017 .014 .016 .016
940615-4

The EU Gradient along Different Directions
CDHPRRISDN BETHEEN i 6 2 ROD PRIMER? GROUNDS
Per Unit Nornaiized

DISTRNCE FRDH CENTER (FEET) 940615-4

163

The IE.” Grodient olong Different Directions
‘ ‘ EET‘.‘.’EEI~I I a. 2 ROO PRIMARY OROONOS

 

 

 

 

Per Unit Normalized NEV

 

 

 

 

DISTANCE FROM CENTER (FEET) 940615-4

Figure 52. Graphic plot of the neutral-to-earth voltage profile data comparison
(per unit normalized) between one rod and two rods of primary
ground systems along the other three directions (length unit is foot)

164

8% of the total voltage higher than for a single ground rod. For the two
ground rods, the voltage was 29% of the total as compared to 21% for
the single ground rod. At 20 feet, the voltage was 13% of the total for
the two ground rods as compared to 9% for the single ground rod. At
30 feet, the voltage was 9% of the total for the two ground rods as
compared to 6% for the single ground rod.

When comparing these results with those of Althouse (1990), it
appears the voltage is higher in the immediate area of the ground rods

when the resistance of the grounding electrode is lowered.

5.3.3 Test of NEV Gradient Distribution, Primary One Ground Rod,
Secondary Isolated One Ground Rod Without Farm Ground

Connection

Table 15 presents the test data of the neutral-to—earth voltage
gradient distribution resulting from one primary ground rod, and one
secondary isolated ground rod without the farm grounding system
connected. The circuit is illustrated in Figure 53. Figure 54 shows the
corresponding three dimensional graphic plot of the voltage gradient
distribution and Figure 55 shows the equi-potential lines of the voltage
gradient distribution in this case.

From the Figure 54 and Figure 55, it can be seen that the voltage
gradient originating from the single primary ground rod created the
highest peak at the very immediate vicinity of the primary ground rod

and then diminished very rapidly along all directions. Note in this case

F

165

 

 

.0.. .0.. N0.. 00.. 00.. 00.. 00.. N0.. 00.. 0.
00.. 00.. 0.. 0.. .0.. .0.. .0.. N0.. .0.. 0.0
00.. 00.. 00.. 00.. 00.. N 00.. 8.. 00.. 0
00.N 00.N 8N 00.N 50.N 00.N 8N 8.N 00.N 0.0
N. .N .. .N 0. .N 0. .N N 0. .N 0. .N 0. .N 0. .N 0
0N.N 0N.N 0N.N 0N.N 0N.N 0. .0.N 0N.N 0N.N . 0.5
00.N 00.N 0.N .0.N 00.N 00.N 00.N 00.N 00.N 5
00.N 0. N0.N 50.N .0.N 0. 00.N 00.N 00.N 0.0
00.N 00.N 50.N 05 .N 05 .N 05.N . 05 .N 00.N N0.N 0
.5.N 55.N N0.N 0.N .0.N N0.N 0.N 00.N 55.N 0.0
50.N 00.N 00 .0 50 .0 .. .0 50 .0 50 .0 00.0 00.N 0
RN 0. .0 0. .0 0.0 00.0 .0 .0 0N.0 0N.0 00.0 0.0
.. 0N .0 00 .0 00.0 00.0 00 .0 .0.0 0.0 .N.0 0
0N.0 00.0 00.0 N0.0 00.0 00.0 N5.0 0.0 00.0 0.0
00.0 00.0 50 .0 00.0 0. .0 0. .0 00.0 .0.0 00.0 0
00.0 0.0 .N.0 00.0 0.0 00.0 0. .0 .0.0 00.0 0.N
00.0 00.0 0.0 50.0 N0.0. N0.0 N.0 00.0 N0.0 N
00.0 00.0 0N.0 .0.0 5. .0 00.0 0. .0 .0.0 00.0 0..
00.0 N5.0 00.0 0N.0 0N.0 0. .0 00.0 0.0 00.0 .
00.0 50.0 00.0 00.0 .0.0 55.0 00.0 00.0 00.0 0.
5.0 0N.0 00.0 0.0 50.0 00.0 .0.0 N0.0 N. .0 0
0 0.0 0 0.N N 0.. . 0. 0 a»: x205. >

 

0|0N5000 $0.00? 258 g..— E... 000.0150. $00 a $.00 :38 hmmb 0.20.85 3!

Coo. a. an: 50:00 5.8288
2320 8:0 30:23 .8536 50883 «on. 8 no. 055..»
28 08205 .0.N—6:88 68 2580 use .095... :80 5033......
25:30 008:; 5.3-9-3230: 05 .o «:3 208333.: 3.0...

.n. 03a...

166

Prlnory Systen
Voltage Source

I000
—~ /

 

 

 

f Prlnoryl
7 ground '7
current No current
PrInory FI0' '" Forn
Systen secondary Grounding
Reststonce Systen
Prlnory Secondary
Ground Rod Ground Rod

 

 

It 400 F!" -l. 6 Feet 4. 600 feet 4

Figure 53. The test circuit for the neutral-to-earth voltage gradient distribution
resulting from one primary ground rod, and one secondary isolated
ground rod without the farm grounding system connected.

 

Figure S4.

167

 

NEUTRAL—TO—EAE’TH VOL TAGE ( L/J

 

 

 

Q"

NEV ORAOIENT.6-Io-94 TEST.°PARI~I R OPEN

Three dimensional graphic plot of the neutral-to-earth voltage
gradient distribution from primary one ground rod, secondary isolated

one ground rod (6 feet separation distance) without farm ground
connection (length unit is foot)

 

 

 

 

 

168

V LINES, 6—10—94 TEST, FARM R OPEN

 

 

 

T 2.8
/

3.2

/’
V. .

l
o 1 2 3 4

Z

 

 

 

 

Figure 55. Equi-potential lines of the neutral-to-earth voltage gradient
distribution from primary one ground rod, secondary isolated one
ground rod (6 feet separation distance) without farm ground
connection (length unit is foot)

169

from Figure 54, the voltage gradient experienced some slight reduction
near the north side mid-point of the test plot. This is due to the
influence from the other isolated secondary ground rod (6 feet from the
primary rod) even without connection to any primary or farm ground
system. The voltage reduction near the secondary ground rod was very
slight as can be seen by the data in Table 15.

It can be concluded that the voltage gradient produced by the
voltage on the primary ground rod is basically undisturbed by the
presence of the secondary ground rod not connected to an electrical

system.

5.3.4 Test of NEV Gradient Distribution, Primary One Ground Rod,
Secondary Isolated One Ground Rod With Farm Ground Connection

Table 16 presents the test data of the neutral-to-earth voltage
gradient distribution resulting from one primary ground rod, and an
isolated secondary ground rod connected to a 5.24 ohm farm ground
resistance. The circuit is illustrated in Figure 56. Figure 57 shows the
corresponding three dimensional graphic plot of the voltage gradient
distribution and Figure 58 shows the equi-potential lines of the voltage
gradient distribution in this case.

From Figure 57 and Figure 58, it can be seen that the secondary
isolated ground rod with the farm ground system attached created a
sharp voltage decrease near the immediate vicinity of the secondary

isolated ground rod.

 

 

 

170

 

 

 

 

00.. N0.. .0.. 0.. 0.. .0.. .0.. 00.. 50.. 0.
00.. 00.. 00.. 00.. 00.. 00.. 00.. 0.. 0.. 0.0
00.. N0.. 00.. 00.. 00.. 00.. 0.. 00.. 50.. 0
.0.. 00.. N0.. 00.. 0.. 50.. 00.. N0.. 00.. 0.0
.5.. 00.. .0.. 00.. 00. 00.. 00.. N5.. 05.. 0
N0.. 00.. 55.. 05.. 00.. 55.. .0.. 00.. 50.. 0.5
50.. N 00.. 50.. 00.. 00.. 00.. 00.. N 5
0N.N 0..N 0..N 0..N .N.N 0..N .N.N 0..N 0..N 0.0
0N.N 00.N 00.N 00.N 0.N 00.N 00.N 00.N .0.N 0
00.N 50.N 0.N 50.N 00.N 00.N 00.N 50.N 0.N 0.0
00.N 00.N 05.N .0.N 00.N 05.N 05.N 05.N .5.N 0
05.N 00.N 00.N N..0 00.0 00.0 00.0 0 50.N 0.0
00.N 00.0 00.0 0.0 0.0 0N.0 0N.0 .N.0 0 0
00.0 0N.0 00.0 00.0 5.0 00.0 00.0 00.0 5..0 0.0
N.0 00 .0 00.0 00.0 0 00.0 05.0 00.0 00.0 0
.0.0 00.0 00.0 00.0 55.0 0.0 00.0 05.0 00.0 0.N
50.0 00.0 0..0 .5.0 5N.0. 05.0 No.0 00.0 50.0 N
00.0 00.0 00.0 0.0 05.0 0N.0 00.0 0.0 00.0 0..
0N.0 00.0 00.0 00.0 0..0 00.0 50.0 50.0 5N.0 .
0N.0 0.0 00.0 0.0 00.0 00.0 00.0 N0.0 0..0 0.
.0.0 00.0 0N.0 00.0 00.0 .0.0 5N.0 0..0 50.N 0
0 0.0 0 0.N N 0.. . 0. 0 .308 x_500v >

 

010N5000 .20.5002200 020000 £000 1h.3 00h¢400. >00020000 a >0¢t.00 .0500 500» 520.0000 :02

98. a. :3 50:0: 3.80586
9590 85.. 3223 Ace—.80.. 508.503 .08 8 co.— 039.0
25 38.0.... bag—80m 6o. 2590 use b.2520 Soc Sanctum?
28680 008.? 5399-333: 2: 00 San Eon—2:32: =0...

.0. 03a...

Prinary Systen
Voltage Source

IumI

171

 

—-.
f Prinary
1? ground
current
Prinary
Systen
ReSIstance
Prinary
Ground Rod

 

I‘T 400 feet

Figure 56.

 

1 Current fron i
prinary flowing .__

In secondary T

 

grounding systen Farn
Grounding
Systen
Secondary
Ground Rod

’l. 6 feet -I- 600 feet ’

 

The test circuit for the neutral-to-earth voltage gradient distribution

resulting from one primary ground rod, and an isolated secondary
ground rod connected to a 5.24 ohm farm ground resistance.

Figure 57.

172

 

AEUTR’AL-TO-EARTH VOL 73465 ( V)
1

 

 

 

 

 

 

 

 

0)

 

 

 

 

 

 

 

 

 

 

 

4%; 5%. WI. ° i

C. ‘5’
’2) . " «3’

0 Q
NEV GRADIENT. 6-18—94 TEST.FARI‘1 R 5 OHM

Three dimensional graphic plot of the neutral-to-earth voltage
gradient distribution from primary one ground rod, secondary isolated
one ground rod (6 feet separation distance) with farm ground
connection resistance 5.2 ohm (length unit is foot)

Figure 58.

173

V LINES, 6—IO—94 TEST, FARM R 5 OHM

 

o I 2 .1 4
I0 I W I 10
9r- +9
A a—\ -

 

 

 

 

 

Equi-potential lines of the neutral-to-earth voltage gradient
distribution from primary one ground rod, secondary isolated one
ground rod (6 feet separation distance) with farm ground connection
resistance 5.2 ohm (length unit is foot)

174

In order to reveal the altering of the voltage gradient distribution
compared to that of the open farm ground connection of the secondary
isolated ground, the neutral-to-earth voltage measurements were
compared in these two cases along the connection line from the south
side center (2, 0) to north side center (2, 10) of the test plot (referring to
sections 4.3.3 and 4.3.4). Table 17 shows this comparison (per unit
normalized) and Figure 59 is the corresponding plot.

From Table 17 and Figure 59, it can be seen that, generally
speaking, secondary isolated ground electrode connected with farm
ground system decreased the local neutral-to-earth voltage gradient
distribution originating from the primary ground connection. The most
significant decrease occurred at the point where the secondary isolated
ground rod was located. At this point (8 feet), the per unit voltage
decreased 72.2%, from 0.194 to 0.054. This decrease is due to the
opposite polarity of the voltage drop on the effective secondary ground
resistance R1 with respect to the voltage gradient that the secondary
ground coupled from the primary ground. This will be discussed in

more detail in next sections.
5.3.5 Verification of the Neutral Isolation Circuit Model

Table 18 and Table 19 present the test results for the circuit
model described in section 2.4 (referring to Figure 2), Chapter 2. The
method used to conduct this research is described in Section 4.3.5.
Refer to Figure 14 of that section to identify the quantities measured

175

Table 17. Data comparison between neutral-to-earth voltage gradient
distributions (per unit normalized) from secondary isolated ground
systems with and without farm ground connection along the direction
of primary and secondary isolated connection line

00190215014 m Fare_R OPEN 6 Far.) 5 on. 940615-3

 

Dist.-inlDist.-ft10ponPerU15 WIWItH 0'. Volt”

 

.267 .226 2.75 2.32

0 .0) .343 .331 3.54 3.4

3 .25 .357 .343 3.68 3.53

6 .50 .371 .359 3. 83 3. 69

9 .75 .41!) .385 4.12 3.96
12 1.00 .411 .399 4.24 4.1
13 1.08 .414 .402 4.27 4.13
14 1. 17 .415 .412 4.28 4.24
15 1.25 .429 .417 4.42 4.29
16 1.33 .418 .419 4.31 4.31
17 1.42 .431 .426 4.44 4.38
18 1.50 .468 .460 4.82 4.73
19 1.58 .483 .475 4.98 4.88
20 1.67 .490 .482 5.5 4.96
21 1.75 .498 .4” 5.13 5.04
22 1.83 .525 .519 5.41 5.34
23 1.92 .632 .613 6.52 .3
24 2.00 1.” 1.“ 10.31 10.28
25 2.08 .839 .825 8.65 8.48
26 2. 17 .593 .569 6. 11 5.85
27 2.25 .524 .5123 5.4 5. 19
28 2.33 .500 .481 5.15 4.94
29 2.42 .474 .458 4.89 4.71
30 2.50 .468 .452 4.83 4.65
31 2.58 .466 .451 4.8 4.64
32 2. 67 .459 .446 4.73 4.59
33 2.75 .453 .441 4.67 4.53
34 2.83 .420 .397 4.33 4.“
35 2.92 .413 .391 4.26 4.02
x 3.00 .405 .36 4. 18 3.96
39 3.25 .388 .39 4 3.79
42 3.50 .376 .356 3.88 3.66
45 3.75 .363 .343 3. 74 3.53
48 4.00 .349 .326 3.6 3.35
51 4.25 .335 .310 3. 45 3.19
54 4.50 .322 .296 3.32 3.04
57 4.75 .312 .284 3.22 2.92
60 5.N .303 .272 3. 12 2.8
63 5.25 .292 .258 3.01 2.65
66 5.50 .281 .247 2.9 2.54
69 5.75 .275 .237 2.84 2.44

6 CD

75 6.25 .259 .220 2.67 2.26
78 6.50 .249 . 197 2.57 2.03
81 6.75 .242 . 1“ 2.5 1.91

 

176

Table 17. (cont’d)

 

Dist.-inloist.-Ft:Wor-Ul5 Mlmltls 01. Volt.”

 

84 7.00 .236 .184 2.43 1.89
85 7.08 .232 .181 2.39 1.86
86 7.17 .223 .175 .3 1.8
87 7.25 .226 .161 2.33 1.65
88 7.33 .224 .166 2.31 1.71
89 7.42 .219 .162 2.26 1.67
90 7.50 .220 .157 2.27 1.61
91 7.58 .218 .138 2.25 1.42
92 7.67 .204 .146 .1 1.5
93 7.75 .212 .138 2.19 1.42
94 7.83 .209 .127 2.15 1.31
95 7.92 .203 .106 2.09 1.09
96 8.00 .194 .054 2 .55
97 8.08 .196 .091 2.02 .94
98 8.17 .196 .116 2.02 1.19
99 8.25 .189 .123 1.95 1.26
100 8.33 .195 .131 2.01 1.35
101 8.42 .197 .135 2.03 1.39
102 8.50 .193 .136 1.99 1.4
103 8.58 .195 .133 2.01 1.37
104 8.67 .192 .136 1.98 1.4
105 8.75 .194 .138 2 1.42
106 8.83 .192 .138 1.98 1.42
107 8.92 .191 .136 1.97 1.4
108 9.00 .190 .138 1.96 1.42
111 9.25 .179 .135 1.85 1.39
114 9.50 .182 .138 1.88 1.42
117 9.75 .180 .134 1.86 1.38
120 10.00 .174 .107 1.79 1.1

 

177

NEV GRADIENT SOUTH-NORTH DIRECTION
COMPARISON BETWEEN Form_R OPEN & Form_R 5 Ohms

 

 

 

 

 

1
a a d E? —9— OPEN
2 .
+ 5 OHM
U
Q)
.'_‘_‘
O
E '6 1 '
O n
z x“
E .4 - {I}; ‘
D E ~?.;.; : ‘
a - = = .1 ‘
D- - = T = : -
.2 -+ - 1‘-
: = : J
I
o , r , r
o 2 4 6 8 1o
DISTANCE FROM SOUTH (FEET) 940615-3
Figure 59- Comparison between neutral-tooearth voltage gradient distributions

(per unit normalized) from secondary isolated ground systems with
and without farm ground connection along the direction of primary
and secondary isolated connection line (length unit is foot)

178

 

 

 

< 386 > 36 > 36 > 2.6 > 8.... < 2&6 > 3.— o 3.... a
< 886 > 36 > 36 > and > v66 < E6 > an; .- n6~ h
(36 >96 >326 >36 >84». (2&6 >34 Chg o
(36 >86 > 86 > :6 >26 (2&6 >$J 2&3 m
(886 >36 >96 >36 >62: («36 >56 can v
(2.36 >96 >26 >36 >26— (and >36 and— n
(:36 >86 >34 >86 >86— <3n6 >nfln on.“ ~
< 86 > 86 > a; > 36 > 26. < n36 > and INA—O —
= mm .m am am .5 am .3
.350 MEX .a 0.8—.5 g .a .850 8.53
.5335 >mz >mz 9388 >mz 33.80 E I85
€2.88 .5... 6.388 a. has... has... has... 82.8 E... .8...
doaaaom .8“. em a .23 8... 9:55
E2583 2.0 was com 6.520 ban—ta 0:0 953 so... :ozflaaom .3502 .8 Sun— .3 03.5.

179

 

 

 

< 236 > 6N6 > 666 > 606 > ...n < 866 > 36 a 3.6 o.
< 9666 > .66 > 3.6 > 066 > 6.6 ( 366 > a a .6. n.
< 6686 > 2.6 > 86 > 6N6 > 2.... < 366 > a a ...MN 6.
< 366 > 86 > 66.. > an > 6.6 < 366 > N0.N ZNmO n.
< n8... > 666 > 86 > 36 > «.6. < :66 > an... a 36 N.
< .686 > 36 > 8.. > 86 > 666. < 866 > an a n6. ..
< 6.86 > 9.. > 3.. > 666 > 666. < N66 > an a ...Q. 6.
< 9:66 > 86 > 36 > an... > a6. < 266 > «Nd 29.0 6
_. ...m .m 8m «m. ..— 8 ...d
32.30 g .n nus—.5 ”.2"... .a .355 8'53
2.3.2... >mz >mz 5388 >mz 3.3.20 >mz 5......
$388 .5... 5388 9 55... has... has... 8.8... .5... =8.
62.2.2.3 .00.... Am a 5.3 new. 6.59.0
E3503 2.0 new 83. 6:220 .025... 03... as“: .3... 3:833 .8502 .o. 5.0 6— 26a...

180

and calculated. The 16 sets of tests described in 4.3.5 were performed.
The animal simulation resistors, R M and R M were not connected.
Tests 1 through 8 were connected with the one primary ground rod and
with primary neutral-to-earth voltage approximately 10 and 5 V,
respectively. These data are shown in Table 18. Tests 9 through 16
were connected with the two primary ground rods and with primary
neutral-to-earth voltage approximately 10 and 5 V, respectively. These
data are shown in Table 19. The resistance of the primary ground rod
RS was measured to be 18.6 ohms by the three-point fall of potential
method; the resistance of the two primary ground rods was measured to
be 11.4 ohms; and resistance of the isolated secondary ground rod was
measured to be 22.0 ohms. The quantities measured for this testing
were: primary source neutral-to-earth voltage (V); farm ground
resistance (ohm); .neutral-to-earth voltage at the transformer primary
ground (V); the neutral-to-earth voltage between the primary ground
and the isolated secondary ground (V); neutral-to-earth voltage at the
transformer isolated secondary ground (V); neutral-to-earth voltage at
the farm ground (V); neutral-to-earth current on the primary ground
(A); and the neutral-to-earth current on the isolated secondary grounds
(A).

The circuit model for the primary and secondary electrical system
with the neutrals separated is shown in Figure 14 of Section 4.35. If the
separation distance is reduced to the point where there is contact
between the primary and secondary ground rods for their entire length,
then R1 as a separate quantity disappears and the points between which

181

E51 are measured merge to become one point. In this situation, the
farm resistance Ri: is simply in parallel with the primary transformer
ground rod Rs- When the primary and secondary ground rods are
separated by a large distance then the only connection between the
ground rods is the earth. The primary circuit and the secondary circuit
will act independent of each other simply connected at one point which
is true earth. In this situation the secondary is no longer a part of the
primary circuit. Between these two extremes the Secondary ground rod
is electrically connected to the primary ground rod. This must be true
because the secondary ground rod is not far enough away from the
primary ground rod to be connected by the full value of the primary
ground rod to earth resistance. There is a primary to secondary ground
rod resistance which if Figure 14 is valid in Section 4.3.5, is made up of
R51 plus R1 which will be less than the primary plus secondary ground
rod resistance to earth RS plus R1. Also if this circuit model is correct,
the measured data from the field can be used to calculate other
measured quantities as well as quantities that cannot be measured.
These calculated quantities will be correct for all conditions if the circuit
model is correct and not just for special cases.

Table 20 provides calculated values of grounding electrode
resistances, circuit resistances, and some voltages. Figure 60 is the
same as Figure 14 of Section 4.3.5, except it is arranged vertically so that
all quantities can be visualized more easily.

The resistance to earth of the primary ground rod at the

transformer (Rs) and the farm system grounding resistance to earth

182

 

 

 

 

JAAA

 

 

 

 

Figure 60 The circuit model of Figure 14 of Section 4.3.5 is arranged vertically so
that all quantities can be visualized more easily.

 

 

 

 

 

..36. I I > ..6. > 8.... :36 on...” a 2...». I o.
.3... I I. >2 .2 >2 .. .2... 2.... .2... I n.
2...... I I > n. .2 > N. .n .8... 3... .2... I z
I 2.... .8... >2... > 2 .n I I .2... ..n: n.
.3... I I >2... >2... .2... .8. .2. I a.
.2... I I >8... >2 .2 .2. .2... .2... I ..
.8... I I >8... > 2.2 .3... .8. 3... I 2
I .2... 2.... > a... > 8.2. I I .2... .8... .
.2... I I > 8.... > 8.. .2... .8. .2... I .
.8... I I > 3.. > x... .n... .8. .2... I ..
.8... I I > 8.. > 8.. .92.. .1... 3...... I .
Mm I .8. .8... >2 .n >8. I I. .8. .8 2 a
1 .8... I I >2 .2 > ...2 3.... 3... .2... I v
.3: I I > ...2 >2 .2 .8... 2.... .2... I a
.8... I I > 8.2 > ...2 .3... 3... .2... I N
I 3.... .8... > n. .2 > n. .2 I I 3... .8... .

2 5. 5. m. ..m + an. .2 2.. 5. a.

.2"... .. 83...... 8.3.1... 8.5.1...
>mz 3.5.... 13...... al.-1... 3. 1.86
9...... E... has... 8....» bl... ...p

 

.2 0...... .3. .2 .3... a.

...o 82. 825...... .32.... 2520 .5 82.82 .2520 .o 82.5 858.6 .8 .1..—

184

(RF) is important for accurate calculations. These resistances to earth
were measured using the three point fall of potential method which is
reasonably accurate but is subject to some error. The accuracy of these
measurements can be checked by using Ohm’s Law to calculate the
resistance when an accurate measurement is made of neutral-to-earth
voltage (Es and Br) and grounding electrode current (IT and 11). When
the circuit to the farm grounding resistance is open as shown in Figure
61 all of the primary ground current (Ir) will flow through the primary
grounding electrode (RS), therefore, the resistance to earth will be Es
divided by the current IT. For the cases where there is current flowing
to the farm grounding electrodes the farm grounding system resistance
to earth will be the farm neutral-to-earth voltage (BF) divided by the
farm grounding system current (11)- The method of calculating the

various other quantities was performed as follows:

1. With the farm grounding open as shown in Figure 61, take the
measurements of Bs (primary neutral-to-earth voltage), B51 (the
voltage between separated primary-secondary neutral), E1
(neutral-to—earth voltage of the secondary isolated ground) and
Rs (the primary ground rod resistance, RS = R31 + R82), and
then use the voltage division law to calculate the resistance R81
and R52 in the circuit model (the results are shown in Table 16c
for tests 1, 5, 9 and 13):

R51 = Rs x (E51 /1~:S) _ (2.1)

 

 

185

 

 

-
H
q
*—
‘AAA
VVT
”
M
—

["0

 

AAAAA

 

 

(p—f‘ -—-———UI¢—[."—.

 

Figure 61

The circuit model of Figure 14 of Section 4.3.5 without farm ground
system connection is arranged vertically so that all quantities can be
visualized more easily.

 

 

 

186

R52 = Rs X (El / Es) (2.2)

For Test 1, RS = 18.29 ohm, E8 = 10.15 V,
E8] = 8.19 V, and E1 = 1.96 V, hence:

 

Rsl = 18.29 x (8.19/10.15) = 14.76 (ohm)
R52 = 18.29 x (1.96/10.15) = 3.53 (ohm)

For Test 9, R8 = 11.22 ohm, E8 = 10.23 V,
E3] = 7.28 V, and E1 = 2.94 V, hence:

Rsl = 11.22 x (7.28/10.23) = 7.98 (ohm)
R52 = 11.22 x (2.94/10.23) = 3.22 (ohm)

According to Kirchhoff’s Voltage Law (KVL), with the farm
grounding closed as shown in Figure 60:

 

E8 = ITXRSI +leR1-I-E1 (2.3)
and R1 = (Es - E1 - I.r x R51) / II (2.4)

Where 1,. and 11 are primary and secondary neutral-to-earth

currents, respectively; and R1 is effective secondary isolated

ground resistance.

Take Test 2 as the example, ES = 10.06 V, E1 = 1.09 V,

187

I, = 0561 A, R51 = 14.76 ohm, and I] = 0.0414 A, from equation
(2.4):

R1 = (10.06 -109 - 0.561 x 14.76)/0.0414 = 16.66 (ohm)

The effective secondary ground rod resistances R1 are shown in
Table 20. Other values in Table 20 were calculated from the measured
values of Table 18 and Table 19. The source resistance (RD) was
determined by dividing the source NEV (ED) by the primary grounding
current (11.). The secondary neutral resistance (RN) was determined by
dividing the voltage drop on the conductor (E1 - BF) by the secondary
grounding current (11).

From Figure 60 it can be seen that the primary neutral-to-earth
voltage (Es) is equal to the sum of the primary to secondary neutral
voltage (E51) and the secondary neutral-to-earth voltage (E1). Keep in
mind that these are all measured values from field tests. Note in Table
20 that E51 plus 131 is very close to the measured value of E.

It is also important to note that the resistance to earth of the
secondary ground rod at the transformer was 22 ohm. The mean value
calculated for the effective secondary ground rod resistance (R1) was
18.69 ohm. Assuming that the resistance to earth value of 22 ohm is
accurate, the effective resistance of the secondary ground rod in these
tests was 85 percent of the ground rod resistance to earth. Althouse
(1990) conducted similar tests but with the primary and secondary
ground rods spaced only six inches apart and with a primary ground rod

188

resistance of 18.9 ohm and secondary ground rod resistance of 21.4
ohm. These are nearly the same ground rod to earth resistances used in
this research. The Althouse (1990) data is shown in Table 21 tests 1
through 12. Values for the effective secondary ground rod resistance
(RI) for the Althouse (1990) data are shown in Table 22. The mean
value of the effective secondary ground rod resistance was 11.4 ohm.
Compared with the resistance to earth of the secondary ground rod of
21.4 ohm, the effective resistance with only a six inch separation is 53
percent. The closer the spacing of the secondary ground rod to the
primary ground rod the smaller will be the effective secondary ground

rod resistance.

5.3.6. Parameters Influencing Effectiveness of Neutral Separation

In order to determine the parameters that influence the

effectiveness of separation of primary and secondary neutrals at a
distribution transformer, it is necessary to develop an accurate circuit
model for the primary and secondary neutral and grounding systems.
The circuit model proposed by Althouse (1990) was subjected to field
testing and the model was found to be an accurate representation of the
data under all field conditions. That circuit model is reproduced here as

Figure 62 in a slightly altered form to aid in examination of some of the

circuit parameters.
The previous analysis of the data collected in this research along

with some data collected by Althouse (1990) indicate‘that the values of

 

189

 

 

 

< ...... ...... > .... ...... > ... < ...... I . .. 2
< ...... ...... > .... ..... > .... < .... I... . ... ..
< ...... ..... > ...... I. > ... < .... I . .. 2
< .8... ...... > .... .. > .2 < ...... I . .. ..
< .8... ...... > a... ..... > ... < ...... I . .. ..
< ...... ...... > 8.. I > 3.. < ...... I . .. ..
< ...... ...... > .... -.. > 5. < ...... I . ... ..
< ...... ..... > 8.. > .... > .... < ...... I . ... ..
< ...... ...... > .... > .... > .... < .... I. . .. ..

< .8... I > .... > .... > .... < ...... I . .. 2

< ...... ...... > .... > ... > .... < 8... I . .. .

< ...... I > ...... > .... > .... < 8... I . .. .

< 8.... -- > .... > .... > ...... < .... I . .... ..

< ...... ...... > .... > .... > .... < ...... I . .... .

< ...... ...... > .... > 8.. > .3 < ..... I . ... .

< 2.8.. ..... > .... > .... > .2. < .... I . .... .

< 88.. ...... > 2.. > .... > .3. < ...... . I . .... .

< ..8... .- > .... > 8.. > 8.. < .... I . .... .

< 88.. ...... > .... > .... > .... < 8... I 29.6 .

_. .m .m ...... m. a. am ....

322.0 g .. on...o> g .. 322.0 8.53
......an >mz >mz e886. >mz ......an >mz In...
88.8.. a... ...-.38.. a 9a.... .53.... .36.... 8.... I... a...

2...... 85...... ...... .. 2. .. 5.2... ... .2 5.2... .... .6222... .8".
.... .... ... 5.2... .. 8.3... .2. ... . 5.3 ...... 8.20 ......80. ..o ....
com 3.5.0 38....— ueo 9...: 58: 0.3052 .0 30,—. cough...“ .3282 .... ...-D ..N 03:.

190

 

 

> 2. .. .... .. I I 2
... > .... -.. a .... I I I ..
-.. > ... ... a .... I I I ..
I > .... ..... a 8. ... I I 2
.. > ... ..... .. .... x I I 2
-.. > .... ..... .. a... .. I I ..
.... > .... .. ... .. I I ..
. .... > ... > ... a F. .. I I ..
...... I ,>.... >8. ...... I I I ..
.. .... I > .... > .... a 2... I. I I 2
.. .... .. > ...... > .... a 2.. .. I I o
a .... ...- > .... > .... a 2.. I I .
.. .... > .... > .2. .. .... ..... I I ..
.. .... .... > 2... > 2.... a .2 .. I I o
.. .... I > .3. > .... a on. .. I I. .
a ..2 > .... > .... . ... I .. .... .
.. .2 I > .... > .... .. ... I .. .... .
.. ... a 8.. > 8. > .2 a .2. ..... I a 2.. .
.. ... . .... > .... > ...... ... I a 2.. .

... .3. 5. .... .... + ...... .... z. 9. ....

5...... .. 8.5.8.. 8.528.. 815......
>mz .3335 .532 2138.. .5. 385
baa... 3...". 588 82.8 in .8...

 

... 22.9 5 a... 8.2. 3.552

80.. 82.223”. 23.6 .2590 ...... 80..qu 92.2.3.0 .o .o:_a> “.0..—3.6

Na 031,—.

191

 

 

 

 

 

 

 

 

Figure 62. Slightly altered form of the neutral separation circuit model
(Althouse,1990)

192

resistances (RD, R51, R52, R1, and RN) in the circuit model are nearly
constant for a particular location and installation. The data show that
the level of voltage and current flow do not result in a significant change
in these resistances. The farm grounding resistance in the circuit model
of Figure 62 has been separated from the circuit so that it can be
studied separately with different values.

All voltages in the circuit model are variable as the value of farm
grounding resistance (RF) changes. This is evident by examining the
voltages in Table 18 and Table 19. However, the open circuit voltage
(VOC) and the neutral-to-earth voltage at the primary ground rod (Es)
were held as constant as practical. But the voltage across the
components of the primary ground rod resistance (Rs) which were R51
and R52 were not necessarily constant. Note from Table 18 and Table
19 that the total current flowing in the primary grounding circuit (11')
increased only slightly as the farm grounding resistance (RF) was
decreased. Also note from Table 18 and Table 19 that as the farm
grounding resistance increased, the secondary grounding current
increased. But note that the increase in secondary grounding current
(11) was much greater than the primary grounding current (IT). The
reason is evident in the circuit of Figure 62. The portion of the primary
ground rod resistance R32 is in parallel with the secondary neutral and
grounding circuit. As the farm grounding resistance is decreased, 3
larger proportion of the current from the primary takes the farm
grounding circuit path rather than completing the circuit to earth
through R52.

193

As was seen in the previous section, and as shown in Table 18,
Table 19, and Table 20, the field test data can be used to determine the
values of all of the resistances of the circuits. If the open circuit voltage
(VOC) is known, the primary side and secondary side neutral-to-earth
voltages can be determined for any value of farm grounding system
resistance to earth using simple circuit solving techniques.

The voltage E52 of Figure 62 is the value of the earth gradient
voltage as measured to a reference ground at any distance from the
primary ground rod. In this research the secondary ground rod was
located a distance of six feet from the primary ground rod. The
resistance Rs2 will decrease and the resistance R51 will increase in value
as the secondary ground rod distance from the primary ground rod is
increased. The converse is true as the distance between these two
ground rods in decreased. Note that for the Althouse (1990) data for
ground rods located only six inches apart that the value of R32 in Table
22 becomes larger. Also note that a greater portion of the primary
neutral-to-earth voltage (Es) appears at the secondary (E1). Clearly the
distance of separation is one parameter involved in the effectiveness of
neutral separation.

In this research the ratio of farm neutral-to-earth voltage (BF) to
primary ground rod neutral-to-earth voltage (Es) was plotted against
the value of farm grounding system resistance to earth (RF). Figure 63
is a plot of the data from Table 18 and Table 19 for two levels of
primary grounding resistance. In Figure 63 each point is the average of

two tests, one at approximately 10 volts on the primary ground rod and

194

   

 

 

25 .4
U
.3.
..‘1 20 —
M
U
m
g: 15 2
M
5 z:
22 1:10 -
§ .. Each nork Is the mrm of no punts
g 5 _ 0 Primary proud 10.29 chm
g I Prlnnry grand 11.25 m
L. o I

0.00 0.05 0310 OTIS 0’50

r,/E,

Figure 63. The ratio of farm neutral-to-earth voltage (Er) to primary ground rod
neutral-to-earth voltage (E8) was plotted against the value of farm
grounding system resistance to earth (RF). The data was from Table
18 and Table 19 for two levels of primary grounding resistance.

195

the other 5 volts. ‘The two values were nearly identical indicating that
voltage level does not influence the parameters in this circuit model.
Note in Figure 63 that a slightly higher percentage of the primary
neutral-to-earth voltage appears at the farm neutral as the primary
grounding resistance to earth is decreased. The data for Figure 63 and
later figures is shown in Table 23.

As was discussed earlier, field data can be used to determine all of
the circuit model resistances. If the open circuit voltage is known, then
the ratio of farm neutral-to-earth voltage to primary ground rod
neutral-to-earth voltage can be determined for any desired value of
farm ground system resistance to earth. That data is plotted in Figure
64. Note that as the farm ground system resistance to earth approaches
zero the ratio Br to Es approaches zero. Clearly a low farm grounding
resistance to earth is extremely important for effective neutral
separation. Note that as the farm ground system resistance to earth
becomes large, the ratio BF to Es approaches a maximum value.
Actually the voltages BF and E1 approach the voltage of the earth
gradient E32. Note in the circuit model of Figure 62 that the effective
secondary ground rod resistance (R1) and the secondary neutral
resistance (RN) are in series with the farm ground system resistance to
earth. As the farm ground system resistance to earth is decreased and
the current in the secondary ground circuit increases the voltage drop
across R1 and RN increase thus leaving a smaller portion of the earth

gradient voltage (E52) to appear at the farm building neutral (BF).

196

 

 

a...” 26 3.3 a 0.2 .3 o a land?
a R .3 3.: a 2: .... o a. 88.3.
a 8. 8.8 3.; a a: .3 o o 1.2.2
a SN 85 3.: a a... .... o n .832
:3 was a a on... f. 2.2 a
and. 83 a a an... .5 2.: .J
2mm :3 a a on... .3 2.2 .J
298 :3 a 2 one: ... o as 3
3.. 33 a «N .5... .5 ... 3
3.2 .38 a 2 «a... do 2 ...
3.8 so... a a 3N... ... a a." 3
220 83 a 2 32: do a. 3
5. new a
£3.20 3. 385 3. 356 i
E... has}, al.... 13. 336 ....F

 

.5233 3.5882? 3.83% 05

.a 8.8:; 5.3.9-3332 com 9520 bag...— o. .0352“ we
0.35— 93 $3381 3262mm 9.6.590 $5225 tom 939.0

Figure 64.

197

 

 

75" ’
1
a '4
g '1
'6; .4
E
a: 50« o
t .
U
.-
9‘
)‘,_ "
V,
3 +
is
3" 1
as; 251 p
O _‘
g 20+ ..t
h- 15-1 .5.
toq -'
U
54 ....
o 0'
0.00 0105 0‘40 oils oTzo
r,/E,

The ratio of farm neutral-to-earth voltage (BF) to primary ground rod
neutral-to-earth voltage (Es) was plotted against the value of farm
grounding system resistance to earth (RF). The data was calculated
from the circuit model in Figure 62.

198

The measured data for the single primary ground rod (Tests 1
through 8) are plotted in Figure 65 as the ratio 13F to Es and farm
grounding system resistance. The circuit model is used to calculate the
ratio corresponding to other values of farm grounding resistance. Note
that the calculated values fit the measured data. Data was plotted from
the data of Althouse (1990) for a primary and secondary ground rod of
nearly identical resistances to earth, but the separation distance
between the ground rods was six inches. The same type of curve is
produced by the data but because of the close spacing of the ground
rods in low resistance earth, the ratio BF to Es slightly exceeds 0.5. But
also note that as the farm grounding system resistance to earth is
decreased, the ratio decreases rapidly approaching zero as the
resistance RF becomes small.

From the data of this research and the data of Althouse (1990)
and the verification of the circuit model, it is evident that there are
several parameters which are important in determining the effectiveness

of neutral separation. These factors are:

1. The resistance to earth of the farm grounding system as seen from

the transformer pole.

2. The spacing between the primary and secondary ground rods at

the transformer pole.

FARR BRDUNDING SYSIEH RESISTANCE

 

Figure 65.

199

Calculated value I) ------
Insured value o———A

:
§ 5' g
3 a
t; 5 _.'
.- 5 ..
f2 5 i
i :-
: . 3
3 5 g
°' :
.3
.a'
.n'

 

0105 0.10 0115 0120 0T2: 0330 alas 0.100 0145 030
50":

The ratio of farm neutral-to-earth voltage (BF) to primary ground rod
neutral-to-earth voltage (Es) was plotted against the value of farm
grounding system resistance to earth (RF) for 6 inch and 6 foot
separation distance. The calculated data was from the circuit model in
Figure 62.

200

3. The resistance to earth of the primary grounding electrode when
the secondary grounding electrode resistance to earth is maintained at a

constant value.

4. The resistivity of the earth in the area of the primary and

secondary ground rods as it determines their resistance to earth.

It could not be determined from the data collected what influence the
resistance to earth of the secondary ground rod would have when the
primary ground rod resistance was held constant. Clearly, the most
important of the parameters is the resistance to earth of the farm

grounding system.

VI. CONCLUSIONS

The computer simulation models were developed and used to
study the neutral-to-earth voltage profile along both single-phase and
three-phase radial distribution line systems with different normal and
abnormal operating conditions. The farm field tests were conducted to
reveal the neutral-to-earth voltage gradient distribution near the ground
electrode systems on the primary and secondary sides of the neutral
isolated distribution transformer. An equivalent circuit of the isolated
primary and secondary ground systems was proposed to study the
parameters that influence the effectiveness of neutral separation. The

conclusions "drawn from this study are:

1. The single-phase DC circuit model developed in this research is
sufficiently valid to predict the changing trends of the
neutral-to-earth voltage profile along a single-phase primary
distribution line. The single-phase AC distribution line model
developed with line inductance and an equivalent transformer
Circuit replacing the load resistor in the DC model produced a
neutral-to-earth voltage profile which was slightly higher in

magnitude and similar in shape to the DC profile.

201

202

The overall neutral-to-earth voltages were substantially lower
along the electric power distribution system of the three-phase

four-wire Yo/Yo connection than that of the single-phase.

The effect of high load demand on neutral-to-earth voltage is
equivalent to having a phase line to neutral fault occurred near
the high demand location and will result in a local increase in the

neutral-to-earth voltage.

Neutral-to-earth voltage change caused by a high resistance
segment in the neutral conductor (bad neutral connection) of a
primary distribution line will be greatest in the local area of the
high resistance segment and will decrease in magnitude as the
distance increases away from the high resistance segment. The
bad neutral connection can be detected by the measurement of
the local primary ground system impedance which presents the

significantly higher value than the normal value.

A high substation resistance-to-earth for a single-phase radial
distribution line will result in an increase in_ neutral-to-earth
voltage along the line near the substation when load current is

carried on the primary neutral.

At any grounding electrode location of the distribution line, the

system ground Thevenin equivalent open circuit voltage was

203

divided in series by the ground resistance and the Thevenin
equivalent resistance (system ground resistance) at this location.
Generally speaking, a lowered ground resistance at a particular
location along the line results in a lowered local neutral-to-earth
voltage. The actual level of neutral-to-earth voltage is determined
by the open voltage and the proportional relation between the
local ground resistance value and the system Thevenin equivalent

resistance value at this location.

Increasing the number of grounding electrodes can decrease the
neutral-to-earth voltage to some extent. A large number of
ground rods could be costly and the result may not be as good as

expected.

Increasing the primary distribution line operating voltage resulted
in a decrease in primary line current and thus the neutral-to-earth
voltage along the line. The reduction of neutral-to-earth voltage
was less significant when the voltage increased above 7 .2 kV than

when the voltage was increase from a lower value up to 7 .2 kV.

A secondary ground fault may cause a significant increase or
decrease in the neutral-to-earth voltage along the distribution line
in the local area of the ground fault depending upon whether the

secondary ground fault current is in-phase or out-of-phase with

10.

11.

12.

13.

204

the primary current. The change in neutral-to-earth voltage is

location dependent along the distribution line.

When a primary phase-to-earth fault occurs anywhere along the
single-phase distribution line, the neutral-to-earth voltages will
increase in line segments near the substation and decrease for line
segments from the middle to the end of the line with the most

significant changes near the substation.

When a primary phase-to-earth fault occurs anywhere along the
distribution line system of a three-phase four-wire Yo/Yo circuit,
the neutral-to-earth voltages will increase in all line segments of

the system with the most significant increases near the substation.

For the earth voltage gradient resulting from a single primary
ground rod electrode, the test value and the theoretical value
calculated from the equation (5.3) were very close within the first
20 inch (051 m) from the ground rod. After this point, the
discrepancy increased as the distance from the center of the rod

increased.

The earth voltage gradient resulting from a single primary ground
rod electrode will be altered with one more additional primary
ground rod 6 feet (1.83 m) away. The two voltage peaks near each
rod will be present. The second peak created by the additional

14.

15.

205

primary ground rod not only elevated the neutral-toearth voltage
significantly along the direction of the straight connection line of
two rods, but also elevated neutral-to-earth voltage along all other
directions to some extent. This elevation diminished as the

distance increased away from the both rods.

The secondary isolated ground electrode connected with the farm
ground system will decrease the local neutral-to-earth voltage
gradient distribution originating from the primary ground. The
most significant decrease occurred at the point where the
secondary isolated ground rod was located. This decrease is due
to the opposite polarity of the voltage drop on effective secondary
ground resistance R1 with respect to the voltage gradient that
secondary ground coupled from the primary ground.

Effectiveness of separation of primary and secondary neutrals is a
function of resistance to earth of the primary grounding electrode,
separation distance between the primary and secondary grounding
electrodes, and the resistance to earth of the farm grounding
system. The most important parameter is the resistance to earth
of the farm grounding system. A low farm grounding system

resistance to earth insures effective neutral separation.

APPENDIX

206

AC CASE, RD9 CHANCE IMPACT ON NEV, 4.8KV

R09 0.086, 1, 5, 20 AND lOOK OHM, LOAD .3 A
15 —+— 0.086 OHM

 

 

  
 

 

 

 

—B-— 1.00 OHM
+5.000HM
+20.00HM
+ 100K OHM

1.0:"
A
>
v
LJJ
(D
<(
*—
_J
O
>
5-1
0 trtITrtTHHIIHIIrrrTWIItIrtrthttrrrtflntrtrttlnn
'- LO Q in 0 Ln 0 In 0 n O IDIN
'- v- N N f’) n ‘3' V’ In LOU)

NODE # (1—57) 940528c

Figure A1. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 5 7th node with variable neutral conductor
resistance RD9 between node 9 and node 10 and a normal customer
load of 3A at node 9 (single-phase AC model simulation)

207

AC CASE, P1333 CHANCE IMPACT ON NEV, 4810/

VOLTAGE (v)

 

 

RD33 0.086, 1, 5, 20 AND 1OOK OHM, LOAD :5 A

_B_

+

_q_

 

 

O 111111111llllTTfiT—TTITITTITIIIIITIITIIITTTITITIIIIIIIIT
v- LO 0 LO Q in O Ln 0 LO Q lDl\
'— v— N N I") 1”) V V L0 1.011)

Figure A2.

NODE # (1—57) 9405260

0.086 OHM
1.00 OHM
5.00 OHM
20.0 OHM
100K OHM

Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with variable neutral conductor
resistance RD33 between node 33 and node 34 and a normal customer

load of 3A at node 33 (single-phase AC model simulation)

208

PDSB CHANGE IMPACT ON NEV, 4.8K 1-PHASE AC

10

  

 

AC MODEL, RD56 0.086, 1, 5, 20 AND 100K OHM, LOAD 3 A

0.086 OHM

 

 

 

—a— 1.00 OHM
—x— 5.00 OHM
8 ‘ 4— 20.0 OHM
—-¢—— 100K OHM
A
a 6 ~
to .....
o
<
[.—
..l
O 4 ‘
>
2-1
0 TIIITTWTITTIIITIIlIITIIITIIIIITIIIIIIIIITIITIFTTIIIIII
-- in 0 tn 0 In 0 l!) O In 0 In I\
v- — N N F) F) V’ V' In In In
NODE # (1 -57) 940530
Figure A3. Prole of the neutral-to-earth voltage along the primary distn'bution

line from substation to 57th node with variable neutral conductor
resistance RD56 between node 56 and node 57 and a normal customer
load of 3A at node 57 (single-phase AC model simulation)

209

HEAVY LOAD IMPACT ON NEV, 4.8K 1—PHASE [LINE

AC MODEL, LOAD 3, 6, 9, 12 8c 15 A AT 404-9
15 —+- 3 A

—B-—6A
+9A
+12A

+15A

O

  

VOLTAGE (v)

U1
1

 

 

 

O TIITIITIIIIIITITIIITITIIIIIIITTIIITjIIITTIIIIITITIIIlrl

v— I!) O In 0 ID 0 I05

NODE # (1 —57) 940529f

Figure A4. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with variable load demand
conditions at node 404-9 (single-phase AC model simulation)

210

HEAVY LOAD IMPACT ON NEV, 4.8K 1—PHASE LINE
AC MODEL, LOAD 3, 6, 9, 12 8c 15 A AT 416—33

15 —-t—3A

 

+611
I ——-—9A
12A
15A

VOLTAGE (v)

 

 

 

 

o IlllllllllllllllIlflrrlllllllllllllllllllIIllllllllIlTT
'- 10 0 l0 0 In 0 In C) In C ION
'- — N N I") I") ‘1' V ID no

NODE # (1 -57) 940529e

Figure A5. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with variable load demand
conditions at node 416-33 (single-phase AC model simulation)

211

HEAVY LOAD IMPACT ON NEV, 4.8K 1—PHASE LINE

AC MODEL, LOAD 3, 6. 9, 12 80 15 A AT 428-57

 

 

 

 

15 + 3 A
—B— 6 A
i + 9 A
1 1
ix + 12A
d” r + 15A
101 ..
A d
> n
V . K
DJ 00 .
3 ‘1
+_ n "=:2|
_1 .. '5‘-
O {‘1‘ ‘43" '==::::‘
> ‘. .:,:Yéygg::=‘
5" 0“ -'.+.L
t‘x‘
VԤb
3' ,
\‘.\¢ lazfl'
§é¥§¥w «'1' f
O IIIIIIIIIIIIIIIITTTIIIITTIIIIIIIIIIFIIIIIIIITTIUIITrI—rf
'- 1n 0 in O 10 O In 0 In C) 10 l\
'- "' N N f') l’) V“ V l0 In In
NODE # (1—57) 940529d
Figure A6. Profile of the neutral-to-earth voltage along the ‘primary distribution

line from substation to 57th node with variable load

demand

conditions at node 428-57 (single-phase AC model simulation)

212

RG1 CHANGE IMPACT ON NEV, 4.8KV LINE AC CS
SUBSTATION R01 0.5, 1, 5, 1O, 20 a: 40 OHM, AC CASE

 

 

 

 

 

30 —+-—O.50HM
I —B—1.00HM
\\ +5.00HM
‘1‘ +100HM
‘-,' +200HM
Azod \‘3 +4OOHM
> ‘\
5 V
<151: ',
t: .. ‘i
O u \.
> A
10: “ \
u ‘
fl ‘.
v 1‘
5« ~
0 IIIflIIIIIIIIIIIIIIIllIITTIIIIIIIIIIIITTIIIIITIIIITTII
'- In 0 In 0 1n O 10 O n O 0.05

e— v- N N f’) f’) V V 10 Inln

NODE # (1 -57) 9405311:

Figure A7. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. variable substation
resistance RS (single-phase AC model simulation)

213

ROSS CHANGE IMPACT ON NEV, 4.8KV LINE AC CS

GROUND R633 1.5, 1.2, 0.9, 0.6 8: 0.3 OHM, AC CASE

 

10 ¢
_.8._
+
a4 ...-...
+
/'\
a 6~
LIJ
O
<
’—
_J
Q In
>
2..

 

 

 

 

O IllI]III111ITITIITTIIIIIITITIII]lllTTlllIlllTIlTlTlllrl

.— In 0 1n 0 In C 10 O n O ION
v- v— N N f") f’) V’ V Ln mm

NODE # (1 -57) 940530—2

1.5 OHM
1.2 OHM
0.9 OHM
0.6 OHM
0.3 OHM

Figure A8. Pro-e of the neutral-to-carth voltage along the primary distribution
line from substation to 57th node with base case vs. ground resistance

RFGBB change at node 33 (single-phase AC model simulation)

214

INCREASE GROUND ROD NUMBER IMPACT ON NEV
ALI. GROUND RESISTANCES 100K CHANGED TO 25 OHMS

 

 

 

 

 

 

10 -'+'“' BASE
C —B— AD__RD
n
8'I
u
A
> 6.. u
LLJ
O u
<
P— --'=
—l 3: ,.,:::: ---- Linc-.1
O 4‘ 5::5“
> '5'.
U 5’4"
7.“
'4
2" 46'
o"‘
4'1
O llllrlrlTITIlUTIllelllIlITTIIrTIlljllllTlTIjlIII][III
'- tn 0 In D n O in O 10 O IDIN
— — N N n n v 1- 10 1am

NODE # (1 —57) 940530-5

Figure A9. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. ground rod

number increasing along the entire system (single-phase AC model
simulation)

215

SEASONAL GROUND R CHANGE IMPACT ON NEV
ALL GROUND RESISTANCES INCREASE TO 2, 3 8c 4 TIMES

 

 

 

 

 

25’ * BASE
-—B— 200%
+3007.

20 +4007;

A
2,15.
LIJ
(D
<
'3
0‘01 ----- I
> --==========“'"‘
’_,"‘=‘
5" 35"
,’)-
0 IIIIIIITIIrIrIITWTFrITIITrTrIrllrrllIIIIIIITIIIIIIIT
P In 0 In C In C In C In 0 Inl\
'- v- N N n I") V V In InIn

NDDE # (1-57) 940530-4

Figure A10. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. seasonal variations

of ground rod resistances all along the entire system (single-phase AC
model simulation)

216

AC MODEL, OUT—OF—PHASE SECONDARY TO EARTH
IMPACT ON NEV, FAULT 3, 6, 9 8c 12 AT 416

 

 

 

 

 

 

10 + BASE
-B—3A
+6A

34 +9A
+ 12A

A

Z. 5«

LrJ

2

...

.J

O 41

>
2.1
O TTTYITTTTITIIIIIIIIIFTTITIITIIIIIIIIIIIIIIIIITIIIIIII
F In 0 In 0 In 0 In C In 0

101x
'- v- N N n n V V In Inln

NDDE # (1—57) 940603-4

Figure All. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. variable secondary
ground fault attached to node 33 (out-of-phase, single-phase AC
model simulation)

217

AC MODEL, IN-PHASE SECONDARY TO EARTH FAUL
IMPACT ON NEv, FAULT 3, 5, 9 8c 12 AT 416

 

 

 

 

 

10 + BASE
- —-a—3A
8‘ +911.
+12A
A
a 6«
LJJ
O
<
l.—
‘J \
O 4*
> .
\
1\
\.
2‘ A
\‘iz
“4‘
‘1
O TTIIITIIHIIIIlIIIIWT—rllITIITIITTIIIIIIIITTIIIITIIIII
— In 0 In 0 In 0 In 0 In 0 on
.- — N N n n v ¢ In InIn

NDDE # (1-57) 940603-3

Figure A12. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. variable secondary
ground fault attached to node 33 (in-phase, single-phase AC model
simulation)

218

AC MODEL, OUT-OF-PHASE SECONDARY TO EARTI—
IMPACT DN NEV, FAULT 3, 6, 9 8c 12 AT 404

 

 

 

 

 

10
I
II
I\
I
A
>
v
Lu
0 .....
<
...
..l
O —*—-BASE
>
+3A
+6A
+9A
+12A
O TITITFFTIIIFTIWI—FTIIIIIIllllIIIIIIITIIIITIIIIIIIIIIITIT
— In o In 0 n o In 0 In 0 Inl'\
.— v- N N n n 1' v n Inln

NDDE # (1 -57) 940603-2

Figure A13. Profile of the neutral-to-earth voltage along the primary distrrbution
line from substation to 5 7th node with base case vs. variable secondary
ground fault attached to node 9 (out-of-phase, single-phase AC model ~
simulation)

219

AC MODEL, IN—PHASE SECONDARY TO EARTH FAUL
IMPACT ON NEv, FAULT 3, 6, 9 8c 12 AT 404

 

 

 

 

 

15 —+—BASE
—B—3A
—*—-SA
+9A
1
1 +12A
}
10:
A y
>
v
LAJ
O
<
I'—
_I
O
>
5-I
0 TIVTIIIIIWITIITIIIIIIWIIIFITIITTIITIITIFTITIIIIIIIIIT
.- In 0 In 0 In 0 In o In 0 Inrx
.- .— N N n n v:- «r In Inn

NODE # (1 —57) 940603-1

Figure A14. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. variable secondary

ground fault attached to node 9 (in-phase, single-phase AC model -
simulation)

220

AC MODEL, OUT-OF—PHASE SECONDARY TO EARTH
IMPACT ON NEV, FAULT 3, 6, 9 8c 12 AT 428

 

 

 

 

15 P BASE
.8— 3A
+6A
1t

# +9A
#
# 7+12A

10" f .

l \ I '0
> v
v I
LLJ
O
< .:3
+— " ’55:.u
S “(33!“ -:=::=T
> oi:,«.:Y:-Yé:;:==:
5‘ T'}:j’on"")*
0 IITITTTTITTTTIFIIIIITIIITTITITIIITTIIIrTTIIITTIIITTFIiI
'— In C In C n O In!"
‘D 9 9 8 m n n <- «- In InIn

NDDE # (1—57) 940603—6

Figure A15. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with base case vs. variable secondary

ground fault attached to node 57 (out-of-phase, single-phase AC
model simulation)

221

AC MODEL, IN—PHASE SECONDARY TO EARTH FAULT
IMPACT ON NEV, FAULT 3, 6, 9 8c 12 AT 428

 

  
  

 

 

 

10 —‘I—' BASE
+3A
8T +9A
+12A
9
V 6"
LIJ
O
1‘5 ..
-’ .. :{e-Tz'g'fé ‘ ----------
O 4 ‘ ,_.,-.'é;_:g=:;;'.'€n’."- -
> 7. ,.,
21
O TlllllllrlllllrlllllTlllIITIlIrTITIIITTIIIIIWIIIIIII

NDDE # (1 -57) 940503-5

Figure A16. Profile of the neutral-tocearth voltage along the' primary distribution
line from substation to 57th node with base case vs. variable secondary
ground fault attached to node 57 (in-phase, single-phase AC model
simulation)

222

PRIIxIARY—GROUND EAULT IMPACT ON NEV, 4.8KV

AC MODEL, FAULT 3, 6, 9 8c 12 A AT 416-O

 

 

 

 

 

NDDE # (1-57) 94053Dc

—-+——BASE
-—8—-:SA
+6A
+12A
10
A
>
V
LLJ
O
E n
_J .
O u‘
> \.
5“ “.
A
11‘
3‘33. .,
\Q‘L
ak‘g La};
‘12“ it"
-. 4'4.
8“." '4;
52'“
O lTTlTlllrllTFlllTllTTTllllllllllllllllTII—ITIIITIIIIII
'— In C) In C In 0 In C In 0 Inl\
'— v— N N n I”) V V In InIn

Figure A17. Profile of the neutral-to-earth voltage along the “primary distribution

line from substation to 57th node with an ungrounded phase conductor
to earth fault at node 416 compared with the normal line (single-phase
AC model simulation)

223

PRIMARY—GROUND EAULT IMPACT ON NEV, 4.811in

5 AC MODEL, FAULT 12 A AT 404—O, 416-O 8c 428-O
1 ,

 

 

 

 

 

+BASE
—a—404
+416
+428
10"
A .4
>
v
LLJ
O
<(
l.—
__J
O
>
5.1
dLV‘T-l.
O r1111lI[IIITTIIIITIIrI[ITIITTITTTrlrrIIIIII]llllllerT
'— In 0 In 0 In C In C In O on

-— -— or N n n v- v- In 1010

NDDE # (1-57) 94053od

Figure A18. Profile of the neutral-to-earth voltage along the primary distribution
line from substation to 57th node with 12 A phase conductor to earth
fault at node 404, 416, or 428 compared with normal line (single-phase
AC model simulation)

REFERENCES

REFERENCES

Althouse, J .R. and T.C. Surbrook, 1990. Voltage Gradient Modifications near
Equipotential Planes. IEEE Paper No. 90CH2823-3-A3, Institute of
Electrical and Electronic Engineers.

Althouse, J .R., 1990. Affecting Neutral-to-Earth Voltage Levels through
Modifications of the Farm Electrical Grounding System. Master’s Thesis,
Unpublished, Michigan State University.

Aneshansley, D.J., R.C. Gorewit, L. R. Price and C. S. Czarniecki, 1990. Milk
Production with Voltage Exposure during Entire Lactation. American
Society of Agricultural Engineers, Paper No. 90-3502, 2950 Niles Rd., St.
Joseph, MI 49085-9659.

Aneshansley and C. S. Czarniecki, 1990. Complex Electrical Impedance of
Cows: Measurement and Significance. American Society of Agricultural
Engineers, Paper No. 90-3509, 2950 Niles Rd., St. Joseph, MI 49085-9659.

Aneshansley, D.J., R.C. Gorewit, D.C. Ludingdon and Z. Xin, 1987. Effects
of Neutral-to-Earth Voltage on Behavior, Production and Water Intake in
Dairy Cattle. American Society of Agricultural Engineers, Paper No. 87-3034,
2950 Niles Rd., St. Joseph, MI 49085-9659.

Appleman, RD. and RJ. Gustafson. Source of Stray Voltage and Effect on
Cow Health and Performance. Journal of Dairy Science, Vol. 68, No. 6, 1985.

Biddle, J. G. Co. Megger Earth Testers (Null Balance). The Instruction
Manual.

Bodman, G. R., L. E. Stetson, H. Shull and H. Benes, 1981. Extraneous
Voltages Incidence in Nebraska Milking Centers. American Society of
Agricultural Engineers, Paper No. MCR-81-502, 2950 Niles Rd., St. Joseph,
MI 49085-9659.

224

10.

11.

12.

13.

14.

15.

16.

17.

225

Cloud, H. A., R. D. Appleman, and R. J. Gustafson, 1987. Stray Voltage
Problems with Dairy Cows. AG-BU-1359. Minnesota Extension Service,
University of Minnesota.

Currence, H.D., B.J. Steevens, D.F. Winter, WK. Dick and GP. Krause,
1987. Dairy Cow and Human sensitivity to 60 Hertz Currents. American
Society of Agricultural Engineers, Paper No. 87-3036, 2950 Niles Rd., St.
Joseph, MI 49085-9659.

Dick, W. K. and D. F. Winter, 1987. Computation, Measurement and
Mitigation of N eutral-to-Earth Potentials on Electrical Distribution Systems.
IEEE Paper No. 0885-8977/87, Institute of Electrical and Electronic
Engineers.

Drenkard, D.V., R.C. Gorewit, N .R. Scott, and R. Sagi, 1985. Milk
Production, Health, Behavior, and Endocrine Responses of Cows Exposed to
Electrical Current during Milking. Journal of Dairy Science, Vol. 68, No.10,
pp 2694-2702, 1985.

Gorewit, R. C., D. J. Aneshansley, L. R. Price and C. S. Czarniecki, 1990.
Holsteins’ Reproductive Performance during Long-Term Voltage Exposure.
American Society of Agricultural Engineers, Paper No. 90-3503, 2950 Niles
Rd., St. Joseph, MI 49085-9659.

Gorewit, R. C., D. v. Drenkard and N. R. Scott, 1984. Physiological EffeCts
of Electrical Current on Dairy Cows. National Stray Voltage Symposium,
Oct. 10-12, 1984, Syracuse, NY. ASAE Pub. 3-85. American Society of
Agricultural Engineers, 2950 Niles Rd., St. Joseph, MI 49085-9659.

Gustafson, RJ., Z. Sun and TD. Brennan, 1988. Dairy Cow sensitivity to
Short Duration Electrical Current. American Society of Agricultural
Engineers, Paper No. 88-3522, 2950 Niles Rd., St. Joseph, MI 49085-9659.

Gustafson, RJ., T.M. Brennan and RD. Appleman, 1985. Behavioral Studies
of Dairy Cows Sensitivity to AC and DC Electric Currents. Transaction of the
ASAE, Vol. 28, No. 5, pp. 1680-1685, 1985. American Society of Agricultural
Engineers, 2950 Niles Rd., St. Joseph, MI 49085-9659.

Gustafson, RJ, 1985. Understanding and Dealing with Stray Voltage in
Livestock Facilities. IEEE Paper Catalog No. 85CH2126-1-CZ, Institute of
Electrical and Electronic Engineers.

18.

19.

20.

21.

22.

23.

24.

26.

27.

226

Gustafson, R.J., H.A. Cloud, 1982. Circuit Analysis of Stray Voltage Sources
and Solutions. Transaction of the ASAE, Vol. 25, No. 5, pp. 1418-1424, 1982.
American Society of Agricultural Engineers, 2950 Niles Rd., St. Joseph, MI
49085-9659.

The Institute of Electrical and Electronic Engineers, 1991. IEEE
Recommended Practice for Grounding of Industrial and Commercial Power
Systems -IEEE Std 142-1991. IEEE, 345 East 47th Street, New York, NY
10017 USA.

The Institute of Electrical and Electronic Engineers, 1986. IEEE Guide for
Safety in AC Substation Grounding. IEEE, 345 East 47th Street, New York,
NY 10017 USA.

The Institute of Electrical and Electronic Engineers, 1983. IEEE Guide for
Measuring Earth Resistivity, Ground Impedance, and Earth Surface
Potentials of a Ground System- IEEE Std 81- 1983. IEEE, 345 East 47th
Street, New York, NY 10017 USA.

Kehrle, AM, 1984. Neutral-to-Earth Voltage Analysis of A Single-Phase
Primary Electrical Distrrbution System. Master’s Thesis, Unpublished,
Michigan State University.

Kraus, J. D., 1953. Electromagrretics. McGRAW-HILL BOOK COMPANY,
INC. New York.

The Institute of Electrical and Electronic Engineers, 1990. National
Electrical Safety Code. IEEE, 345 East 47th Street, New York, NY 10017
USA.

Norell, R.J., RJ. Gustafson and RD. Appleman, 1982. Behavioral Studies of
Dairy Cattle'Sensitivity to Electric Currents. American Society of
Agricultural Engineers, Paper No. 82-3530, 2950 Niles Rd., St. Joseph, MI
49085-9659

Pearce, J .A., J .D. Bourland, W. Neilsen, LA. Geddes, and M. Voelz, 1982.
Myocardial stimulation with ultrashort duration current pulses. PACE,
Volume 5, pp 52-58.

Phillips, D.S.M, 1969. Production Losses from Milking Plant Voltage. New
Zealand Journal of Agriculture Vol. 119, No. 2, pp.45-47.

29.

30.

31.

32.

33.

34.

35.

36.

227

Phillips, D.S.M. and R.D.J. Parkinson, 1963. The Effects of Small Voltages
on Milking Plants; Their Detection and Elimination. Dairy Farming Annual,
pp. 79-90, New Zealand.

Prothero, J.N., B.W. Lukecart and CM. DeNardo, 1988. Primary
N eutral-to-Earth Voltage Levels as Impacted by Various Wiring System
Treatments. American Society of Agricultural Engineers, Paper No. 88-3528,
2950 Niles Rd., St. Joseph, MI 49085-9659.

Reinemann, D. J ., L. E. Stetson and N. K. Laughlin, 1994. Response of Dairy
Cattle to Transient Voltages and Magnetic Fields. EBB Rural Electric
Power Conference Paper No. 94-C5.

Sodcrholm, LH., 1982. Stray-Voltage Problems in Dairy Milking Parlors.
Transaction of the ASAE, Vol. 25, pp. 1763-1767, 1774, 1982. American
Society of Agricultural Engineers, 2950 Niles Rd., St. Joseph, MI 49085-9659.

Stetson, LE, G.R. Bodman and H. Shull, 1984. An Analog Model of
Neutral-to-earth Voltages in a Single-Phase Distribution System. Transaction
of the Industry Applications Society of the Institute of Electrical and
Electronic Engineers, Volume IA-20, Number 2, Pages 418-424, EBB, New
York, NY.

Stetson, LE, G.R. Bodman and H. Shull, 1982. Digital Voltrneter for
Checking Connections in Neutral Conductors. American Society of
Agricultural Engineers, Paper No. 82-3506, 2950 Niles Rd., St. Joseph, MI
49085-9659.

Surbrook, T.C., N.D. Reese and CM. Li, 1989. Trouble-shooting
Neutral-to-Earth Voltage. Conference Record of the 1989 EBB Industry
Applications Society Annual Meeting, Paper No. 89CH2792-0.

Surbrook, T.C., N .D. Reese and J .R. Althouse, 1988. Parameters Affecting
Neutral-to-Earth Voltage along Primary Distribution Circuits. Transaction of
the Industry Applications Society of the Institute of Electrical and Electronic
Engineers, Volume 24, Number 5, pp. 798-804, EBB, New York, NY.

Surbrook, T.C., J .R. Althouse and ND. Reese, 1988. Stray Voltage
Diagnostic Procedure. American Society of Agricultural Engineers, Paper
No. 88-3520, 2950 Niles Rd., St. Joseph, MI 49085-9659.

37.

38.

39.

41.

42.

43.

228

Surbrook, T.C., N.D. Reese and J .R. Althouse, 1987. Training Power
Supplier Personnel to Find NEV. Sources. American Society of Agricultural
Engineers, Paper No. 87-3549, 2950 Niles Rd., St. Joseph, MI 49085-9659.

Surbrook, T.C., N.D. Reese and A.M. Kehrle, 1986. Stray Voltage: Sources
and Solutions. Transaction of the Industry Applications Society of the
Institute of Electrical and Electronic Engineers, Volume IA-22, Number 2,
Pages 210-215, EEE, New York, NY.

Surbrook, T.C. and ND. Reese, 1981. Stray Voltage on Farms. American
Society of Agricultural Engineers, Paper No. 81-3612, 2950 Niles Rd., St.
Joseph, MI 49085-9659.

Tagg, C. F., 1964. Earth Resistances. George N ewnes Limited, London.

U. S. Department of Agriculture - Agricultural Research Service, 1991.
Efiects of Electrical Voltage/Current on Farm Animals - How to Detect and
Remedy Problems. Agricultural Hand Book No. 696.

Woolford, M.W., 1972. Small Voltages on Milking Plants. Proceedings of the
Second Seminar on Farm Machinery and Engineering. Ruakura Agricultural
Research Center, Hamilton, New Zealand. pp. 41-47.

Woolford, M.W., 1971. Recording Transient Voltage Pulses in Milking
Plants. New Zealand Journal of Agricultural Research, Vol. 14, No. 1,
February, p. 248.

HICHIGAN STATE UNIV. LIBRARIES
IIIII IIIIII III IIIIIII IIIIII IIII III III IIII III IIII IIIII IIIIII
31293014099372