energies

Article
A New Method of Ground Fault Location in
2 × 25 kV Railway Power Supply Systems
Jesús Serrano 1 , Carlos A. Platero 1, *, Máximo López-Toledo 1 and Ricardo Granizo 2
1 Department of Electrical Engineering, ETS Ingenieros Industriales, Universidad Politécnica de Madrid,
C/José Gutierrez Abascal, 2, Madrid 28006, Spain; jserranoalv@hotmail.com (J.S.);
maximo.lopez.t@gmail.com (M.L.-T.)
2 Department of Electrical Engineering, ETS Ingeniería y Diseño Industrial,
Universidad Politécnica de Madrid, C/Ronda de Valencia, 3, Madrid 28012, Spain; ricardo.granizo@upm.es
* Correspondence: carlosantonio.platero@upm.es; Tel.: +34-91-336-31-29; Fax: +34-91-336-30-08

Academic Editor: William Holderbaum

Received: 7 December 2016; Accepted: 3 March 2017; Published: 10 March 2017

Abstract: Owing to the installation of autotransformers at regular intervals along the line, distance
protection relays cannot be used with the aim of locating ground faults in 2 × 25 kV railway power
supply systems. The reason is that the ratio between impedance and distance to the fault point is not
linear in these electrification systems, unlike in 1 × 25 kV power systems. Therefore, the location of
ground faults represents a complicated task in 2 × 25 kV railway power supply systems. Various
methods have been used to localize the ground fault position in 2 × 25 kV systems. The method
described here allows the location of a ground fault to be economically found in an accurate way
in real time, using the modules of the circulating currents in different autotransformers when the
ground fault occurs. This method first needs to know the subsection and the conductor (catenary or
feeder) with the defect, then localizes the ground fault’s position.

Keywords: ground faults; protection; 2 × 25 kV; fault location; railways

1. Introduction
Currently, the most widely used electrical configuration to feed high-speed trains is the 2 × 25 kV
railway power system. The use of this system, compared to the traditional 1 × 25 kV railway power
system, allows the required number of traction substations erected along the line to be reduced, because
it has lower losses and lower voltage drops along the line [1].
In this 2 × 25 kV system, each traction substation feeds two sections, one in each direction
of the line. Each section consists of several subsections of similar length that are delimited by
autotransformer power stations (ATS). These autotransformers have two terminals; one is connected to
the positive conductor rated 25 kV (catenary), while the other is connected to the negative conductor,
also rated 25 kV (feeder). The midpoint of its winding is connected to the rail and ground. Another
autotransformer (SATS) is installed at the end of the section and is connected in the same manner as
the intermediate autotransformers (Figure 1).
Electrical installations need adequate protection to ensure a reliable and safe operation [2]. This
need is accentuated in the case of electric railway traction systems due to the increased number of
failures caused by the sliding of the pantograph on the contact wire, which can cause its breakage.
These faults cause severe economic and social problems by disrupting rail services. Therefore, it is
very important to detect and locate the point where the ground fault has occurred in order to isolate
the subsection at the fault, and to repair such a failure as quickly as possible [3].
In AC electrical traction systems, the most-used method for measuring the distance from a traction
substation to a ground fault due its stability, reliability and simplicity, is the impedance method [4,5].

Energies 2017, 10, 340; doi:10.3390/en10030340 www.mdpi.com/journal/energies

Energies 2017, 10, 340 2 of 14

However, in 2 × 25 kV traction power systems, it is not possible to determine the distance from the
traction substation to the ground faults in an acceptable manner with this method [6]. This is due to
the presence of autotransformers that make the ratio impedance-distance nonlinear [7].
To be able to use this location method and to know the distance to the ground fault through the
impedance seen from the traction substation, it is necessary to disconnect the autotransformers and
separate the catenary from the feeder circuits when a ground fault is detected. Thus, the subsection
is divided into two circuits with only one conductor and return cable, as in the case of 1 × 25 kV
systems. Following this procedure, a ground fault can be located by measuring the impedance in each
conductor and return cable circuits. This method is used in some 2 × 25 kV traction power systems,
but it is very slow, requiring several operations of connection and disconnection of switches with
waiting times between each maneuver [8,9].
Another traditional method used for fault location in railway power systems that employs
auto-transformers is the AT neutral current ratio fault location method [10]. However, the capacity
of the main transformer is much higher than that of the autotransformer stations. For this reason,
this method can be insufficiently accurate and identifies failures in the wrong subsections. Therefore,
a method of compensation using the neutral currents circulating in adjacent autotransformers to those
of the faulty section was proposed in [11].
The location method using travelling waves, in addition to transmission and distribution lines,
has been applied to the lines of railway traction power systems, aided by advances in the theory of
wave propagation, microelectronics and data processing [4]. In this method, when a ground fault
occurs, there are various types of technologies to locate ground faults according to the detection of the
generated travelling wave at one end [12] or at both ends [13], and if it is performed by the wave itself
or by a pulse injected into one end when such a fault occurs [14]. However, this method is difficult to
implement because of its limited reliability, complex setup and high economic investments [15]. This
difficulty is due to the specific characteristics of traction facilities with numerous discontinuities in the
contact line and multiple loads changing position continuously [16].
This article describes a new ground fault location method that allows the position in the railway
line where a ground fault has occurred to be obtained in real time, and also whether it has been in the
catenary (positive conductor) or feeder (negative conductor). The method locates the position of the
failure using the values of the modules of the currents in different autotransformers when a ground
fault occurs, and is based on the knowledge of the subsection and the conductor where the ground
fault has occurred following the method described in [17]. The method that identifies the subsection
with a ground fault is summarized in Section 2. In Section 3, the distribution of currents in the
autotransformers is presented for different ground fault locations by computer simulations. Section 4
describes the principles of the new method for ground fault location. In Section 5, the experimental
results of tests carried out in the laboratory are introduced. Finally, Section 6 summarizes the main
contributions of the paper.

2. Brief Description of Ground Fault Identification Method for 2 × 25 kV Railway Power

Supply Systems
Figure 1 shows the simplified scheme of an electric section that is fed by a 2 × 25 kV power system
with three subsections (A, B and C) delimited by intermediate autotransformers ATS1 and ATS2. At the
end of the section, an autotransformer SATS connects the three conductors. Also, the distribution of
currents in the conductors is represented when a train travels in the intermediate section B.
According to [17], when in a line of a 2 × 25 kV rail power system supply, a fault occurs between
the catenary and rail or between the feeder and rail, there is a significant increase in the current
flowing in the windings of the closest autotransformers to the position where the fault has occurred.
In addition, the angles between these currents and voltages between the catenary and rail change at
the location of the autotransformers.
Energies 2017, 10, 340 3 of 14
Energies
Energies2017,
2017,10,
10,340
340 33of
of14
14

Section
Section

Subsection
SubsectionAA Subsection
SubsectionBB Subsection
SubsectionCC

33 Catenary 11
II I Catenary I
22 I 22 I
++25
25kV
kV 2I2I
11 11
22II 22II

II Rail
Rail II
11 11
I 22II
22 I 11
II I
22 I
TS - 25 kV
TSTraction
TractionSubstation
Substation - 25 kV ATS
ATS11 Feeder
Feeder ATS
ATS22 SATS
SATS

Figure 1.Simplified
Figure1.1.
Figure Simplifieddiagram
Simplified diagram ofa a2a 2×
diagramofof 2 ××25
25kV
25 kVrail
kV railpower
rail powersystem
power systemand
system andcurrent
and currentdistribution
current distributionwhile
distribution whileaaatrain
while train
train
travels
travelsininsubsection
subsection B.B.
travels in subsection B.

The
The power
Thepower factor
powerfactorfactorofof the
ofthe high-speed
thehigh-speed
high-speedtrains trains
trainsisis close
isclose
closeto to the
tothe unity,
unity,soso
theunity, the
sothe impedance
theimpedance
impedanceseen seen from
seenfrom
fromthe the
the
autotransformer
autotransformer is
is mainly
mainly resistive.
resistive. In
In aa line
line feeding
feeding trains
trains in
in normal
normal
autotransformer is mainly resistive. In a line feeding trains in normal operation without any faults, operation
operation without
without any
any faults,
faults,
these angles are close ◦.
these
theseangles
anglesare closetoto
areclose 180°.
to180
180°.
But
Butifififaaaground
But ground
groundfault fault arises,
faultarises,
arises,thethe impedance
theimpedance
impedanceseen seen from
seenfromfromthethe autotransformer
theautotransformer
autotransformercan can
canbebe the
bethe catenary
thecatenary
catenaryor or
or
feeder
feeder inductance.
inductance. So,
So, the
the angle
angle between
between currents
currents and
and voltages
voltages in
in
feeder inductance. So, the angle between currents and voltages in the autotransformers closest to the the
the autotransformers
autotransformers closest
closest toto the
the
fault changes increasing ◦ ordecreasing
90° ◦ becausethey
fault
faultchanges
changesincreasing
increasing90 90°or or decreasing
decreasing90 90° because they
90°because theyare are reactive
arereactive currents
reactivecurrents without
currentswithout resistance
withoutresistance
resistance
component.
component. There is mainly
mainly reactive
reactive power
power inin the
the 2 2× × 25 kV
component. There is mainly reactive power in the 2 × 25 kV supply system. In the case ofground
There is 25 kV supply
supply system.
system. In
In the
the case
case of
of ground
ground
fault
fault between the catenary and ground, the angle between the currents and voltages in the nearest
fault between
between the
the catenary
catenary and
and ground,
ground, the
the angle
angle between
between the
the currents
currents and
and voltages
voltages inin the
the nearest
nearest
autotransformers ◦ . However,
autotransformersisis
autotransformers approximately
isapproximately
approximately90 90°.
90°. However,
However,ifififthe the fault
thefault occurs
faultoccurs between
occursbetween
betweenthe the feeder
thefeeder
feederandand ground,
andground,
ground,
the angle between thethe currents andand voltages in the ◦ This last
the
theangle
angle between
between currents
the currents voltages
and voltages in the the nearest
in nearest autotransformers
autotransformers
nearest autotransformers is now is now
now. 270°.
is270 270°. This
This last
typelast
type
oftype of fault
faultofandfault and
theand the
angles angles
thebetween between
the currents
angles between the currents
and voltages
the currents and
and involtages in
the nearest
voltages the nearest autotransformers
autotransformers
in the are shown are
nearest autotransformers in
are
shown
shown2.in
Figure in Figure
Figure 2. 2.
Subsection
SubsectionAA Subsection
SubsectionBB Subsection
SubsectionCC

VVC1
C1
VVC2
C2
VVC3
C3

IIA1
A1
IIA2
A2
IIA3
A3

Fault
Fault

Traction
TractionSubstation
Substation ATS
ATS11 ATS
ATS22 SATS
SATS

IIA1
A1 270º
180º 270º 270º
180º 270º

VVC1
C1
VVC2
C2
VVC3
C3

IIA2 IIA3
A3
A2

Figure
Figure 2. Angular difference between currents (IA ) and voltages in catenary (VC )C)ininautotransformers
Figure 2.2. Angular
Angular difference
difference between
between currents
currents (I(IAA)) and
and voltages
voltages in
in catenary
catenary (V
(VC) in autotransformers
autotransformers
with
withfault between feeder and ground in subsection C.
with fault
fault between
between feeder
feeder and
and ground
ground in
in subsection
subsection C. C.
 Energies 2017, 10, 340 4 of 14
Energies 2017, 10, 340 4 of 14

Therefore, if the currents and voltage values, including the phase angle difference between them,
areTherefore,
measuredifinthethecurrents
autotransformers,
and voltagewhen a fault
values, occurs
including thebetween any conductor
phase angle difference and the ground,
between them,
areit measured
is possibleintothe
identify in which subsection
autotransformers, the fault
when a fault took
occurs place, and
between which conductor
any conductor and thehad such a
ground,
fault.
it is possible to identify in which subsection the fault took place, and which conductor had such a fault.

3. 3.
Current Distribution
Current inin
Distribution Autotransformers forfor
Autotransformers Different Ground
Different Fault
Ground Locations
Fault 2×
in in
Locations 2525kV
2× kV
Rail Power
Rail Supply
Power SupplySystem
System
AsAsexplained
explained before,
before, if if
a ground
a ground fault takes
fault takesplace between
place between thethe
catenary
catenaryandandrail
railoror
between
between thethe
feeder
feeder and
andrail in in
rail a power
a power system
system fedfed
byby2× 2 ×2525kV, the
kV, values
the valuesofof
the currents
the flowing
currents flowing inin
thethe
windings
windings
ofof
thethe
closest autotransformers to the point where the fault has occurred experience a
closest autotransformers to the point where the fault has occurred experience a very significant very significant
increase.
increase. These
Thesevalues
valuescan canbebeeasily
easilyobtained
obtainedby bymeasuring
measuringthethecurrent
currentininthose
those windings.
windings.
ToTosimulate
simulate thethe
current
current distributions
distributions in inautotransformers
autotransformers when
when ground
ground faults
faults occur
occur along
alongallall
positions of a section, the circuit shown in Figure 3 was used. The modified nodal
positions of a section, the circuit shown in Figure 3 was used. The modified nodal circuit analysis circuit analysis
method
method [18] is is
[18] applied
applied to to
solve
solvesuch
sucha circuit using
a circuit MATLAB
using MATLAB (MathWorks
(MathWorks Inc.: Natic,
Inc.: Natic,MA,
MA, USA).
USA).

ZC [Ω/km] VC1 VC2 VC3

Fault
IA1 IA2 IA3
ZR [Ω/km]

ZF [Ω/km]

ATS 1 ATS 2 SATS

Subsection A Subsection B Subsection C

0 km 10 km 20 km 30 km

Figure
Figure 3. 3. Simulated
Simulated 2×2 ×2525kV
kVpower
powersupply
supplysystem
systemwith
withtwo
twoconductors
conductorsand
andreturn
returngrounded
grounded wire.
wire.
Fault in subsection C between catenary and
Fault in subsection C between catenary and rail. rail.

In these simulations, a value of 27.5 kV has been given to each source of alternating current (AC)
In these simulations, a value of 27.5 kV has been given to each source of alternating current (AC)
voltage. The impedance values per unit length of each conductor are listed in Table 1, where ZC is the
voltage. The impedance values per unit length of each conductor are listed in Table 1, where ZC is the
impedance value per unit length of the catenary, ZF is the corresponding impedance of the feeder;
impedance value per unit length of the catenary, ZF is the corresponding impedance of the feeder; and
and ZR is the impedance of the rail that is grounded. Likewise, the values of the mutual impedances
ZR is the impedance of the rail that is grounded. Likewise, the values of the mutual impedances per
per unit length between the catenary, feeder and rail are also shown in Table 1.
unit length between the catenary, feeder and rail are also shown in Table 1.
Table 1. Self and mutual impedance values of railway conductors.
Table 1. Self and mutual impedance values of railway conductors.
Conductor Self-Impedance Ω/km
Conductor
Catenary ZC 0.1197 + j0.6224 Ω/km
Self-Impedance
Feeder
Catenary ZZCF 0.1114 + j0.7389
0.1197 + j0.6224
Feeder Rail ZZFR 0.0637 + j0.5209
0.1114 + j0.7389
Rail Conductors ZR 0.0637 + j0.5209
Mutual Impedance Ω/km
Catenary-Feeder
Conductors ZCF 0.0480 Impedance
Mutual + j0.3401 Ω/km
Catenary-Rail ZCR 0.0491 + j0.3222
Catenary-Feeder ZCF 0.0480 + j0.3401
Feeder-Rail ZFR 0.0488 + j0.2988
Catenary-Rail ZCR 0.0491 + j0.3222
Feeder-Rail ZFR 0.0488 + j0.2988
The values of the impedance parameters of the transformer of the substation and of the
autotransformers used in the simulation are shown in Table 2.
The values of the impedance parameters of the transformer of the substation and of the
autotransformers used in the simulation are shown in Table 2.
Energies 2017, 10, 340 5 of 14

Table 2. Transformers and autotransformers ratings used in the simulation.

Energies 2017, 10, 340 Main Transformer 5 of 14

Energies 2017, 10, 340 5 of 14

Voltage ratioand autotransformers

Table 2. Transformers × 27.5used in thekV
400/2 ratings simulation.
Table 2. Transformers
Rated powerand autotransformers60 ratings used in the
MVAsimulation.
Short-circuit voltage Main Transformer
10
Main Transformer %
Voltage ratio 400/2 × 27.5 kV
Voltage ratio
Autotransformers400/2 × 27.5 kV
Rated power 60 MVA
Rated power 60 MVA
Short-circuit
Voltage ratio voltage 10
55/27.5 %
kV
Short-circuit voltage 10 %
Rated power Autotransformers
15 MVA
Autotransformers
Voltage
Short-circuit ratio
voltage 155/27.5 kV
%
Voltage ratio 55/27.5 kV
Rated power 15 MVA
Rated power 15 MVA
Short-circuit voltage 1 %
Short-circuit voltage 1 %
This circuit has been resolved making faults between the catenary and feeder conductors and rail
This
in all positions
This ofcircuit has
has been
the three
circuit resolved
resolved making
subsections
been A, B and
making faults between
C that
faults the
the catenary
compose
between the 30and
catenary kmfeeder
and section.
feeder conductors and
Thus, different
conductors and
rail
values of in
the all positions
currents inof the
the three subsections
autotransformers A, B
haveand C that
been compose
obtained the
with30 km section.
ground Thus,
rail in all positions of the three subsections A, B and C that compose the 30 km section. Thus, different of
faults at different
all points
values
the section
valuesin of
both
of the currents
the cases:
currentswhenin
in the autotransformers
thethe fault is between
autotransformers have been
havethe obtained
catenary
been with
obtainedand ground
withrail and faults
ground whenat
faults atitall
is points
all between
points of
of the
the section
the section in both
in both cases:
cases: when the
when the fault
fault is between the
is between in catenary
theFigures and
catenary4and rail and when it is between the
feeder and rail. The results obtained are represented andrail
5. and when
Figure it is between
4 shows the of
the values
feeder
feederI and
and rail. The
TheIresults
rail.and results obtained
obtained are
are represented
represented in
in Figures
Figures 44 and
and 5.
5. Figure
Figure 44 shows
shows the the values
values of
of
the currents
the A1 , IA2
currents I A3 for catenary-rail faults; and Figure 5 shows the values of the currents IA1 ,
A1, IA2 and IA3 for catenary-rail faults; and Figure 5 shows the values of the currents IA1,
the currents IA1, IA2 and IA3 for catenary-rail faults; and Figure 5 shows the values of the currents IA1,
IA2 andIA2
IA3and
forIA3feeder-rail
for feeder-railfaults.
faults.
IA2 and IA3 for feeder-rail faults.

7000
7000
IA1
6000 IA1
6000
IA2
IA2
5000
5000 IA3
IA3
[A]

4000
Current[A]

4000
Current

3000
3000

2000
2000

1000
1000

0
00 5 10 15 20 25 30
0 5 10 15 [km]
Distance 20 25 30
Distance [km]

Figure 4.Figure 4. Currents

Currents obtained obtained in simulations
in simulations of the of the autotransformers
autotransformers ATS1,ATS1,
ATS2ATS2 and SATS
and SATS for
for different
Figure 4. Currents obtained in simulations of the autotransformers ATS1, ATS2 and SATS for
different
fault locations. fault locations. Fault between catenary and rail.
different Fault betweenFault
fault locations. catenary andcatenary
between rail. and rail.

7000
7000
IA1
6000 IA1
6000
IA2
IA2
5000 IA3
5000 IA3
[A]

4000
Current[A]

4000
Current

3000
3000

2000
2000

1000
1000

0
00 5 10 15 20 25 30
0 5 10 15 [km]
Distance 20 25 30
Distance [km]

Figure 5. Currents obtained in simulations of the autotransformers ATS1, ATS2 and SATS for different
fault locations. Fault between feeder and rail.
Energies 2017, 10, 340 6 of 14

Figure 5. Currents obtained in simulations of the autotransformers ATS1, ATS2 and SATS for different
Energies 2017, 10, 340 6 of 14
fault locations. Fault between feeder and rail.

As As
shown
shown in Figures 4 and
in Figures 5, the
4 and module
5, the module of the current
of the currentin each autotransformer
in each autotransformer hashas
its highest
its highest
value when
value whenthethe
fault occurs
fault occursvery
veryclose
closetotothe
thecorresponding
corresponding autotransformer.
autotransformer. ItIt isisalso
alsoremarkable
remarkablethat
thatthe
thecurrent
currentofofone
oneautotransformer
autotransformerisisapproximately
approximately equal equal to
to that
that of
of the
the adjacent
adjacentautotransformer
autotransformer at
at the
themidpoint
midpointofofthethesection
sectiondelimited
delimited byby these
thesetwo
two autotransformers.
autotransformers. This feature
This featureis observed
is observedwhenwhen
thethe
fault takes
fault place
takes between
place thethe
between catenary
catenaryandand
rail,rail,
andand
alsoalso
between
betweenthethe
feeder andand
feeder rail.
rail.

4. New Method
4. New for for
Method Ground Fault
Ground Location
Fault in 2in×225×kV
Location 25 Traction Power
kV Traction Systems
Power Systems
This new
This method
new methodfor for
ground fault
ground location
fault location × kV
in 2in×225 power
25 kV systems
power is divided
systems intointo
is divided three
three
different stages
different (Figure
stages 6). 6).
(Figure

Subsection A Subsection B Subsection C

MEASSUREMENT
Step 1

VC1 VC2 VC3

IA1 IA2 IA3

Fault

Traction Substation ATS 1 ATS 2 SATS

SUBSECTION and CONDUCTOR

IDENTIFICATION
Step 2 IA3
IA2
Phase comparator.

IA1 180º 90º 90º

VC1 VC2 VC3

FAULT LOCATOR
Step 3
Current [A]

Previously calculated
short-circuit current vs distance.
Look-up Table. IA3
IA2

Distance [km]
Fault Location

Figure 6. Simplified
Figure layout
6. Simplified of the
layout location
of the method.
location method.

4.1.4.1.
Measurement
Measurement
In the firstfirst
In the stage, thethe
stage, voltages andand
voltages thethe
currents
currentsin the
in thedifferent autotransformers
different autotransformersshould be be
should
measured, as well as the phase angle between them (VC1(V
, IA1;, VC2, IA2; VC3, IA3).
measured, as well as the phase angle between them C1 IA1 ; VC2 , IA2 ; VC3 , IA3 ).
In the case of a ground fault, the current in some of the autotransformers
In the case of a ground fault, the current in some of the autotransformers would would exceed
exceed a
a prefixed
prefixed threshold,
threshold, and the and the location
location systemsystem
shouldshould be activated.
be activated.
Energies 2017, 10, 340 7 of 14
Energies 2017, 10, 340 7 of 14

4.2.
4.2. Subsection
Subsectionand
andConductor
ConductorIdentification
Identification
Afterwards,
Afterwards, the the subsection
subsection between
between autotransformers
autotransformersandand the
the positive
positive or
or negative
negative conductor
conductor
where
where the fault has occurred are identified by analysing the phase angle between the
the fault has occurred are identified by analysing the phase angle between the voltages
voltages and
and
currents in the autotransformers, as was summarised in Section 2. The completed method is
currents in the autotransformers, as was summarised in Section 2. The completed method is described described
in
in [17].
[17].

4.3.
4.3. Fault
FaultLocator
Locator
Finally, thefault
Finally, the fault locator
locator is based
is based on the previous
on the previous calculationcalculation of in
of the currents thethecurrents in the
autotransformer
autotransformer
for a ground fault foralong
a ground
all the fault alongof all
positions the positions
a section, of a section,
as per Section 3. The as per Section
calculations 3. The
should be
calculations should be performed
performed specifically taking intospecifically
consideration taking
the into consideration
precise data of thethe
powerprecise data The
system. of the power
values of
system. The values of the fault current should
the fault current should be stored as look-up tables. be stored as look-up tables.
The
The fault
fault location
location isis performed
performed by by aa comparison
comparison of of the
the current
current measured
measured in in the
the two
two adjacent
adjacent
autotransformers
autotransformers to to the pre-calculated value storage in the look-up tables. In this way, the
value storage in the look-up tables. In this way, the location of location
of
thethe fault
fault can
can bebe found.
found. AA flowchart
flowchart of of
thethe ground
ground fault
fault location
location method
method is presented
is presented in in Figure
Figure 7. 7.

Start

- GET VALUES OF CURRENTS IA1 , IA2 , IA3

Step 1 AND VOLTAGES VA1 , VA2 , VA3
WHEN GROUND FAULTS HAPPENS

- GET SUBSECTION WITH FAULT

Step 2 - GET POSITIVE OR NEGATIVE CONDUCTOR
WITH FAULT

- COMPARE WITH CURRENT VALUES

IA1 (n), IA2 (n), IA3 (n) IN LOOK-UP TABLES

Yes
Subs = A
No IA1 (n) < IA1 yes
n=n+1
No
yes E A
Subs = B
Yes
No IA2>IA1
Step 3
No
Yes Yes
IA1 (n) > IA1 IA2 (n) < IA2
n=n+1 n=n+1
No No
E A E A
Subs= C Yes
Yes
No IA3>IA2
No
Yes Yes
IA2 (n) > IA2 IA3 (n) < IA3
n=n+1 n=n+1
No No
E A E A
Error
n=position

Flowchartof
Figure7.7.Flowchart
Figure ofthe
theground
groundfault
faultlocation
locationmethod.
method.

This method could

This method couldbebeeasily
easily implemented
implemented in 2in
× 225×kV
25power
kV power systems,
systems, as the voltages
as the voltages and
and currents
currents in the autotransformers
in the autotransformers are recorded are recorded
by the protectionbyrelays
the installed
protection relays
in the installed in and
autotransformers, the
autotransformers,
this information is and this sent
normally information is normally
to the Supervisory sent to
Control andthe Supervisory
Data AcquisitionControl and Data
System “SCADA”.
Acquisition System “SCADA”.
 Energies 2017, 10, 340 8 of 14
Energies 2017, 10, 340 8 of 14
Energies 2017, 10, 340 8 of 14
4.4. Fault Location for Sections with n Subsections
4.4.Fault
4.4. FaultLocation
Location forSections
Sectionswith
withn nSubsections
Subsections
This methodfor is also valid for any section that has at least two subsections, since the pattern that
follows This
This the values
method
method ofalso
isisalsothe modules
valid
valid forany
for ofsection
any the currents
section in at
thathas
that has such cases
atleast
least two
two is totally similar
subsections,
subsections, tothe
since
since that in sections
thepattern
pattern that
that
with
follows
follows three
the
the subsections.
values
values of ofthethe modules
modules of the
of the currents
currents in such
in such casescases is totally
is totally similar
similar to thattointhat in sections
sections with
with Figures
three 8 and 9
subsections.
three subsections. show the values of the current modulus as a function of the distance to the fault
in aFigures
section
Figures8with
8and
andtwo 9 subsections,
show the both
values of for
the catenary
current and feeder
modulus asfaults.
a function of the distance
9 show the values of the current modulus as a function of the distance to the fault to the fault
inina asection
sectionwith
withtwo
twosubsections,
subsections,bothbothfor
forcatenary
catenaryandandfeeder
feederfaults.
faults.
7000
7000 IA1
6000 IIA2
A1
6000 IA2
5000
5000
[A] [A]

4000
Current

4000
Current

3000
3000
2000
2000
1000
1000
0
0 2 4 6 8 10 12 14 16 18 20
0 Distance [km]
0 2 4 6 8 10 12 14 16 18 20
Distance [km]
Figure 8. Currents in autotransformers as a function of the distance from the traction substation when
Figure 8. Currents in autotransformers as a function of the distance from the traction substation
aFigure
ground fault occurs
8. Currents between the catenary
in autotransformers and rail
as a function along
of the different
distance fromzones in a section
the traction withwhen
substation two
when a ground fault occurs between the catenary and rail along different zones in a section with
subsections.
a ground fault occurs between the catenary and rail along different zones in a section with two
two subsections.
subsections.

7000
7000 IA1
6000
IA1
6000 IA2
5000 IA2
5000
[A] [A]

4000
Current

4000
Current

3000
3000
2000
2000
1000
1000
0
0 2 4 6 8 10 12 14 16 18 20
0 Distance [km]
0 2 4 6 8 10 12 14 16 18 20
Distance [km]
Figure 9. Currents in autotransformers as a function of the distance from the traction substation when a
Figure 9. Currents in autotransformers as a function of the distance from the traction substation when
ground fault occurs between the feeder and rail along different zones in a section with two subsections.
aFigure
ground fault occurs
9. Currents between the feeder
in autotransformers and railof along
as a function different
the distance zones
from in a section
the traction withwhen
substation two
subsections.
a ground fault occurs between the feeder and rail along different zones in a section with two
Furthermore
subsections. in Figures 10 and 11, the values of the current modulus are represented as a function
of the Furthermore
distance to theinground fault
Figures 10 inand
a section with
11, the four subsections
values for both
of the current catenary
modulus are and feeder faults.
represented as a
function of the distance
Furthermore to the 10
in Figures ground fault
and 11, in values
the a section
of with four subsections
the current modulus for
areboth catenary as
represented anda
feeder faults.
function of the distance to the ground fault in a section with four subsections for both catenary and
feeder faults.
 Energies 2017, 10, 340 9 of 14
Energies 2017,
Energies 10,10,
2017, 340340 9 of 14 14
9 of

7000
7000
IA1
IA1
6000
6000 IA2
IA2
5000 IA3
5000 IA3
IA4
IA4
Current[A][A]

4000
4000
Current

3000
3000

2000
2000

1000
1000

0
00 5 10 15 20 25 30 35 40
0 5 10 15 Distance
20 [km] 25 30 35 40
Distance [km]

Figure10.10.Currents
Currentsininautotransformers
autotransformers asa afunction
functionofofthe
thedistance
distancefrom
fromthe
thetraction
tractionsubstation
substation
Figure
Figure 10. Currents in autotransformersasas a function of the distance from the traction substation
when
when a ground
a ground fault occurs between the catenary and rail along different zones in a section with four
when a groundfault
faultoccurs
occursbetween
between the
the catenary and rail
catenary and railalong
alongdifferent
differentzones
zonesinin a section
a section with
with four
subsections.
four subsections.
subsections.

7000
7000
IA1
6000 IA1
6000 IA2
IA2
5000 IA3
5000 IA3
IA4
Current[A][A]

4000 IA4
4000
Current

3000
3000

2000
2000

1000
1000

0
00 5 10 15 20 25 30 35 40
0 5 10 15 Distance
20 [km] 25 30 35 40
Distance [km]

Figure Currents
11.11.
Figure in autotransformers
Currents as a function
in autotransformers of the of
as a function distance from thefrom
the distance traction
the substation when a
traction substation
Figurefault
ground 11. occurs
Currents in autotransformers
between the feeder and as along
rail a function of the
different distance
zones in a from with
section the traction
four substation
subsections.
when a ground fault occurs between the feeder and rail along different zones in a section with four
when a ground fault occurs between the feeder and rail along different zones in a section with four
subsections.
subsections.
From all the results illustrated in previous figures it can be concluded that this ground fault
locationFrom all the
process results illustrated
is applicable in previous
to any section figures
with two it can
or more be concluded that this ground fault
subsections.
From all the results illustrated in previous figures it can be concluded that this ground fault
location process is applicable to any section with two or more subsections.
5. location process
Experimental is applicable to any section with two or more subsections.
Tests
5. Experimental
In addition toTests
the computer simulations, numerous tests have been performed in the laboratory
5. Experimental Tests
to obtain
In addition to thetables
the necessary in order
computer to store in
simulations, memory tests
numerous the currents of the
have been autotransformers
performed used into
in the laboratory
In addition to the computer simulations, numerous tests have been performed in the laboratory to
the circuit represented in Figure 12 when several ground faults are performed.
obtain the necessary tables in order to store in memory the currents of the autotransformers used in the
obtain the necessary tables in order to store in memory the currents of the autotransformers used in the
Therepresented
circuit experimental setup used
in Figure in the
12 when laboratory
several groundis faults
shown inperformed.
are Figure 13. It simulates the circuit of
circuit represented in Figure 12 when several ground faults are performed.
Figure 12. In this experimental setup, the traction substation was implemented using a power source
(1) and the traction transformer was made using two transformers (2) and (3) providing 25 V AC each
(see Figure 13).
 Energies 2017, 10, 340 10 of 14

Subsection A Subsection B Subsection C

Energies 2017, 10, 340 10 of 14
Energies 2017, 10, 340 10 of 14

VC1 VC2 VC3

Subsection A Subsection B Subsection C
L1 L2 L L
Fault IA1 IA2 IA3
VC1 VC2 VC3

L1 L2 L L
Fault IA1 IA2 IA3

TSTraction Substation L ATS 1 L ATS 2 L SATS

Figure 12. Laboratory circuit to simulate a section fed by a 2 × 25 kV system with two conductors and
ground return. A fault between the catenary and rail in subsection A is represented.
TSTraction Substation L ATS 1 L ATS 2 L SATS
The experimental setup used in the laboratory is shown in Figure 13. It simulates the circuit of
Figure 12. Laboratory circuit to simulate a section fed by a 2 × 25 kV system with two conductors and
Figure
Figure In Laboratory
12.12. circuit tosetup,
this experimental simulate
the atraction
section fed by a 2 ×was
substation 25 kV system with using
implemented two conductors and
a power source
ground return. A fault between the catenary and rail in subsection A is represented.
(1)ground
and thereturn. A transformer
traction fault betweenwas the made
catenary andtwo
using rail transformers
in subsection A
(2)isand
represented.
(3) providing 25 V AC each
(see Figure 13).
The experimental setup used in the laboratory is shown in Figure 13. It simulates the circuit of
Figure 12. In this experimental setup, the traction substation was implemented using a power source
(1) and the traction transformer was made using two transformers (2) and (3) providing 25 V AC each
(see Figure 13).

Figure 13. Laboratory setup for 2 × 25 kV power system fault location testing.
Figure 13. Laboratory setup for 2 × 25 kV power system fault location testing.

The section is divided into three subsections by autotransformers ATS1, ATS2 and SATS (4). The
The section
distributed is divided
impedances into
of the three subsections
catenary and feeder wereby autotransformers ATS1, ATS2
simulated using inductive and SATS
impedances L (5).(4).
The distributed Figure 13. Laboratory setup for 2 × 25 kV power system fault location testing.
Moreover, the impedances
rail impedance ofwas
the considered
catenary and feeder and
negligible werenosimulated
impedance using inductiveToimpedances
was installed. simulate
L (5). Moreover, the rail impedance was considered negligible and no impedance
faults (7) along different points of each subsection, variable inductors L1 and L2 (6), were used. was installed.
Such
The section is divided into three subsections by autotransformers ATS1, ATS2 and SATS (4). The
To inductors
simulate faults (7) along different points
can be observed in detail in Figure 14.of each subsection, variable inductors L1 and L2 (6), were
distributed impedances of the catenary and feeder were simulated using inductive impedances L (5).
used. Such inductors
The values can
of the be observedsetup
experimental in detail in Figure
impedances per14.
unit length of the catenary, feeder and rail
Moreover, the rail impedance was considered negligible and no impedance was installed. To simulate
areThe values
shown of the3.experimental
in Table setup impedances
No mutual impedances have been per unitinlength
taken of the catenary, feeder and rail
consideration
faults (7) along different points of each subsection, variable inductors L1 and L2 (6), were used. Such
are shown in Table 3. No mutual impedances
inductors can be observed in detail in Figure 14. have been taken in consideration
Table 3. Experimental setup impedance values of railway conductors.
The values of the experimental setup impedances per unit length of the catenary, feeder and rail
Table 3. Experimental setup impedance values of railway conductors.
are shown in Table 3. No mutual impedancesSelf-Impedance
Conductor have been takenΩ/km in consideration
Catenary
Conductor
Z C 0.10 + j1.00
Self-Impedance Ω/km
Table 3. Experimental
Feeder setupZF impedance 0.10 +values
j1.00 of railway conductors.
Catenary ZC 0.10 + j1.00
Rail
ConductorZR 0
Self-Impedance Ω/km
Feeder ZF 0.10 + j1.00
Catenary ZZ
Rail 0
0.10 + j1.00
RC
Feeder ZF 0.10 + j1.00
Railtransformer
Additional information about the ZR 0
and autotransformers used in the experimental
setup is shown in Table 4.
Energies 2017, 10, 340 11 of 14
Energies 2017, 10, 340 11 of 14
Additional information about the transformer and autotransformers used in the experimental
setupEnergies 2017, 10, 340
is shown in Table 4. 11 of 14
Table 4. Experimental setup transformers and autotransformers ratings.
Additional information about the transformer and autotransformers used in the experimental
Table 4. Experimental setup transformers and autotransformers ratings.
setup is shown in Table 4. Main Transformer
Main Transformer
Voltage ratio 400/2 × 25 kV V
Voltage
Rated ratio
Table 4. Experimental
power setup transformers 400/2 × 25 kV VA Vratings.
and autotransformers
1600
Rated power
Short-circuit voltage 5.38 1600 % VA
Main Transformer
Short-circuit voltage 5.38 %
Voltage ratio Autotransformers400/2 × 25 kV V
Autotransformers
Rated
Voltage power
ratio 50/25 1600 V VA
Voltage ratiovoltage
Short-circuit 50/25
5.38 V
Rated power 300 VA %
Rated power 300 VA
Short-circuit voltageAutotransformers
2.17 %
Short-circuit voltage
Voltage ratio 2.17
50/25 %V
Rated power 300 VA
With this
With this setup,
setup,the
thecurrents
currents IA1
IA1 , I,A2IA2
Short-circuit , and IA3Iwere
, and
voltage A3 were measured.
measured. These
2.17These currents
currents
% were were flowing
flowing in
in the
the three
three autotransformers
autotransformers usedused
whenwhen
a fault a fault was executed
was executed between
between the catenary
the catenary andThese
and rail. rail. faults
These
faults were made
With this along
setup, the
the section
currentsthat
I , simulated
I , and I a length
were of 30
measured. km with
These three
currents
were made along the section that simulated a length of 30 km with three subsections of 10 km,
A1 A2 A3 subsections
were of
flowing 10
in km,
the
performing
performing measurements used
three autotransformers
measurements everywhen
every 0.5 km.
0.5 km.
a faultFigure
Figurewas1515 shows the
executed
shows the evolution
evolution
between of these
theseand
the catenary
of currents depending
rail. These
currents faults
depending
on were
the made
position along
where the
the section
fault that
was simulated
developed fora length
a of 30 km
catenary-ground
on the position where the fault was developed for a catenary-ground fault. with three
fault. subsections of 10 km,
performing measurements every 0.5 km. Figure 15 shows the evolution of these currents depending
on the position where the fault was developed for a catenary-ground fault.
Variable inductance 10 x 0.5 Ω

(a) (b)
(a) (b)
Figure 14. Variable inductance (L1, L2) used to simulate faults at any point of any subsection; (a)
Figure 14. Variable inductance (L1, L2) used to simulate faults at any point of any subsection; (a) Picture;
Figure
Picture; 14. Variable
(b) Electrical inductance (L1, L2) used to simulate faults at any point of any subsection; (a)
(b) Electrical diagram.diagram.
Picture; (b) Electrical diagram.

1200
1200
I
A1
1000 IA1
1000
IA2
IA2
800
800 I
Current [mA]

A3
IA3
Current [mA]

600
600

400
400

200 200

0 0
0 0 10 10 20 20 30 30 4040 50
50 60
60
Distance
Distance [km]
[km]

Figure 15. Autotransformers

Figure current
15. Autotransformers distribution
current in in
distribution thethe
tests with
tests witha afault
faultbetween
betweenthe
thecatenary
catenary and rail.
rail.
Figure 15. Autotransformers current distribution in the tests with a fault between the catenary and rail.
Energies 2017, 10, 340 12 of 14
Energies 2017, 10, 340 12 of 14

Also
Alsoininthis
thisfacility, thethe
facility, currents flowing
currents flowingin the three
in the autotransformers
three autotransformers IA2
(IA1(I, A1 and
, IA2 IA3IA3
and ) were
) were
measured when a fault was made between the feeder and rail every 0.5 km along the section.
measured when a fault was made between the feeder and rail every 0.5 km along the section. The The results
are shown
results inshown
are Figurein16.Figure 16.

1200

IA1
1000
IA2
800
IA3
Current [mA]

600

400

200

0
0 5 10 15 20 25 30
Distance [km]

Figure 16.Autotransformers
16.
Figure Autotransformerscurrent
currentdistribution
distributionin
inthe
thetests
tests with
with aa fault
fault between the feeder
feeder and
and rail.
rail.

AsAs
it it
is isobserved,
observed,inincase
caseofofground
groundfault,
fault,the
thecurrent
currentdistributions
distributionsobtained
obtainedthrough
throughexperimental
experimental
results (Figures 15 and 16) are similar to those obtained through computer simulations
results (Figures 15 and 16) are similar to those obtained through computer simulations (Figures 4 and (Figures 4 and
5).
5). Therefore the simulation model is validated, as well as the
Therefore the simulation model is validated, as well as the location method. location method.

6.6.Conclusions
Conclusions

WeWe have
have described
described a new
a new method
method of of ground
ground fault
fault location
location forfor
2×2 ×2525kVkVtraction
tractionpower powersystems
systems
based
based onon the
the measurement
measurement ofof
thethe currents
currents andand voltages
voltages inin the
the autotransformers
autotransformers along
along a section.
a section.
Thesubsection
The subsectionand andthe thepositive
positiveorornegative
negativeconductors
conductorswhere wherethe theground
groundfault faulttakes
takesplaceplaceare
are
determined
determined by analyzing
by analyzing the angle
the phase phasebetween
angle the between
voltagestheandvoltages
the currents and theautotransformers
in the currents in the
atautotransformers at both ends of the subsection.
both ends of the subsection.
Thefault
The fault locationis isperformed
location performedbyby a control
a control device,
device, inin which
which the
the values
values ofofthethe autotransformers
autotransformers
currents,
currents, inin the
the case
case ofofa aground
groundfaultfaultalong
alongthethesection,
section,have havebeen
beenpreviously
previouslyrecorded
recorded ininitsitsmemory
memory
as look-up tables. This control device receives the readings from the autotransformers
as look-up tables. This control device receives the readings from the autotransformers protection relays. protection
Byrelays. By comparing
comparing the measured the measured
currents incurrents in the autotransformers
the autotransformers to therecorded
to the previously previously recorded
values, the
values,
fault the fault
position can beposition can be determined.
determined.
Thismethod
This methodis isalsoalsoeconomical,
economical,asasthe thevoltage
voltageand andcurrent
currentmeasurement
measurementtransformers
transformersand and
protection
protection relays
relays are
are already
already installed
installed inin
thethe autotransformer
autotransformer substations.The
substations. The newnew location
location method
method
could
could bebe implemented
implemented in the
in the Supervisory
Supervisory Control
Control andandDataData Acquisition
Acquisition System System
“SCADA” “SCADA” of 2 ×of252 kV × 25
kV power systems, in which the information from autotransformer protection
power systems, in which the information from autotransformer protection relays is normally available. relays is normally
available.
This methodThis method
has been testedhasbybeen testedsimulations
computer by computer and simulations
experimental and experimental
laboratory tests, laboratory
both obtaining tests,
both obtaining
suitable results. suitable results.

Author Contributions:Jesús
AuthorContributions: JesúsSerrano
Serrano developed
developed the study and and included
includedthe
themathematical
mathematicalformulation
formulationofof
the
the tractionpower
traction powersystems
systemsininthe
the simulated
simulated models.
models. Carlos
Carlos A.
A. Platero
Platero and
and Ricardo
RicardoGranizo
Granizoperformed
performedthethe
laboratory tests and checked the validity of the new method. Máximo López-Toledo revised and improved the
laboratory tests and checked the validity of the new method. Máximo López-Toledo revised and improved the
MATLAB model. All the authors contributed to writing this article, including their conceptual approaches to the
MATLAB
solution model. All the authors contributed to writing this article, including their conceptual approaches to the
obtained.
solution obtained.
Conflicts of Interest: The authors declare no conflict of interest.
Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations
Energies 2017, 10, 340 13 of 14

Abbreviations
The following abbreviations are used in this manuscript:

AC Alternating current.
AT Auto-transformer.
ATS Auto-transformer station.
IA1 Catenary current in autotransformer 1.
IA2 Catenary current in autotransformer 2.
IA3 Catenary current in autotransformer 3.
I Autotransformer number: 1, 2, 3, . . . N.
IAi Current input from catenary in autotransformer “i”.
L Inductive impedance.
L1 Adjustable Inductive impedance 1.
L2 Adjustable Inductive impedance 2.
SATS Autotransformer station at the end of the section.
V1 Voltage source 1.
V2 Voltage source 2.
VCi Catenary voltage in autotransformer “i”.
ZC Catenary self-impedance value per unit length.
ZF Feeder self-impedance value per unit length.
ZR Rail self-impedance value per unit length.
ZCF Mutual impedance Catenary-Feeder value per unit length.
ZCR Mutual impedance Catenary-Rail value per unit length.
ZFR Mutual impedance Feeder-Rail value per unit length.

References
1. Han, Z.; Zhang, Y.; Liu, S.; Gao, S. Modeling and Simulation for Traction Power Supply System of High-Speed
Railway. In Proceedings of the 2011 Asia-Pacific Power and Energy Engineering Conference (APPEEC),
Wuhan, China, 25–28 March 2011; pp. 1–4.
2. Gao, B.; Wei, W.; Zhang, L.; Chen, N.; Wu, Y.; Tang, Y. Differential Protection for an Outgoing Transformer of
Large-Scale Doubly Fed Induction Generator-Based Wind Farms. Energies 2014, 7, 5566–5585. [CrossRef]
3. Knights, D.C. Protection for DC Substations and AC Track Feeder Stations. In Proceedings of the 6th IET
Professional Development Course on Railway Electrification Infrastructure and Systems, London, UK,
3–6 June 2013; pp. 97–107.
4. Xu, G.; Zhou, Y.; Chen, Y. Model-Based Fault Location with Frequency Domain for Power Traction System.
Energies 2013, 6, 3097–3114. [CrossRef]
5. Zhou, Y.; Xu, G.; Chen, Y. Fault Location in Power Electrical Traction Line System. Energies 2012, 5, 5002–5018.
[CrossRef]
6. Richards, S.H. Application Benefits of Modern Microprocessor Distance Protection for AC Electrified
Railways. In Proceedings of the Sixth International Conference on Developments in Power System Protection,
Nottingham, UK, 25–27 March 1997; pp. 338–341.
7. Hewings, D.B.; Chen, J.; Storey, A.P. Application of a New Distance Protection Scheme to a Single-Pole
Autotransformer Electrification System on the UK West Coast Main Line Railway. In Proceedings of the IET
9th International Conference on Developments in Power System Protection, Glasgow, UK, 17–20 March 2008;
pp. 56–61.
8. Agarwal, K.K. Automatic Fault Location and Isolation System for the Electric Traction Overhead Lines.
In Proceedings of the 2002 ASME/IEEE Joint Railroad Conference, Washington, DC, USA, 23–25 April 2002;
pp. 117–122.
9. Sezi, T.; Menter, F.E. Protection Scheme for a New AC Railway Traction Power System. In Proceedings of the
1999 IEEE Transmission and Distribution Conference, New Orleans, LA, USA, 11–16 April 1999; pp. 388–393.
10. Fujie, H.; Miura, A. Fault Location System in Autotransformer Feeding Circuit of AC Electric Railways.
Electr. Eng. Jpn. 1976, 5, 96–105. [CrossRef]
Energies 2017, 10, 340 14 of 14

11. Wang, C.; Yin, X. Comprehensive Revisions on Fault-Location Algorithm Suitable for Dedicated Passenger
Line of High-Speed Electrified Railway. IEEE Trans. Power Deliv. 2012, 27, 2415–2417. [CrossRef]
12. Zhang, D.; Wu, M.; Xia, M. Study on Traveling Wave Fault Location Device for Electric Railway Catenary
System Based on DSP. In Proceedings of the 2009 International Conference on Sustainable Power Generation
and Supply, Nanjing, China, 6–7 April 2009; pp. 1–4.
13. Duan, J.; Xue, Y.; Xu, B.; Sun, B. Study on Technology of Traveling Wave Method Fault Location in Traction
Power Supply System. In Proceedings of the 10th IET International Conference on Developments in Power
System Protection, Manchester, UK, 29 March–1 April 2010; pp. 1–5.
14. Xia, M.; Wu, M. Traveling Wave Fault Location Scheme for Jing-Jin Dedicated Passenger Line Electric Railway
Traction System. In Proceedings of the 2nd International Conference on Power Engineering, Energy and
Electrical Drives, Lisbon, Portugal, 18–20 March 2009; pp. 447–449.
15. Chen, P.; Xu, B.; Li, J. The Optimized Combination of Fault Location Technology Based on Traveling Wave
Principle. In Proceedings of the 2009 Asia-Pacific Power and Energy Engineering Conference (APPEEC),
Wuhan, China, 28–31 March 2009; pp. 1–5.
16. Wei, G. Application of Traveling Wave in Fault Location of Traction Power System. In Proceedings
of the 2014 IEEE Conference and Expo on Transportation Electrification Asia-Pacific, Beijing, China,
31 August–3 September 2014; pp. 1–3.
17. Serrano, J.; Platero, C.A.; López-Toledo, M.; Granizo, R. A Novel Ground Fault Identification Method for
2 × 25 kV Railway Power Supply Systems. Energies 2015, 8, 7020–7039. [CrossRef]
18. Wedepohl, L.M.; Jackson, L. Modified nodal analysis: An essential addition to electrical circuit theory and
analysis. Eng. Sci. Educ. J. 2002, 11, 84–92. [CrossRef]

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).