The Effects of Earthing Strategies On Rail Potentlal and Stray Currents in D.C. Transit Railways

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

303

THE EFFECTS OF EARTHING STRATEGIES ON RAIL POTENTlAL


AND STRAY CURRENTS IN D.C. TRANSIT RAILWAYS

J G Yu

Cegelec Projects Limited, U K

SYNOPSIS When the running rails are used as the


traction current return conductor in a rail transit syst-m,
the control of rail potential and stray current is an
important aspect of system design. The choice of
earthing strategies within the transit system ha!; a
profound influence on rail potential and stray curi'ent
control. The tasks of controlling both rail potential m d
stray current are conflicting, therefore a balance has to be
struck between the two. This calls for a comprehensive
study before a strategy is chosen. Previous research
work at the University of Birmingham in collaboration
with Cegelec Projects Limited [ 1-41 established an
approach to the definition of the performance criterion
for both rail potential and stray current. A computer
program was also developed to simulate the
performances of systems against the defined criterion.
The simulator takes account all relevant system
PiC) (d)
parameters that affect the power supply system. This Figure 1 : Schematics of Different Earthing Schemes
paper describes the follow-on work, following filrther (a). Totally Floating Earth (b). Directly Connected Earth
developments on the simulator, having taken account (c). Diode Earth (d). Floating Earth with RPCD
the latest technological advancements on rail potential
control devices (RPCD). The merits of floating eartt- in Positive Conductor
combination with RPCDs are examined in detail. The
simulation model has been validated against practical :est
results. Sample simulation results for a test track are
presented for analysis.

1. INTRODUClrION
1 Running Rails

In most rail transit systems, with the exception of London


Underground Limited, the running rails are used as the
return conductor for traction current!;. This arrangement
has the distinctively economic advantage in that no
dedicated return conductor is required. The disadvanta,:es
associated with such an arrangement are rail potential and
stray current problems:
(b) Rail Potential For a Floating Earth

Rail potential rise or fall may be hazardous above "1x1


certain thresholds, in the forms of touch and/or step t I

- voltages.
Stray current may cause or accelerate electrochemical
corrosion to metallic structures in the vicinity of the
transit system.

There are many rncasures that havc bccn adopted in the (C) Rail PGlenlial FGr a Direclly Connected Earlh or a Dlode Earth
attempt to mitigate the harmful effects of the two
problems.
Figure 2: The Effect of Earthing Arrangement
on Rail Potential

International Conference on Developments in Mass Transit Systems, 20 - 23 April 1998, Conference Publication No. 543 OlEE 1998

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY KANPUR. Downloaded on October 03,2020 at 03:28:59 UTC from IEEE Xplore. Restrictions apply.
304

Among them the choice of the overall earthing strategy the measured voltage exceeds a pre-set voltage-time
has a profound influence on the performance of the characteristic: the thyristor gates will be pulsed and the
system. contactor will be instructed to close. The appropriate
thyristor will conduct, depending on the polarity of the
Broadly the earthing strategy can be classified as follows: voltage across the thyristors. This will then be followed
by the action of the contactor. The contactor takes longer
0 Totally floating earth (with no intentional connection to switch ON than the thyristor due to its inherent and
between the rail and earth) intentionally built-in delays. This is desirable since the
Directly connected earth contactor rating is usually chosen to be less than the
Diodeearth thyristors.
0 Floating earth with RPCDs (RPCD earth)
Once the contactor is switched ON, the voltage across the
These are illustrated in Figure 1. Figure 2(a) shows a thyristors will collapse and the conducting thyristor will
feeding situation with one substation and one train only. It be switched OFF since the conduction current is
is assumed that the rail to earth resistance is transferred to the contactor.
homogeneous. Figure 2(b) shows the rail potential when
there is no intentional earth in the system, while Figure The contactor will remain ON until the current through
2(c) shows the rail potential profile when the rail is the RPCD falls below a pre-set current threshold gnJ its
connected to the substationearth. The rail to earth leakage ON time is longer than a pre-set minimum ON time. The
current density follows the rail potential profile. thyristor will not be pulsed for the next switch ON cycle
until the contactor is in the fully open position.
2. RAIL POTENTIAL CONTROL DEVICES
In some applications, the contactor does not have a
The RPCDs can be in many forms, while the modern required minimum ON time. The contactor will be
practices use thyristor controlled switches. switched OFF once the current through it falls below the
pre-set current threshold.
In the Ankara Rapid Transit System of Turkey, the device
is called Floating Negative Automatic Ground Switches 3. SIMULATION MODEL
(or FNAGS). In the Lantau & Airport Railways of Hong
Kong, it is called Over Voltage Protection Devices (or A computer simulation model has been developed to
OVPD). The two devices are similar in that both are simulate the rail potential and stray current for d.c. transit
thyristor controlled, bi-directional switches. However, the systems. The methodologies are explained in [3,4].
design and operating details differ in certain aspects. Briefly the input and output of the simulation are as
Figure 3 shows a schematic diagram of such a device. follows:
These devices have been installed in the respective
systems and tested for their functions in the integrated The basic data input starts with track profiles, including
systems. gradients, curvatures, speed limits, substatiodtrack
parallel hut locations, passenger station locations, train
characteristics, train headways, electrical power network
Negative Bus
paramctcrs including earthing data, etc. The results are
given for train performance including minimum voltages,
train speeds and journey times, electrical plant load
cycles, etc.

To obtain results for rail potential and stray current, the


distributed rail-earth circuit is represented by lumped
circuit parameters. A large amount of information is made
available for analysis. This includes rail potential against
distance at any particular time of simulation (figurcs 5 &
& 6) , rail potential against time at any particular chosen
location (Figure 9). Similar output is made for rail to
earth current (Figures 10 & 11). The following
parameters are defined to assess the performance of a
system in terms of rail potential and stray current.
Figure 3: Schematic of a Rail Potential Control Device
3.1 Rail Potential Parameters
Refer to Figure 3. Where a RPCD unit is inbtalled, the
voltage between the rail and the local earth is
continuously monitored. The device will be activated if

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY KANPUR. Downloaded on October 03,2020 at 03:28:59 UTC from IEEE Xplore. Restrictions apply.
305

At all instants of time during the simulation (usually at 1


second’s interval), the positive maximum potential and (4)
the negative maximum potential of the rails are found and
their locations recorded. The overall maximums are also
recorded. These give an indication of what touch voltages 3.3 RPCD Model
can be expected in the system [4].
For the rail potential control devices, 3 setting parameters
Over the simulatedperiod of time, the maximum positive are modelled:
and negative potential can be plotted against time, as
shown in Figure 7. Voltage Setting - This is the voltage threshold to
switch the RPCD ON. If the measured voltage is
3.2 Stray Current Parameters greater than the voltage setting, the device will be
switchedON.
Faraday’s Law states that the amount of metal dissolved Current Setting - This is the current threshold to
in a corrosive reaction is proportional to the electrical switch the RPCD OFF. If the measured current
charge that causes the reaction. The performance of a through the RPCD falls below the current setting, the
transit system can therefore be measured by the g-oss RPCD will be switched OFF, provided the time
leakage charge that is emitted from the transit system. setting is satisfied as well.
The definitionsare as follows [ 3 ] . Time Setting - This is the minimum required time for
the RPCD to be in the ON state before it is switched
Total Stray Current This is defined as the total curpent OFF. If the device has conducted for a period up to
that is leaked to the earth. the time setting, it will be switched OFF, provided the
current setting is satisfied as well. Where the time
setting is not required, it can be set to zero.

k=l practice, all three setting parameters are built with


certain ranges specified by the users, so that the users
where it,,(t) is the total stray current at timet, irc(t,k)is the have the flexibility to fine tune the settings to suit their
rail to earth current at electrical node k and at time t, N(t) particular systems.
is the total number of electrical nodes at time t. A plot of
total stray current vs. time is shown in Figure 8. The diode is modelled as an ideal switch.

Gross Leakage Charge The gross leakage charge Q, (in 4. SIMULATION RESULTS AND ANALYSIS
Coulombs) is obtained by integrating the total slray
current itsc(t) against time for the simulated duration of A test line is used to simulate the effects of earthing
time (t,,tz)(in seconds). strategies on rail potential and stray currents. This is a
double track line with a route length of 18km and 16
passenger stations. See Figure 4 for the gradient profile of
the uproad track, simulated train speed, voltage and
current.

Usually the simulated duration varies for different studies, The line is powered by 7 traction substations at a nominal
depending on the train headways. To make a meaningful voltage of 15OOV d.c. The traction substations are located
comparison, Q, is normalised to one hour’s equivalent Qh near selected passenger stations. Each train consists in 6
(also in Coulombs): cars with induction motor drives and regenerative
braking. A headway of 120 seconds is used.

A number of cases are studied and the results are


summarised in Table 1. Graphical output plots are shown
in figures 5 to 1 1.
Qhcan also be quoted in the unit of Ampere-hours.
For each case, it is assumed that each earth position has a
Mean Total Stray Current Dividing the gross leakage resistance of 0.5 Q and the rail to earth leakage resistivity
charge by the simulated duration of time, the mean total is at 50 Q-km . For cases A and B, it is assumed that the
stray current (in Amperes) during the simulated load rails are floating freely. For cases C to E, 7 earths are in
cycle is obtained: place, one at each substation. For Case F, two extra
RPCD earths are added: one at Stop-2 and the other at
Stop-14. For cases G to M, 16 RPCDs are installed, one at
each passenger stationisubstation.

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY KANPUR. Downloaded on October 03,2020 at 03:28:59 UTC from IEEE Xplore. Restrictions apply.
Table 1 - Summary of Simulated Results

Cases Scenarios Overall Max. Overall Min. Qh


Potential (V) Potential (V) (Ah)
A totally floating earths, with 10.25 mR/km rail resistance 67.2 -72.0 3.3
B totally floating earths, with 7.50 mRikm rail resistance 54.2 -54.4 2.5
C Case “A” + 7 RPCD earths. settings at 60V,50A, 1s 106.2 -72.0 3.4

L Case “A” + 16 diode earths 115.0 -11.2 11.3


M Case “A” + 16 directly connected earths 61.3 -61.5 268.2
Note Qh = 1 hour’s equivalent Gross Leakage Charge (in Ampere-hours).

TRACK GRADIENT PROFILE

no ITS*‘

LOO
STOP-2 STOP4 STOPS ST0P-S STOP 10 STOP 12 ASTOP-14 STOP 18
STOP 1 STOP 3 STOP 5 STOP 7 STOP 9 STOP 1 1 STOP 13 STOP 15

T W N SPEED AND SPEED LIMIT PROFILE __--5Peed lm


-Tran %p4

tm

€a

Eo
Wh
43

20

00 10
I I
20 30
I
40 50
I 1 1 1 1
50 70 80 90 100
I 110
l 120
l 130
I 140
I 1
150 160 170
I 180 190
astance (h)

T W N CURRENT ANDVOLTAGE
M iA1

m
3511

Figure 4: Test Track Simulation Results for Train Performance

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY KANPUR. Downloaded on October 03,2020 at 03:28:59 UTC from IEEE Xplore. Restrictions apply.
307

occurred at 54 second. In this particular case, the


maximum and minimum occurred near passenger stop-&.

The maximum occurred when a train accelerated to leave


the stop, while the minimum occurred when a train
braked regenerativelyto approach the stop.

Figure 6 shows the rail potential snapshots at 80 second


for 3 cases (A, D & F). This illustrates how the rail
potential varies when the earthing conditions are changed.
Figure 7 shows the maximum rail potential vs. time for 3
cases (A, D & F). Figure 8 shows the total stray current
vs. time for 3 cases (A, D & F).

Figure 9 shows the rail potential vs. time at passenger


stop-2 for 2 cases (D & F). This illustrates the necessity to
install rail potential control devices at such locations.

Figure 10 shows the earth to rail current at passenger


stop-2 for 2 cases (H & I). This illustrates how the earth
to rail current is affected by RPCD current settings.

Figure 1 1 shows the earth to rail current at passenger


stop-2 for 2 cases (J & K). This figure, in conjunction
with Figure 10, illustrates how the earth to rail current is
affected by RPCD voltage settings and time settings.

Observations made from these simulated results are


discussed as follows.

Rail Resistances Both cases A and B have totally


floating earths with different rail resistances. The
I

simulation results indicate that the rail resistance has a


for Cases A, D iF fundamental role in rail potential and stray current
control.

All other conditions being equal, different earthing


strategies have a significant effect on rail potential and
stray current control. In the cases studied, Case A is
the base case. Cases C to M are based on Case A, with
variations on earthing arrangement, as identified in
Table 1.

Totally Floating Earth The totally floating earth


comes out as the best option in terms of stray current
control, as shown in cases A and B. All forms of
intentional earthing, such as directly connected earths,
diode earths and earths with RPCDs, increase the
gross leakage charge.
-
for Cases A, D 8( F Directly Connected Earths The directly connected
earth schemes result in very high gross leakage
Figure 5 shows the rail potential snapshots at two charges, as illustrated in cases E and M.
different instants of time for Case J. This illustrates how
the rail potential changes when the loading conditions Diode Earths In the cases studied, t h e diode earths
have changed. The overall maximum potential occurr-d are not acceptable since the rail potential cannot be
at 114 second and the overall minimum potential effectivelycontrolled, as shown in cases D and L.

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY KANPUR. Downloaded on October 03,2020 at 03:28:59 UTC from IEEE Xplore. Restrictions apply.
308

Diode earths are inflexible since the diodes are not for cases H and K are shown in figures 10
controllable. Besides, diodes are unidirectional and 1 1. Therefore it is recommended that the
devices and are ineffective in reducing positive rail voltage settings should be made as high as is
potential. This makes it impossible to control the rail acceptable.
potential at locations other than at traction substations.

-
for Cases A. D & F Stop-2 for Cases H & I

Figure
for Cases D & F Stop-2 for Cases J & K

RPCD Earths A floating earth scheme with rail + Current Settings Within the RPCD
potential control devices is an effective scheme for schemes, a higher current setting means that
rail potential control. The RPCDs have the flexibility the RPCD already switched on can be
to make the system totally floating by setting the switched off sooner, therefore reducing the
voltage threshold high enough. (Case A is equivalent gross leakage charge. The disadvantage of
to 16 RPCD earths with voltage settings at 75V or the higher current setting is the heavier duty
higher. Case B is equivalent to 16 RPCD earths with on the contactor. This is compared between
voltage settings at 55V or higher.) The gross leakage cases H and I, shown in Figure 10. The
charge depends on the three settings chosen. results show that it is worthwhile to have a
heavier duty contactor in exchange of a
+ Voltage Settings Within the RPCD schemes, lower gross leakage charge.
a higher voltage setting reduces the
occurrences of the RPCD being switched on + Time Settings Within the RPCD schemes, a
and reduces the gross leakage charge. This shorter time setting means that the RPCD
is compared between cases G, H and K. Plots already switched on can be switched off

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY KANPUR. Downloaded on October 03,2020 at 03:28:59 UTC from IEEE Xplore. Restrictions apply.
309

sooner, which results in a lower gross


leakage charge. The disadvantage of the 0 Once the decision has been made on the overall
shorter time setting is the incrcased system earthing strategy, the simulation results can
frequency of RPCD SM. itching. which rt:suIts be used to determine device locations and ratings. In
in more wear and tear for the devices. This is the case o f rail potential control devices, optimum
compared between cases H and J , shown in settings can also be derived through simulation
figures 10 and I 1. studies.

0 Passenger Stations Since the RPCDs are bi- The simulation model has been verified against system
directional devices, they can be installed in locations integration test results from the Phase-I track of the
other than traction substations, such as passenger Ankara Rapid Transit System in Turkey, where rail
stations, where high rail potential usually occurs due potential control devices are used. Excellent agreement
to train acceleration. has been found between simulation and test results.

Case F represents two extra RPCDs at passenger The simulation results for a test track demonstrate that
stations that are located in the two longest feeding the floating earth scheme with rail potential control
sections. The results show that the rail potential is devices is the best choice.
reduced substantially compared with Case C. This
means that these two extra RPCDs are adequate for 6. ACKNOWLEDGEMENT
the system under normal operational conditions. They
may not be adequate under substation outage The author would like to thank the directors of Cegelec
conditions which are not studied here. For a passenger Projects Limited for permission to publish this paper. The
station located within a relatively short feeding author also acknowledges the assistance from his
section, a RPCD installation may become a necessity colleagues for the preparation of the paper, particularly
if one of the substations feeding the section is oLt of the comments by Mr K Loring, and the discussions on
service. the rail potential control devices with Messrs S Forde
and N Groves.
Compared with Case F, the results in Case H indizate 7. REFERENCES
that more installations of RPCDs increase both rail
potential and gross leakage charge. However, this can [ I J Yu,J.G. & Goodman,C.J. "Modelling of Railway
be avoided if the voltage settings at the added Potential Rise and Leakage Current in DC Supplied
installations are made higher, so that Case H becomes Railway Systems", IEE Colloquium Digest
equivalent to Case F, under normal feecling No. 19901129. Presented to IEE Colloquium on Stray
arrangement. The added installations can be made Current Effects o f DC Railways and Tramways.
active in emergency feeding conditions only. London, October 1990.

5. CONCLUSIONS [Z] Yu,J.G. & Goodman,C.J. "Computer Simulation of


Stray Currents in DC Supplied Rail Transit Systems
A computer simulation model has been developed and and Their Corrosive Effects", Proceedings of the
sample simulation results are presented and analysed. IMechE International Conference on Transit 2020,
The model is essential for effective system design pp.121-127,London, October 1990.
concerning rail potential and stray current control, in
addition to its normal function o f assessing ot'ler Yu,J.G. & Goodman,C.J. "Stray Current Design
aspects of system performance and designs. Parameters for DC Railways", Proceedings of the
ASMEiIEEE Joint Railroad Conference, pp. 19-28,
The model can simulate different earthing schemes, Atlanta, USA, March,l992.
including diode earths and floating earths with 1,ail
potential control devices. These can be used for tiyo Yu,J.G. & Goodmar1,C.J. "Computer Analysis of
aspects of system design: Touch and Step Voltages for DC Railways",
Proceedings of the Third International Conference on
Computer Aided Design, Manufacture and Operation
0 The performance of the system in terms of rail
in the Railway and Other Advanced Mass Transit
potential and stray current can be assessed through
Systems, Washington DC, USA, August 1992.
simulation studies. This, together with other systrm
performance indicators, can be used to determine the
overall system configuration (e.g., system voltage
level, maximum feeding section length, maximLlm
rail resistance, etc.) and the overall system earthi 7g
strategy;

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY KANPUR. Downloaded on October 03,2020 at 03:28:59 UTC from IEEE Xplore. Restrictions apply.

You might also like