Stray Current Calculation

and Monitoring in
DC Mass-Transit Systems
Interpreting Calculations for Real-Life Conditions
and Determining Appropriate Safety Margins

T
Charalambos A. Charalambous,

ING
his article delivers useful practical contempla-

LISH
Pete Aylott, and David Buxton tion of stray current calculation and monitoring

PUB
endeavors in dc mass-transit systems. We

RAM
focus on interpreting stray current calcula-

ING
BY
tions—carried out at the design stage for real-life

ED
conditions—and on determining safety margins to
ENS
LIC
cope with calculations following oversimplifying
CKS

assumptions. We also discuss the general speci-

TRA

fications and benefits of the direct stray cur-

AIN
, TR

rent monitoring method, through addressing

M/chaluk

the implications that arise from implement-

ing the alternative rail potential monitoring
.CO

method informatively quoted in European

OTO
KPH

(EN) Standards.
TOC
©IS

Introduction
TS
BOL

Over the past few decades, the stray

IC
CTR

current modeling and monitoring

ELE

endeavors in dc mass-transit sys-

CK,

tems can be broadly summarized

STO
PHIC

as follows:
GRA

■■ Existing railway stray current

BY

model applications have the

ED
ENS

ability to compute rail voltage

LIC

to remote earth and current

AGE

flow [1], [2] in the modeled

D IM
OUN

components under various

KGR

scenarios, depending on
BAC

their design and level of

LUE

Digital Object Identifier 10.1109/MVT.2015.2477419

complexity [3], [4].
©B

Date of publication: 31 March 2016

24 ||| 1556-6072/16©2016ieee IEEE vehicular technology magazine | june 2016

Authorized licensed use limited to: Mahidol University provided by UniNet. Downloaded on July 26,2023 at 04:35:51 UTC from IEEE Xplore. Restrictions apply.
 ■■ Assessments of the corrosion impacts are made as qualita-
tive assessments using a mix of engineering judgment and
simple spreadsheet applications of Faraday’s laws to
assess the cumulative mass of metal loss over the target
operating period [5].
■■ Application of Faraday’s law requires consideration of cur-
rent flows, whereas the most common site of validation
measurement is corrosion potentials in railway system
structures and utility assets to a local reference [6], [7].
■■ Railway stray current flows are time-variable in
response to timetabled train operations and bidirection-
al as a function of dynamic changes in multiple train
positions and train regeneration characteristics [8], [9].
■■ The impacts measured on affected structures and ser-
vices present the net effect from these variable factors.
■■ Current impact assessment techniques are limited to
simple time averaging and linear extrapolation of
current flows from either static or dynamic model
outputs [10].
■■ By contrast, current standards (e.g., EN 50122-2 [11] and
EN 50162 [12]) apply criteria based on exceedance of
absolute or averaged corrosion potential thresholds with-
out regard to current flows.
In particular, EN 50122-2 specifies requirements for protec-
tive provisions against the effects of stray currents, which
result from the operation of dc traction systems. This
applies to all metallic fixed installations, which form part of
the traction system, and also to any other metallic compo-
nents located in any position in the earth, which can carry
stray currents, resulting from the operation of the railway
system. To this end, EN 50162:2004 completes EN 50122-2
by establishing the general principles to be adopted for
minimizing the effects of dc stray current corrosion on
buried or immersed metal structures.

Stray Current Calculations

Interpreting Stray Current Magnitude

Stray current magnitude depends on the traction
current, the rail resistance, and the resistance to
the earth value. Applying the equation in the EN
51022-2:2010 Annex C to a data set with a maxi-
mum expected traction current of 2,000 A that
is equally returning through a 500-m section
of two rails (i.e., 1,000 A on each side) with
a resistance of 40 mW/km of rail and a
resistance to earth of 100 W/km results in
a I stray of 50 mA.
However, care should be taken
when interpreting any simulation

JUNE 2016 | IEEE vehicular technology magazine ||| 25

Authorized licensed use limited to: Mahidol University provided by UniNet. Downloaded on July 26,2023 at 04:35:51 UTC from IEEE Xplore. Restrictions apply.
 Traction
n Traction
T Table 1 The stray current magnitude scaling
Substation
nA 2,000 A Substation
Sub B for real-life conditions.

Calculated Values
V ac 1,000 A
34.5-kV 1,000 A 34.5
34.5-kV ac
Stray Current Stray Current
Running Rails with Single Train with Single Train
Rail-to-Earth Drawing 2,000 A Drawing 4,000 A
2.0 Resistance (X km) 500 m from Each at 500 m from
Stray Current Pro le (mA)

1.5 Single Track Substation Each Substation

1.0 EN 50122-2: 2.52 A 5.04 A
50 mA
0.5 permitted
0.0 minimum: 2 X km
–0.5 Design recommen- 50.1 mA 100.63 mA
–1.0 dation: 100 X km
–1.5 Service operation: 125.34 mA 250.21 mA
–2.0 40 X km
0 200 400 600 800 1,000
Distance Between Traction Substations (m)

Figure 1 The stray current profile under the worst static conditions
in a floating system. the modeling results. A variation in track current will
influence the leakage current distribution due to the
resulting alteration of rail-to-earth potential. A doubling in
outputs. For example, Figure 1 illustrates the stray current track current will lead to a doubling in voltage and,
profile simulated (under a resistive type model [1]) along the hence, doubling of leakage current density along the rail.
same 1-km section of floating track with two substations at This is a linear effect.
remote ends, supplying a train with 2,000 A at the midpoint. The rail-to-earth potential is determined by the cur-
This profile constitutes a snapshot of the worst distribu- rents flowing through the rail. Unless there is significant
tion stray current along the length of the rail (0–1,000 m). In leakage that causes a change in the rail current flow, the
the floating system, modeled by means of an example, this leakage current density is proportional to the resistance
profile will appear on the rails as +2 mA near the train and of the rail insulation. Variations in the rail-to-earth resis-
-2 mA near the two substations. A positive figure implies a tance can, therefore, be taken to have a linear relation-
current leaking out of a conductor by corrosion, and a neg- ship with the leakage current density.
ative figure implies a current leaking into a conductor. At The 50-mA stray current value in Figure 1 may be low-
250-m down the track, the voltage to remote earth will be er when compared to measured numbers on some sys-
0 V, thus no leakage current activity occurs. Although the tems; however, it is realistic at rail-to-earth resistance
maximum stray current under the static condition simu- levels of 100 X km, which is driven by design and con-
lated is about 2 mA at 500 m along the rail (see Figure 1), struction targets. To this end, it is noted that the whole
the sum of total stray current leaving the rails, between track system is usually planned, installed, and main-
250 and 750 m, is 50 mA. tained to ensure particular insulation levels (e.g., 40 or
The principles previously described for interpreting 10 X km) can be sustained under operational conditions.
the stray current simulation outputs equally hold for However, industrial practice suggests using a value of
diode-bonded systems, albeit for their intrinsic char- insulation level higher than the aforementioned values
acteristics. In such systems, the diodes can either be in as a benchmark for the design—that is, to ensure that
turn-on or turn-off status. When the rail is at negative when the track is new, clean, and dry (i.e., just after the
potential with respect to the earth, the system is floating installation process), it has a typical value of 100 X km.
(i.e., the diode is turned off). The diode, however, will If, however, the rail-to-earth resistance is pushed to
appear as a short circuit (i.e., the diode is turned on) 2 X km, (specified as minimum in EN 51022-2) or any oth-
when the rail potential moves positive with respect to er value lower than 100 X km, then the numbers from the
the earth. The general effect is to increase the stray cur- model would come out significantly higher (see Table 1).
rent level because it holds the negative potential of the The linear scaling rule does not apply when adjusting
rail at or near earth potential and raises the peak rail the expected maximum stray current to account for vary-
voltage, with respect to the floating mode of the system. ing distances between two supplying traction substations.
To this end, we note the total stray current leaking from a
Applying Scaling Factors for Real-Life Conditions floating system can be conveniently described using
A determination of the stray current levels under real-life 2
conditions can be achieved by applying scaling factors to I stray = I $ rt $ l , (1)
8rc

26 ||| IEEE vehicular technology magazine | JUNE 2016

Authorized licensed use limited to: Mahidol University provided by UniNet. Downloaded on July 26,2023 at 04:35:51 UTC from IEEE Xplore. Restrictions apply.
 Stray current performance monitoring
is a recommended practice in EN standards
covering dc railways.

where I is the traction current in amperes, rt is the resis- especially once the system is in operation, it is more
tance of the track (i.e., two parallel rails) in ohms per kilo- reasonable to assume that there will be a concentra-
meter, l is the distance between the train and the two tion of stray current return to the rail over a possibly
supplying substations in kilometers when the train is at short distance. To apply a more realistic case, a range
the midpoint, and rc is the resistance to the earth of the of scenarios can be investigated based on a more cred-
tracks. Thus, doubling the distance between any two sup- ible method for calculating the average stray current
plying substations, ^ l h will result in increasing the stray per length of a single-track line, illustrated in the fol-
current level ^ I stray h by a factor of four. The concept is lowing equation:
clearly illustrated in Figure 2, where the maximum stray
p # Is
current level expected is interpolated for different equal J= , (2)
d
distances between the train and the supplying traction
substations (i.e., the train sits at midpoint). where p is the percentage (%) of the stray current that will
return to the rails within a specified shorter length d and
Interpreting Stray Current Leakage Density I s is the total stray current flow from the rails. If, for exam-
EN 50122-02:2010 states that there is no damage in the ple, the total calculated stray current flow is 200 mA (1-km
tracks over a period of 25 years, if the average stray cur- section), the worst case stray current leakage under a
rent ^ I max h per unit length, that is, current leakage density, design level of 50 X km into the track bed concrete would
does not exceed 2.5 mA/m. In particular, Annex C of the be 2 mA/m when it is assumed that 30% of the total stray
EN 50122-02:2010 specifically states that the 2.5 mA/m is a current flow will have a concentrated return to the rails
conservative figure based on simplifying assumptions within 3% of the rail length (see Figure 3).
and prompts for more detailed investigation. Similarly, when it is assumed that 30% of the total
In relevant modeling endeavors identified in the liter- stray current flow will have a concentrated return to
ature, as well as the method described in EN 50122-02, the rails within a 30% section of the rail length, then
the stray current density calculation assumes that the the stray current leakage would be at 0.25 mA/m. Fig-
rail-to-earth resistance at each baseplate along a sec- ure 3 provides the results for two more scenarios that
tion of track is uniform, and it provides a combined ef-
fective resistance to the earth (e.g., 100 X km) for the
entire traction section considered. However, this is an
500
Total Maximum Stray Current (mA)

improbable condition and the real resistance is more

450
likely to result from a small proportion of baseplates
400
with a lower than expected resistance. Furthermore,
350
the longitudinal rail resistance is also uniform—i.e.,
300
there is even wear along the length of the rail and the
250
welded joints are all of uniformly low resistance. It is
200
important to note that variation in longitudinal rail
150
resistance may also arise due to track bonding issues
100
(stolen bonds and degraded connections. The method
50
described in EN 50122-02 also assumes the track bed
0
reinforcement and concrete, together with the tunnel 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5
systems, are also electrically uniform and that the ex- Equal Distance of Train from Two Substations (km)
ternal soil environment is uniform.
Given that this degree of system uniformity across Figure 2 The stray current level variation for different substations
all of the infrastructure components is highly unlikely, locations in the floating system.

JUNE 2016 | IEEE vehicular technology magazine ||| 27

Authorized licensed use limited to: Mahidol University provided by UniNet. Downloaded on July 26,2023 at 04:35:51 UTC from IEEE Xplore. Restrictions apply.
 To acceptably interpret any corrosion
alarms under the rail potential monitoring
method, a mix of engineering judgment
and experience is required.

investigate smaller percentages (e.g., 10% and 20%) of or the voltage against the earth resulting from train
stray currents’ concentrated returns to the rails. Thus, operation (see Figure 4).
the method shown in (2) may facilitate stray current con- These rail potential measurements are providing
trol designers to determine safety margins to credibly information to the systems’ operators and owners to
test a number of configurations of their designs as well restore their systems back into line with a reference
as service conditions. condition. Thus, such endeavors do not measure the
effects of stray current but they merely concentrate on
Stray Current Monitoring Methods its source. The drawback of this endeavor lies in iden-
Stray current performance monitoring is a recommend- tifying an appropriate reference condition that will
ed requirement in EN standards covering dc railway sys- serve as a healthy condition-benchmarking metric.
tems. The performance monitoring is utilized to enable The reference condition should be able to account for
proactive maintenance. It merely relies on the system’s elements that are semideterministic but also varying.
condition data and ensures continued stray current con- These elements include scheduled daily or seasonal
trol over railways and third-party infrastructure. traffic trends, occasional traffic peaks (e.g., a major
sporting event), weather/environmental conditions,
Rail Potential Monitoring Method rail insulation condition, faults, and track pollution.
A basic specification for continuous monitoring is given Most importantly, the reference healthy condition
in EN 50122-2 as an informative (not normative) should be defined once all third-party measurement
Annex  B, and on this a number of commercial systems issues have been resolved and should be under oc-
have been developed and marketed. The philosophy of casional reassessment. Therefore, to acceptably in-
these systems is that direct measurement of stray cur- terpret any arising alarms under the rail potential
rents is difficult; therefore, they are based on measure- monitoring method, a mix of engineering judgment
ments of the resistance of the return circuit to the earth and experience is required.

10% Current 20% Current 30% Current

2.5 Positive Bus Bar Overhead
Stray Current Density (mA/m)

Contact System
2

1.5 Running
Rail
1

0.5
Stray Current Collection
0 System
3 6 9 12 15 18 21 24 27 30 V
Track Section for Current Return (%)

Figure 3 The leakage current density (amperes per meter length of

the rail) assuming concentrations of stray current return to the rail
over shorter distances. Figure 4 The rail potential monitoring method.

28 ||| IEEE vehicular technology magazine | JUNE 2016

Authorized licensed use limited to: Mahidol University provided by UniNet. Downloaded on July 26,2023 at 04:35:51 UTC from IEEE Xplore. Restrictions apply.
 To partially lift the uncertainties
associated with the rail potential
monitoring method, a direct stray current
monitoring method can be deployed.

Direct Stray Current Monitoring Method criterion can be converted to a maximum stray cur-
To partially lift the uncertainties associated with the rail rent value for each section, given the known stray
potential monitoring method, a direct stray current moni- current grid resistance and traction earth cable resis-
toring method can be deployed. The objective of the tance. The latter value is used for performance mea-
stray current monitoring method is to determine the per- surement purposes.
formance of the package of protective measures used to ■■ The tunnel reinforcement potential shift is subject to
control stray current, measure the impact of stray current a maximum limit of +0.2 V (EN 50162:2004, Table 1)
on the corrosion of the system structures, and allow the which EN 51022-2:2010 interprets as “the average
location of stray current faults to be determined. value in the hour of highest traffic.” A normal indus-
The reference specification of a stray current moni- trial practice is to measure this value as a corrosion
toring method for a tunnel metro system can be summa- potential using an embedded sensor.
rized as follows. ■■ Taken together, these two parameters (longitudinal
■■ In each rectifier substation of the traction system, a 0.1-V limit and potential shift subject to a maximum
wall-mounted stray current cabinet is usually provid- limit of +0.2 V) will allow quantification of the stray
ed. This cabinet is equipped with a sufficient number current magnitude and direction at the measurement
of suitable size terminals to terminate the necessary location and confirm whether the metro system is
cables carrying the potential of specific parts of the exporting and importing traction stray current
traction earth and structural earth systems accord- through the tunnel walls to and from the outside
ing to the stray current monitoring design. After com- environment. This will both quantify the corrosion
pletion of construction and putting into operation of threat to the tunnel reinforcement and the risk of
the tunnel stretch, measurements have to be carried stray current corrosion to external pipes and servic-
out via the stray current cabinet to check if the maxi- es. The tunnel-wall measurements will also allow
mum allowable value of 0.1 V for the longitudinal volt- detection of imported stray current from outside sys-
age drop, caused by operation in the tunnel, is tems, such as pipeline cathodic protection systems.
not exceeded. ■■ To achieve this, a network of current and tunnel
■■ EN 50122 applies voltage limits in two ways: 1) longi- corrosion sensors can be applied to locations
tudinal voltage drop in tunnel reinforcement and 2) across the metro system. Data acquisition units are
structure to the earth potential shifts in tunnel rein- then installed at each sensor and the digital output
forcement. Stray current designs assume that the lon- is transmitted to the stray current cabinets at each
gitudinal 0.1 V limit is applied over individual section rectifier substation. The number and position of
lengths of the stray current grid (collection system), sensors are usually determined during detail
where a length is defined as being between two trac- design to allow the operators to locate potentially
tion substations or two dielectric joints should the dangerous track insulation failures. The design pro-
system be segregated, accordingly. The remote mea- cess takes account of the traction power and stray
surement of the end-to-end voltage drop, which current control designs, as well as the distribution
would require sense cables to be routed from each of the different tunnel and station construction
end to the nearest stray current cabinet, is not rec- across the system and significant interfaces with
ommended as the sense cables would be susceptible external systems and services. As a minimum, cur-
to electromagnetic interference. The presence of the rent and tunnel corrosion sensors are located at
stray current grid allows direct measurement of the each metro station (one pair per track) and in the
stray current flow, either in the reinforcing steel or in tunnels at midpoints between stations (one pair
the traction earth cable using local sensors. The 0.1-V per tunnel).

JUNE 2016 | IEEE vehicular technology magazine ||| 29

Authorized licensed use limited to: Mahidol University provided by UniNet. Downloaded on July 26,2023 at 04:35:51 UTC from IEEE Xplore. Restrictions apply.
 Direct Stray Current Monitoring Method Benefits suitable minimum configuration would be a six-element
The direct measurement of stray current in the stray cur- sensor comprising a carbon-steel sense electrode, a car-
rent grid offers a more direct correlation with rail insula- bon-steel coupon electrode, an auxiliary or counter elec-
tion performance than is achievable from either trode, a reference element or electrode, a reinforcement
rail-to-earth voltage measurements or longitudinal volt- connection, and a temperature sensor. The carbon-steel
age drop measurements. The use of the data at the oper- electrode used for measurement can be independent of
ation control center (OCC) can follow the approach that used for measurement.
defined in EN 50122-2 Annex B, with the rail-to-earth volt-
age measurement replaced by direct stray current mea- Measurement Module
surement. Measurement of current flow will allow A local data measurement and acquisition unit (DAU)
estimation of stray current corrosion to the stray current can be provided at each sensor location to perform
grid components as current flows out of the grid to the required measurements both on demand from the
return to the rail. control system and to preconfigured schedules. The
The inclusion of tunnel corrosion sensors will reduce DAU is also used to convert the results into a digital
the scope for external baseline and operational surveys format. The DAU is usually capable of bidirectional
as any stray current reaching external services will have communication with its corresponding local control
first to traverse the tunnel wall. These measurements unit (LCU) at the rectifier substation stray current
will give a direct measure of corrosion rates in the tun- cabinet. Power for the DAU can be taken locally within
nel reinforcement and provide a contextual reference to the tunnel.
the stray current grid measurements allowing key per-
formance index (KPI) criteria and performance visual- Stray Current Grid and RS Stray
izations to be defined. Current Cabinet Measurements

Tunnel Corrosion Monitoring Stray Current Grid Measurement

Measurements can be made from current in the stray
Measurement Sensors current grid system using a shunt resistor, installed as
The tunnel reinforcement to the earth potential can be part of the transverse bond connection, or from currents
measured using sensors fixed to the reinforcement cage within the traction earth cable using a suitable Hall-effect
prior to concrete pour. Each sensor can be of the same type sensor, or other suitable means (see Figure 5). In
type and configuration such that it can be used to make each case, the measurement sensor and equipment
a series of stray current and corrosion related measure- should be capable of withstanding over current during
ments (see Table 2). short-circuit fault conditions on the rail system.
The sensors can provide sufficient components to
allow the required measurements to be undertaken. A RS Stray Current Cabinet Grid Measurement
The RS local measurements depend on existing mea-
surement requirements from traction power design.
These could include rail-to-earth voltage measurement
Table 2 The types of sensors and monitoring methods. and traction return current measurement.
ID Name Method
1 Tunnel reinforcement Corrosion potential
potential measurement between
reinforcement and integral
reference element
2 Tunnel-wall stray Current flow between rein- Positive Bus
Bar Overhead
current forcement and an integral
carbon-steel element Contact System

3 Tunnel reinforcement Corrosion rate using linear

corrosion rate polarization or galvanostatic
pulse method of the Running
reinforcement Rail
4 Tunnel coupon Corrosion rate using linear
corrosion rate polarization or galvanostatic Hall Type
pulse method of an integral Sensor
carbon-steel coupon element Stray Current Collection
5 Temperature Temperature within the System
tunnel-wall concrete using
an appropriate sensor
Figure 5 The direct stray current monitoring method.

30 ||| IEEE vehicular technology magazine | JUNE 2016

Authorized licensed use limited to: Mahidol University provided by UniNet. Downloaded on July 26,2023 at 04:35:51 UTC from IEEE Xplore. Restrictions apply.
 Local Control Units Author Information
An LCU is provided at each stray current cabinet to Charalambos A. Charalambous ([email protected])
manage communications with the tunnel and station earned his Class I B.Eng. (honors) degree in electrical and
DAU units. The LCU can have local data storage facili- electronic engineering in 2002 and his Ph.D. degree in
ties. The storage capacity is defined at detail design to electrical power engineering in 2005 from UMIST, United
ensure adequate data redundancy in conjunction with Kingdom. Currently, he is an assistant professor in the
OCC data management procedures. Department of Electrical and Computer Engineering at
the University of Cyprus.
Data Management System Pete Aylott earned his M.A. (honors) degree in natural
sciences from the University of Cambridge, United King-
Measurement Control and Configuration Software dom, in 1981. From 1985 to 2014 he was with CAPCIS Ltd.
Usually, some software is provided at the OCC to allow (now Intertek) Manchester, United Kingdom, where he
direct configuration of all aspects of the stray current was the commercial director. His consultancy experience
monitoring system, including individual measurement covered the management of stray current on light rail and
sequences, sampling frequencies, LCU data storage, and heavy rail underground systems.
data transfer functions. Diagnostic routines are included David Buxton ([email protected]) earned
to allow troubleshooting of individual DAU performance. his B.S. (joint honors) degree in chemistry and materials
The system can be configured to allow central rail-traffic in 1988 and his Ph.D. degree in corrosion science in 1994
controller synchronization across the LCU/DAU network from UMIST, United Kingdom. He has been at CAPCIS Ltd.
to OCC systems. (now Intertek) Manchester, United Kingdom, since 2000,
undertaking cathodic protection and stray current con-
Data Storage and Management Systems sultancy and monitoring activity, and he is currently the
A database management system can be provided to stray current business manager.
store, archive, and analyze the outputs from the stray
current monitoring system with the capability to output References
[1] I. Cotton, C. Charalambous, P. Ernst, and P. Aylott, ‘‘Stray current
data in a variety of forms—from high-level diagrammatic control in DC mass transit systems,’’ IEEE Trans. Veh. Technol., vol.
system-health visualizations down to detailed perfor- 54, no. 2, pp. 722–730, Mar. 2005.
[2] J. G. Yu and C. J. Goodman, ‘‘Modelling of rail potential rise and
mance-versus-time graphs at single or multiple sensor leakage current in DC rail transit systems,’’ in Proc. IEE Colloqui-
locations under user control. System alert and alarm lev- um Stray Current Effects DC Railways Tramways, Oct. 11, 1990, pp.
2/2/1–2/2/6.
els shall be configurable with KPI performance tracking [3] C. A. Charalambous, I. Cotton, and P. Aylott, ‘‘A simulation tool to
on a weekly and monthly basis. Such systems should be predict the impact of soil topologies on coupling between a light
rail system and buried third-party infrastructure,’’ IEEE Trans. Veh.
able to provide estimated location data for alert and Technol., vol. 57, no. 3, pp. 1404–1416, May 2008.
alarm conditions. [4] C. Charalambous and I. Cotton, ‘‘Influence of soil structures on
corrosion performance of floating-dc transit systems,’’ IET Res. J.
Electr. Power Aplicat., vol. 1, no. 1, pp. 9–16, Jan. 2007.
Conclusions [5] P. J. Aylott, ‘‘Stray current is for life-not just for christmas stray cur-
rent corrosion management strategies for dc traction systems,’’ in
This article aims to benefit stray current control design- Proc. IEE Seminar dc Traction Stray Current Control—Offer a Stray a
ers by highlighting the relative significance of two mutu- Good Ohm? Oct. 21, 1999, pp. 7/1–7/6.
[6] L. Ardizzon, P. Pinato, and D. Zaninelli, “Electric traction and elec-
ally important elements that should be addressed at the trolytic corrosion: A software tool for stray currents calculation,’’ in
design stage of dc mass-transit systems. These two ele- Proc. IEEE PES Transmission Distribution Conf. Exposition, Sept. 7–12,
2003, pp. 550–555.
ments are 1) indicative stray current calculations and [7] C. A. Charalambous and I. Cotton. (2011). Stray current control
assessments and 2) the specifications of the stray cur- and corrosion limitation for dc mass transit systems. The Univ.
Manchester. Manchester, England, U.K. [Online]. Available: https://
rent monitoring system that should be subsequently www.­e scholar.manchester.ac.uk/uk-ac-man-scw:141928
installed along the route of the system. [8] R. J. Hill, Y. Cai, S. H. Case, and M. R. Irving, “Iterative techniques for
the solution of complex dc-rail-traction systems including regenera-
It is worth noting at this point that the stray current tive braking, Generation,’’ IEE Proc. Transmiss. Distrib., vol. 143, no.
calculations following simplifying assumptions, carried 6, pp. 613–615, Nov. 1996.
[9] C. A. Charalambous, I. Cotton, and P. Aylott, “Modeling for prelimi-
out at the design stage, should be interpreted with care; nary stray current design assessments: The effect of crosstrack
keeping in mind that some standard-based approaches regeneration supply,” IEEE Trans. Power Delivery, vol. 28, no. 3, pp.
1899–1908, July 2013.
can be generic, or in some occasions, quite conservative. [10] C. A. Charalambous and P. Aylott, “Dynamic stray current evalua-
tions on cut-and-cover sections of dc metro systems,” IEEE Trans.
Veh. Technol., vol. 63, no. 8, pp. 3530, 3538, Oct. 2014.
Acknowledgment [11] Railway Applications. Fixed Installations. Electrical Safety, Earth-
The authors sincerely acknowledge the information and ing and the Return Circuit. Provisions Against the Effects of Stray
Currents Caused by dc Traction Systems, EN Standard 50122-2, Oct.
practical knowledge received by their late colleague P. 2010.
Aylott from Intertek. This article is a tribute to his long [12] Protection Against Corrosion by Stray Current from Direct Current Sys-
tems, EN Standard 50162, Aug. 2004.
service in stray current control and corrosion applica-
tions across the world. 

JUNE 2016 | IEEE vehicular technology magazine ||| 31

Authorized licensed use limited to: Mahidol University provided by UniNet. Downloaded on July 26,2023 at 04:35:51 UTC from IEEE Xplore. Restrictions apply.