CIGRE TB 559 - Impact of EMF On Current Ratings and Cable Systems

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/338388664
CIGRE TB 559 - Impact of EMF on current ratings and cable systems
Technical Report · December 2013
CITATIONS READS
0 187
5 authors, including:
Paolo Maioli Jean Hoeffelman

Prysmian Group 3 PUBLICATIONS 1 CITATION
5 PUBLICATIONS 12 CITATIONS
SEE PROFILE
SEE PROFILE
All content following this page was uploaded by Jean Hoeffelman on 18 September 2022.
The user has requested enhancement of the downloaded file.

559
Impact of EMF on current ratings

and cable systems
Working Group
B1.23
December 2013
Impact of EMF on Current ratings and Cable Systems
Members
H. Orton, Convener (CA), P. Maioli, Secretary (IT),
T. Barnes (AU), H. Brakelmann (DE), J. Bremnes (NO), F. Lesur (FR), J. Lopes (BR),
J. Saenz Orella (ES), J. Smit (NL)
Corresponding Members
S. Cherukupalli (CA), J. Hoeffelman (BE),
Contributions also received from
F. Cochet (CH), J. Stammen (DE), J. Brüggmann (DE)
Copyright © 2013
“Ownership of a CIGRE publication, whether in paper form or on electronic support only infers right
of use for personal purposes. Are prohibited, except if explicitly agreed by CIGRE, total or partial
reproduction of the publication for use other than personal and transfer to a third party; hence
circulation on any intranet or other company network is forbidden”.
Disclaimer notice
“CIGRE gives no warranty or assurance about the contents of this publication, nor does it accept
any responsibility, as to the accuracy or exhaustiveness of the information. All implied warranties
and conditions are excluded to the maximum extent permitted by law”.
ISBN : 978-2-85873-254-8
Page 1
IMPACT OF EMF ON
CURRENT RATINGS AND
CABLE SYSTEMS
Table of Contents
INTRODUCTION .......................................................................................... 5
1 GLOSSARY AND DEFINITIONS .................................................................... 6
2 SCOPE OF WORK AND TERMS OF REFERENCE............................................ 9
2.1 Scope of work ..................................................................................... 9
2.2 Terms of reference ............................................................................. 9
3 GENERAL PRINCIPLES ............................................................................. 10
3.1 Salient References ............................................................................ 10
4 CURRENT (DE)RATING PRINCIPLES .......................................................... 11
5 DERATING DEFINITION AND PRINCIPLES ................................................. 12
6 EMF MITIGATION METHODS: IMPACT ON CABLE RATINGS ...................... 13
6.1 Cable management ........................................................................... 14
6.2 Passive loops .................................................................................... 17
6.3 Metallic plates .................................................................................. 21
6.4 Ferromagnetic raceways ................................................................... 23
6.5 Steel pipes ........................................................................................ 25
6.6 Special cable design ......................................................................... 27
7 INTERACTION WITH NON THERMAL CABLE SYSTEM DESIGN ................... 31
7.1 Change of cable route ...................................................................... 31
7.2 Increase depth of laying ................................................................... 32
Page 3
7.4 Phase splitting .................................................................................. 35

7.5 Mitigation techniques with metallic components .............................. 36
8 ECONOMICAL EVALUATION OF MITIGATION MEASURES ......................... 41
8.1 Economical evaluation ...................................................................... 41
9 CASE STUDIES / EXAMPLES OF INSTALLATIONS ...................................... 42
9.2 Passive loops .................................................................................... 67
9.3 Metallic plates .................................................................................. 71
9.4 Ferromagnetic raceways ................................................................... 76
9.5 Steel pipes ........................................................................................ 80
9.6 Special cable design ......................................................................... 87
9.7 Magnetic field mitigation techniques for a 330kV cable .................... 94
10 ASSESSMENT AND METHODS TO LIMIT DERATING ................................ 97
10.1 Summary table ................................................................................ 97
10.2 Assessment of derating................................................................... 98
11 CALCULATION METHODS AND SOFTWARE CHARACTERISTICS .............. 99
12 MF MEASUREMENT TECHNIQUES ........................................................ 100
12.1 References .................................................................................... 100
CONCLUSIONS ........................................................................................ 101
FURTHER READING .................................................................................. 102
Page 4
INTRODUCTION
Numerous methods have been devised by electric utilities and various research organizations to manage power
frequency magnetic fields levels in the vicinity of underground cable systems. Information is available in Cigré TB
373, from WG C4.204, to implement the various methods, evaluate their impact on construction and their cost
effectiveness. Their impact on cable rating, losses, installation and operational costs need to be evaluated. In
particular, there are differing opinions about the derating effects of HV transmission cables placed in ferromagnetic
shielding enclosing structures such as pipes or casings. Past work at CIGRE and elsewhere addressed magnetic
field calculation procedure (with and without ferromagnetic components); however, they do not address the current
rating reduction of the magnetic field management methods or their practical applications to electric utility systems.
In most cases mitigation measures have disadvantages: either the current rating may decrease or the costs will
increase. Some mitigation methods are very cost effective, e.g. an optimal choice of the phase sequence for
double circuits/systems.
Where mitigation is required, the first question to answer is:
Is it
a) to reduce the magnetic field directly above the cable circuit, or

b) to reduce the field to a certain limit at a specified distance from the cable circuit, or
c) to minimize the width of the corridor within which a specified field limit is exceeded?
Some mitigation devices are more suited for mitigating field directly above the cable circuit, while others are
extremely effective in mitigating the field at a distance.
Installation in air is more easy than directly buried: probably derating is negligible in some cases.
This TB deals only with systems of balanced currents, that is with no zero sequence current.
Page 5
1 GLOSSARY AND DEFINITIONS
The main physical quantity discussed throughout the guidelines is the magnetic flux density (B) which is a
vector quantity. For simplicity, further in the text this quantity is referred to as magnetic field.
The units of measurements used throughout this document are MKSA units (Meter, Kilogram, Second and
Ampere), which forms part of the SI system (for the French "Le Système International d'Unités"). In the SI system
the unit for the magnetic flux density is Tesla (T), while microTesla (µT) is the most commonly used sub-unit.
Another commonly used unit for describing the magnetic flux density is Gauss (G). A useful conversion is 1
microTesla = 10 milliGauss.
When an RMS value is applied B is a scalar quantity.
The magnetic field intensity, H, is another physical quantity which is used for describing the magnetic field strength.
The SI system unit of measurement is A/m.
B is related to H via permeability of the medium, that is B = µH.
The word mitigation is used in this TB to represent an operation (or set of operations) aimed to limit the magnetic
field at the point-of-interest. In the technical literature, reduction and attenuation are used as synonyms of
mitigation, as are screening and shielding.
Regarding materials, the term purely conductive is used for materials having conductivity other than zero, and
relative magnetic permeability equal to 1. On the other hand the term purely ferromagnetic is used for materials
that have large relative magnetic permeability and negligibly low conductivity.
Some keywords used in this TB:
Cable management: an alternative geometric arrangement of the cable circuit resulting in an intentional reduction
of the magnetic field at the point-of-interest.
Mitigation device: an additional object/conductor/element intentionally added to mitigate the magnetic field.
Mitigation technique: a combination of methods implemented with the aim of reducing the magnetic field that is
created at a point of interest.
Passive loop: a mitigation device made with a closed cable or coil (loop), where the induced current reduces the
original magnetic field, according to Lenz’s law.
Metallic plate: a mitigation device composed of flat sheets, whereby induced eddy current in the plate reduces the
original magnetic field according to Lenz’s law.
Ferromagnetic raceway: a mitigation device, made from high permeability ferromagnetic plates, which completely
surrounds the cable circuit.
Shielding tape: a mitigation device made from high permeability ferromagnetic tape embedded into the cable.
Shielding Factor (SF): at a given point in space, the Shielding Factor (SF) of a given mitigation technique is
defined here as the ratio between the magnetic field modulus without and with the adoption of the mitigation
device.
Corridor width (also named Right-of-Way width (ROW)): lateral overall dimension of the transmission path inside
which the magnetic field is above a fixed value.
Page 6
Right-of-Way (ROW) (Corridor width): lateral overall dimension of the transmission path outside which the
magnetic field has to be reduced below a fixed value.
Derating: use of a cable circuit at a restricted current level than that for which it was originally designed, due to the
application of mitigation techniques.
Near-field and far-field:

o Near-field is the term used for the magnetic field close to individual cables, such that it may be dominated
by one or two of the phase currents. E.g., the magnitude of the magnetic field at the outer sheath of a
three-core cable will vary with position on the circumference, and will be highest directly outside the
individual cores.
o Far-field is the term used for the magnetic field sufficiently far away for individual phase influence to be
insignificant. Application of far-field approximations depends on burial depth and phase separation.
Resultant magnetic field, BR
o BR , being the resultant magnetic field, is the relevant quantity with respect to MF levels
(Refer for example to IEC 61786 or IEEE Std 1308-1994):
BR  Bx  By  Bz  Bmax  Bmin
2 2 2 2 2
where Bmax and Bmin are as defined in IEC 61786.
o For the two-dimensional case:

t = θ + 90°
Bmax t = θ + 0°
Bmax = t = θ + 180°
Bmin
t = θ + 180°
y t = θ + 270°
x
t = θ + 90°
Bmin t = θ Bmax Bmin = 0
t = θ + 0°
t = θ + 180°
t = θ + 270°
General case: Special case #1: Special case #2:

Elliptically polarised Circularly polarised Linearly polarised
Page 7
Some abbreviations used in this TB:
ELF: Extremely Low Frequency
EMF: Electromagnetic Field
FEA: Finite Element Analysis
FEM: Finite Element Method
ICNIRP: International Commission on Non-Ionizing Radiation Protection
MF: Magnetic Field
SF: Shielding Factor
A more complete list of physical quantities [SI units]:
H: magnetic field vector (or magnetic field intensity) [A/m]
B: magnetic flux vector (or magnetic induction or simply magnetic field) [T]
BR: resultant magnetic field [T]

2
J: current density [A/m ]
t: time variable [s]
f: frequency [Hz]
: angular frequency [s-1]
: magnetic permeability [H/m]

-7
µ0 : magnetic permeability of free space = 4 10 [H/m]
r: relative magnetic permeability [no unit]
: electrical conductivity [S/m]
δ: skin depth or penetration depth [m]
Page 8
2 SCOPE OF WORK AND TERMS OF REFERENCE
2.1 Scope of work

The work shall focus on single conductor, high voltage, AC, land cable systems, excluding pipe type and GIL
cables, considering:
1 Extruded dielectric insulation,
2 Laminar dielectric insulation,
3 Single-point bonding or balanced cross-bonding, so that there is no ground return current.
2.2 Terms of reference
1 To define correct terminology for field management techniques.

2 To review practical magnetic field management methods that are currently used for underground transmission
cable systems.
3 To quantify the shielding effectiveness of practical methods.
4 To review practical design and construction considerations relating to engineering, standardization of
components, scalability, environmental suitability of component material, impact by third party damage, reduction of
rating due to air inclusions, corrosion, theft of material, logistics and workers skill level.
5 To review the effectiveness of field management methods.
6 To quantify the cable current de-rating aspects of the various field management methods.
This WG shall neither cover any environmental or biological effects of EMF, nor discuss any specific levels of EMF.
Page 9
3 GENERAL PRINCIPLES
The thermal design of HV underground cable circuits is calculated according to related international standards and
commercially available software packages. It is calculated using the thermal properties of the cable and its
surrounds and is referred to maximum conductor temperature of the cables.
The current carrying capacity of a cable circuit is the maximum steady state current that can be transmitted
without exceeding the limit established for the conductor temperature at any point along the circuit, according to the
local environmental conditions and load factor. The calculation method for continuous load (load factor 100%) is
given in IEC 60287, whereas for cyclic loads (load factor less than 100%) is given in IEC 60853. For emergency
situations where a magnetic field limit is to be achieved, the transient temperature calculations are performed
according to IEC 60853.
The design of any mitigation device will be based upon continuous, emergency or cyclic load conditions, and will
be selected after defining the magnetic field limit.
Especially in urban areas, where most of the magnetic field mitigation devices are installed, there are variations in
the laying configuration or depth, lengths of parallels with other circuits or crossings with heat sources, that locally
modify the current carrying capacity of the cables. The HV system designer considers all these different conditions
and finally determines the rating of the circuit as the minimum of all the local current carrying capacities of the
cables.
The losses induced by the shielding devices introduce additional heat sources into the cable trench that increase
surrounding reference soil temperature seen by power cables, and the rating has to be recalculated. The possible
reduction of current carrying capacity depends mainly on the shielding method and on a variety of minor additional
factors, applies locally to the shielded part of the circuit and is equivalent to an elevation of ambient temperature.
The Shielding Factor (SF) is defined as the ratio between the magnetic flux density at a given point (P) in absence
of mitigation means (B0) and in presence of mitigation means (BS), namely SF(P)= B0(P)/ BS(P). The SF can be
calculated at any point such as directly over the cable circuit or at the nearest location specified for the magnetic
field limit. For design purposes, the SF has to be evaluated at the position P where the field is requested to be
reduced, because the SF is not constant in space, especially if the point P is close to a part of the shielding device.
3.1 Salient References
The work done for this TB is in strict continuity with the electrotechnical state-of-the- art, that can be found in
many publications. A non-exhaustive list of salient references is reported here to facilitate the comprehension and
development of the subject.
[1] Cigrè TB 373, “Mitigation Techniques of Power Frequency Magnetic Fields Originating from Electric Power
Systems”, February 2009.
[2] Cigré TB 375, “Technical Guide for Measurement of Low Frequency Electric and Magnetic Fields near
Overhead Power Lines”, April 2009.
[3] Cigré TB 320, “Characterization of ELF Magnetic Fields”, April 2007.
[4] IEC 60287 International Standard: Electric cables – Calculation of the current rating, 2006.
[5] IEC 60853 International Standard: Calculation of the cyclic and emergency current rating of cables, 2002.
[6] IEC TR 62095 Technical Report: Electric cables – Calculation for current rating – Finite element method, 2003.
Page 10
[7] IEC 61786 International Standard: Measurement of low-frequency magnetic and electric fields with regard to
exposure of human beings – Special requirements instruments and guidance for measurements, 1998.
4 CURRENT (DE)RATING PRINCIPLES

The design of a HV circuit is generally based on the cheapest solution, comprising in the total sum the cost of
cables, installation, maintenance, capitalization of the losses and other items, according to local experience and
expertise. The approach followed for the description of the mitigation impact on the circuit starts from the basic
point-of-view that the normally installed solution is the most convenient solution according to the weight of the
various addenda presented in this TB.
When a designer plans a new circuit, he computes the magnetic field generated by the cables and compares it
with the mandatory values that must be respected. If the value is lower than that specified, the project is completed,
but for those situations where there is an excess, mitigation may be foreseen.
The approach followed here is the Shielding Factor (SF) approach, which consists of determining the ratio
between the computed value and the specified value at the sensitive location (for example at the nearest house).
For each shielding factor there is in literature a variety of different mitigation methods, with technical characteristics
and installation procedures. The designer has to investigate each possibility and propose a working solution.
The installation of mitigation devices in close proximity to power cables inevitably generates additional losses,
and these losses usually generate heat and possible derating of the cables. The closer the proximity of the
mitigation device to the cables, the greater the mitigation is, but also the greater the derating. Other mitigation
measures, for example cable management or phase rearrangement, do not necessarily increase the losses or
decrease the rating.
A standard solution is a solution that maximises the rating and minimises the costs. Derating, or current rating
reduction, is not always present, but it is the starting point to a standard solution. For example, the introduction of
additional losses, the reduction of cables interaxial distance or the modification of other parameters, may reduce
the local current carrying capacity of the circuit.
All the practical considerations discussed in this TB and the examples reported start from the point-of-view that
cables are installed in standard trenches, avoiding any augmentation of the magnetic field that is not strictly
necessary.
Derating is calculated on the hottest cable in the section where mitigation measures are applied. It could result in a
derating of the whole link, if it is or becomes the limiting section.
The rating of the circuit should always be reassessed after the application of magnetic field reduction measures.
Page 11
5 DERATING DEFINITION AND PRINCIPLES
Derated Operation (as defined by IEEE Dictionary).
Use of equipment or a system at a more restricted performance level than that for which the equipment or system
was originally designed. Derated operation is usually implemented either to forestall failures or as a result of
system component failure.
Derating of a Cable Circuit: use of a cable circuit at a more restricted current level than that for which it was
originally designed, due to the application of mitigation techniques.
Current derating in this publication is expressed as a percentage of the unshielded current rating.
Page 12
6 EMF MITIGATION METHODS: IMPACT ON CABLE RATINGS

INTRODUCTION
For a single current flowing in a straight infinitely long conductor, the magnetic field far from the conductor (far field)
can be computed according to the following equation 6.1 (refer e.g. to TB 373, page 17)
0  I
B 6.1
2   r
where:
I is the conductor current [A],
r is the distance of the computation point from circuit centre [m].
For balanced three phase circuits eq. 6.1, again in far field approximation and after an algebraic arrangement,
becomes:
0  k  d
B I 6.2
4   r 2
where
d is the interaxial distance between the conductors [m],
k is the geometric coefficient (refer e.g. to TB 373, page 27)
and where the other symbols have the same meaning of eq. 6.1. The important difference is the presence of
exponent “2” at the distance from circuit centre, which determines a rapid decay of the magnetic field with distance.
There are some parameters which can influence the magnetic field, even if they are not strictly present in the
computation formulae. These are for example the voltage of the line, because it increases the insulation thickness
and can increase the interaxial distance. An increase in the conductor section increases the cable diameter and
also enables to transmit higher current, so that losses in the shielding device will also be higher. Solid bonding of
the sheaths, instead of single-point bonding, allows circulating sheath currents that generate additional losses.
Mutual heating also necessitates an increase of interaxial distance.
Finally, for multiple circuits laid in parallel, the phase sequence and arrangement determine a cancellation or
summation of the field components of single circuits, giving rise to different shapes of the magnetic field distribution
curves: the selection of the best phase sequence and optimized arrangement can provide large magnetic field
reduction and exponents in eq. 6.2 higher than 2.
CIRCUIT MANAGEMENT
As a first approach, increase the distance of the circuit from areas that may be affected, for example diverting
cable route or by increasing trench depth.
Circuit management would be the simplest technique for attempting to minimize magnetic fields. Circuit
management is the movement of the whole circuit without changing the cable circuit geometry. This will increase
Page 13
the distance of the circuit from the magnetic field measurement point and may be achieved by diverting the cable
route or by increasing the trench depth.
Depending on the distance of the diversion, diverting the cable trench will have a negligible cost impact and,
depending on the magnitude of the circuit movement, increasing trench depth will have the detrimental effect of
requiring the individual phases to be spaced further apart or will result in de-rating the cable. Spacing the cables
further apart causes an increase in magnetic field possibly negating the benefits achieved by moving the circuit.
6.1 Cable management

Cable management is the design of the cable installation with respect to a reduced magnetic field, but without
modifying the cable construction itself. Relevant parameters are the positions of the cables (laying arrangement),
the bonding of sheath or screens, and – in case of more than one system – the electrical lay-out and the way of
operation: the latter point means the phase-sequences rearrangement, the splitting of the phases in more than one
conductors as well as the operation of double circuits in a six-phase mode.
Cable management is the easiest way to reduce the EMF directly at the source, acting on the position of the
conductors that generate the electromagnetic field. For example, trefoil formation is often used as an alternative to
flat formation, giving a reduction of the EMF of √2 for the same interaxial distance and distance from the circuit, in
far field approximation: the cost to benefit relationship is good, and the derating depends mainly on the sheath
construction
It is also possible to let the currents circulate into the metallic sheath, by proper electrical connections (solid
bonding at both ends). Solid bonding may give a significant reduction of the magnetic field, but the derating in
these cases is so important that this technique is rarely applied to underground cables to mitigate EMF.
For any laying geometry and for any electrical connection of the sheaths, the standards IEC 60287 and IEC 60853
contain the information required for the computation of the conductor temperatures and losses.
6.1.1 SINGLE CIRCUIT
Contrary to the case of more than one circuit, for single circuits the phase sequence is not relevant for the
magnitude of the magnetic field. Longitudinal currents in normally designed sheaths or screens will not lead to
significant field reductions, but to extensive losses and derating. Substantial effects are only achievable by
increasing sheath/screen cross sections to sizes which are similar to the conductor cross section. Interesting
results for such cables, with and without integrated high permeable magnetic foils are discussed in section 6.6.
Straight-forward methods to reduce the magnetic field are:
 to increase the cable laying depth,

 to minimise the cable interaxial distance, and
 to prefer triangle arrangements of the cables instead of flat formations.
For cables in soil, all these measures may have a strong influence on installation costs as well as on the losses
and the current rating of the cables, i.e. on the operational costs, too.
Triangle formation as an alternative to a flat formation with the same interaxial distance allows an EMF reduction by
a factor of √2, in far field approximation. The cost/benefit relationship is good. For cables with aluminium or lead
sheaths installed with small interaxial distances, the triangle arrangement may reduce the cable losses, as
compared with the flat arrangement.
Depending on the required EMF reduction (the SF), optimisations of cable arrangement, laying depth and distance
with respect to the total costs (trench, cables and losses) can be achieved, as discussed in the case studies
(section 9).
Page 14
6.1.2 DOUBLE CIRCUIT
In many cases, more than one cable system is installed in the same trench, so that two different situations must
be evaluated: the cable systems are operated separately or in parallel.
In the first case, for two or more cable systems which are operated separately, the choice of the phase sequences
in the systems does (normally) not influence the distribution of the conductor current, but definitely influences the
sheath/screen currents as well as the EMF.
For the second case, high voltage AC single-core cables are most common for bulk power transmission, which are
designed as a twin system, so that electrically in each phase two cables are working in parallel. Hereby, the
partitioning of the total phase current, as well as the sheath/screen currents, will be strongly dependent on the
phase sequence as well as on the geometrical arrangement of the six cables.
For such double systems, not only the interaxial distance inside the systems, but also the separation between the
systems can be optimised for any given trench width. Depending on such arrangements and in combination with
the chosen phase sequence of the cable cores, magnitude and phase of the conductor currents as well as of the
sheath/screen currents (depending on the type of bonding) are decisively influenced, resulting in specific losses,
current ratings and EMF.
Comparing flat with triangle formations of the cable systems shows that there may be different optimal phase
sequences for the two cases. As with single circuits, the influence of the sheath currents on the EMF is small,
unless very large screen sections are used.
Depending on the required EMF reduction (SF), optimisations of cable arrangement, laying depth and distances in
combination with the phase sequences with respect to the total cost (trench, cables, losses) can be achieved, as
discussed in the case studies (section 9).
6.1.3 SPECIAL CASES - OPERATIONAL MODES
There could be special ways of operation of systems which are based on the phase-sequences rearrangement: in
particular the splitting of the phases in more than one conductors as well as the operation of double circuits in a six-
phase mode. These cases are not common and may be applied only in special cases. Their theoretical aspects
and benefits are described here below.
6.1.3.1 Split phases
Utilizing two single-core cables per electrical phase introduces the possibility of active field compensation, using
two trefoils as “building blocks” as illustrated by Fig 6.1.3.1. Each trefoil will generate a rotating magnetic field at
their shared geometrical axis, and because the two trefoils are rotated 180 ° relative to each other, their fields will
be opposing and nearly cancel each other out.
Page 15
L1 I1 = I ej0° I3 = I/ 2 ej120° L3 L2 I2 = I/ 2 e-j120°
ωt mmf ωt
I1 = I/ 2 ej0 ° L1 mmf ≈ 0 L1 I1 = I/ 2 ej0°
j120 ° -j120 °
I3 = I e I2 = I e at common
centre/axis
L3 L2 I2 = I/ 2 e-j120° I3 = I/ 2 ej120°
L2 L3
Fig. 6.1.3.1 Illustration of the inherent EMF cancellation of a split-phase hexagonal cable configuration
To maintain approximately the same power rating, at the same operating voltage, the conductors of the hexagonal
configuration should have approximately half the size of the trefoil conductors. The improved cooling (larger
surface) will allow more power to be transmitted if the conductor size is exactly 50% of the trefoil. If the circuits are
defined to have Milliken conductors and cross bonded screens, the improved power rating will solely be due to
improved cooling. If non-Milliken conductors and/or bonded screens are used, an additional decrease in power loss
will apply for the hexagonal configuration.
With respect to EMF, sharing the same total phase current equally between opposite conductors, as shown in Fig.
6.1.3.1, will immediately result in approximately 50% near-field reduction. Connecting oppositely facing conductors
to the same electrical phase will generate two opposing (rotating) magnetic fields, resulting in a more dramatic
reduction of the far-field.
Consequently, the hexagonal configuration is deemed highly interesting when achieving low EMF without suffering
de-rating penalties. Compared to the trefoil configuration, negative aspects may include higher cable cost due to
increased number of cables, use of “Y” joint and possibly more complex installation (more, but smaller, single-core
cables). Positive aspects include a reduced far-field magnetic field and increased power rating for the same
amount of conductor material. Additionally, the integrity of the EMF mitigation over time will be unquestionable as it
is an inherent part of the high-voltage cables themselves.
The most attractive aspect of this approach is the fact that it will actively and directly reduce the source of EMF,
rather than attempting to reduce a generated dipole field. A quantification example illustrating the effect of this
scheme is presented in section 9.1.4.
The well-known “low-reactance” split-phase configuration used for overhead lines can be viewed as a deformed
hexagonal configuration, with more favourable/practical installation requirements. On a general basis there is
reason to expect higher field values if the axial distance is increased, or a higher conductor temperature if the same
configuration width is maintained. This case is also included in section 9.1.3.
6.1.3.2 Six-phase system
Bulk power transmission by means of six-phase HVAC cables was first proposed with the aim to enable HVAC
single-core submarine cables for high voltages and great conductor cross sections. The principal laying
arrangement, which applies also for land cables, is shown in Fig. 6.1.4.1: six single-core cables are laid in pair,
each pair with two close-by cables. Each pair of adjacent conductors is related to one phase of the supplying three-
phase system in a way that their AC voltages and currents are in phase opposition.
The voltages of opposite phase may be achieved by using a transformer with a tap at the middle of the secondary
windings or by connecting two transformers with opposition wound (anti-parallel) secondary windings. The latter
are commonly used in combination with power converters. At the opposite end of the cables, identical transformers
Page 16
are to be connected in the same manner.
The voltages in the individual cables in Fig. 6.1.4.1 can be described as:
U L12  U L11 ; U L2 2  U L2 1 ; U L32  U L31 , (6.1.4.1)
with the two AC systems with opposite phase angle:
U L2 1  a  U L11 ; U L31  a  U L11 ; U L2  2  a  U L1 2 ; U L3 2  a  U L1 2 ; a  e j 120

2 2
The transmission capacity of the proposed configuration is elucidated in references [1-9] of chapter 9, with specific
losses and costs. The high redundancy of six-phase systems is important: if one pair fails, four cables are left, with
a higher transmission power than in the case of a damaged three-phase double system.
b) L1-1 L1-2 L2-1 L2-2 L3-1 L3-2
UM
1 2 3
-U M
l0 l0
2 l0
Fig. 6.1.4.1: Six-phase HVAC cable system: laying arrangement and principle of cross-bonding
Thermal properties including derating need further investigation for this potential energy saving application.
6.2 Passive loops
Passive loop techniques make use of conductors which are laid parallel to the power cables but are electrically
connected at both ends, creating a loop; currents are induced into them resulting in a cancellation of the
inducing magnetic field so that the total field is reduced.
The passive loop technique is one of the most recently applied methods for EMF shielding. The passive loops
are easily installed into the trench together with HV cables, with negligible impact on the laying operations of the
main cables: standard practice is to limit the increase of ambient temperature of the cables to a few degrees
Celsius, with a careful choice of the section of the conductors and of their position. Low voltage cables are
normally used, due to the low value of the potential induced into the passive conductors. The loops are
arranged into the trench, joint bay or manhole, either on the surface of the compacted backfill or at the same
level of the HV cables.
Page 17
Currents tend to circulate mainly in the loops placed closer to the power cables contributing also to higher
losses and thermal influence on the main cables due to their proximity. Balancing is required at design stage to
select the section and the position of the passive cables: the calculation is not simple and needs dedicated
software, due to the presence of many variables.
6.2.1 HEATING INDUCED BY PASSIVE LOOPS
There are many ways to compute the thermal influence of the passive loops on the main cables. A general
approach could be to be use FEM methods, describing all the heat sources and other thermal characteristics,
but it would be time consuming when dealing with standard designs. To simplify the approach, it is better to
distinguish the two main cases where the cables are directly buried or installed in air.
6.2.2 DIRECTLY BURIED INSTALLATION
The approach followed here to evaluate the thermal influence of the passive loops is to compute the increase of the
ambient temperature of the power cables due to the presence of passive cable. It is straightforward that the
influence increases with the proximity of the passive cables and their losses, giving rise to the well-known “mutual
heating” between adjacent cables, that is normal practice in HV cable design. The induced current in the passive
loop can be computed according to an electrotechnical manual; see for example IEC 60287-1-3 Electric cables –
Part 1-3: Current rating equations (100% load factor) and calculation of losses - Current sharing between parallel
single-core cables and calculation of circulating current losses.
IEC Standard 60287-2-1 § 2.2.3.1 defines the contribution to heating due to each external source, according to
its position, induced losses and thermal resistivity of the soil (non-drying out soil is considered here).
Let Δθkp be the thermal increment, above ambient temperature, of the p cable, HV cable, due to the presence of
the k cable (passive cable). The overall increment of ambient temperature, Δθ p, at each HV cable p is given by
the sum of the q contributions Δθkp computed for all the passive cables:
q
 p    kp  1 p   2 p  ..... kp  ..... qp (6.2.2.1)
1
The terms Δθkp are given by the following formula:
t  d pk
'

 kp   Wk  ln   (6.2.2.2)
2 d 
 pk 
k’
d‘pk
Soil surface
k
dpk
p Page 18
Fig. 6.2.2.2 Thermal image of the passive cable used in eq (6.2.2.1).
where:
ρt is the thermal resistivity of the soil (K*m/W),
W k is power dissipated into the passive cable k (W/m),
‘
d pk is the distance from the axis of power cable p to the axis of cable k’ symmetric of the passive cable k relative
to soil surface (m),
dpk is the distance from the axis of the power cable p to the axis of the passive cable k (m).
In the previous equation the losses W k [W/m] of the passive cable k are given by
Wk  Rk  I k2 (6.2.2.3)
where:
Rk is the electrical resistance of the passive cable k at operating temperature (Ohm/m),

Ik is the current flowing into the passive cable k (A).
From equation 6.2.2.2, when the passive cable is close to the soil surface, the argument under the logarithm
approaches the value of one and the thermal influence becomes negligible.
Each term Δθp at equation 6.2.2.1 expresses the increment of the temperature of the cable p. For the rating
point of view it is important to determine the temperature of the hottest cable of the power circuit. After these
considerations, “hottest” means the hottest power cable after shielding and it will not necessarily be the same
power cable as it was before the shielding device was applied (for example the central one or the deepest).
In this sense, the derating has to be considered: the new rating shall be computed with the temperature of the
hottest power cable after shielding. All the p contributions term Δθ p in equation 6.2.2.1 shall be evaluated and
summed to the rating temperature of the respective cables. Comparison of these new temperatures will
determine the new hottest cable, case by case.
For the sake of precision, let’s assume, for example, that the hottest cable of a circuit at nominal rating without
shielding is at 90°C and a lateral cable is at 87°C. With application of the passive loops the new temperatures
are computed respectively as 91°C and as 92°C. In this case, the effective temperature of the hottest point has
increased by 2°C, that is from 90°C to 92°C and this is the new position of the cable that will determine the new
rating when it is reported to be 90°C.
Mutual heating acts also among passive cables, so that a recursive computation could be required to determine
the resistance of the passive conductors at correct working temperature.
6.2.3 INSTALLATION IN AIR
When passive cables are installed inside in air, the heat dissipated by the passive cables is conveyed by air to
the walls of the trough or of the joint chamber. This situation is different from the directly buried case, because
of the particular path that the heat travels before reaching the unperturbed temperature. In this case, ambient
Page 19
temperature rise Δθ [°C] of the power cables is computed directly on the base of the sum W t [W/m] of the losses
of all the passive cables k are given by
  Wt  T4 k (6.2.3.4)
where
W t   W k
k
(6.2.3.5)
It is assumed that the air temperature inside the trough or joint bay increases uniformly and that the effect of
radiation is neglected, due to the small viewing angle between cables.
The thermal resistance T4 [K*m/W] encountered by W t [W/m] is given by the surface resistance T 4w at the walls
of the trough summed to the resistance T 4t of the trough towards open air:
T4  T4 w  T4t (6.2.3.6)
The resistance T4w [K*m/W] can be computed using equation at IEC 60287-2-1 § 2.6.2.2 here reported:
1
T4 w  (6.2.3.7)
3 p
where p is the perimeter of the trough (m).
The external resistance of the trough T4t [K*m/W] can be computed with the formula reported in IEC 60287-2-1
§ 2.2.2.
T4t 
t
2

 ln u  u 2  1  (6.2.3.8)
where:
ρt is the thermal resistivity of the soil (K*m/W),
W k is power dissipated per unit length by the passive cable k (W/m),
2 L
u is the following expression u 
Dt
L is the distance from the surface of the ground to the axis of the trough (m),
Dt is the internal equivalent diameter of the trough (m).
For values of u exceeding 10, the following simplified formula can be used instead of eq. (6.2.3.8):
t
T4t   ln 2  u  (6.2.3.9)
2
Page 20
Each passive cable shall be operated at a temperature less than or equal to its rated temperature; for the
computation of the temperature of each single passive loops refer to IEC 60287-2-1 § 2.2.1.
To enable free air convection, care shall be taken to avoid mechanical contact between passive loops and
power cables.
6.3 Metallic plates
The use of magnetic and/or conductive plates to mitigate MF is described in TB 373. Conductive plates can have a
significant shielding effect far from the source whereas ferromagnetic plates are only effective near the source
(unless they fully encompass the source as for a raceway). Conductive plates generally have lower losses and are
more robust with respect to installation imperfections.
In this section a hypothetical, comparative example is analysed using a two-dimensional finite element tool, having
a geometry illustrated by Fig 6.3.1, and assuming perfectly balanced 50 Hz phase currents with a magnitude of
1360 A rms.
BR [µT] BR [µT]
Fig. 6.3.1. Sample configuration for quantification and comparison of plate shielding effects: inactive
shielding plate (left) vs. conductive, non-magnetic copper plate (right)
Results shown in Fig. 6.3.1 are: magnetic flux lines (snapshot, no scale), and BR in air (contour plot).
The shallow depth of the shielding plate results in an uneven reduction of the unperturbed field that must also be
categorised as a near-field effect. The difference in shielding effect directly above the cables is significant, while the
shielding effect at 5 m distance is about 35 % better for copper than for ferromagnetic steel, as can be seen from
Fig. 6.3.2. Given a shallow burial depth (0.6 m), the far-field approximation presented in TB 373:
BR, flat r  
3d
I
5r 2
does not yield perfect correlation with unshielded FEA results directly above the cables, as can be seen from Fig.
6.3.2. In other words: near-field effects reach up to the ground surface in this case. Maximum deviation between
FEA and far-field approximation is limited to 13%. Increased phase separation, from the current 0.24 m, would
result in increased deviation.
Page 21
Fig. 6.3.2 Comparison of magnetic field (left) and shielding factor (right) at soil surface
The graphs in Fig. 6.3.2 show BR along a path perpendicular to the cables’ axes at the soil surface for all cases,
and corresponding shielding factors obtained using metallic plates. The obtained shielding factors for this
hypothetical case are by no means impressing anywhere except directly above the centre cable.
Improved shielding factors can be obtained by increased burial depth, increased plate width and/or additional
shielding plates, e.g. vertical plates on both sides as described in paragraph 9.3.1, and/or finally by optimizing the
distance between cables and shielding plate(s). However, in addition to increased cost and complexity of
installation, improved EMF shielding by metallic plates will come at some cost of added shielding loss, again
implying that the impact on current rating must be adverse.
The power loss distribution in the ferromagnetic shielding plate is illustrated by Fig. 6.3.3.
1,60E+04
1,40E+04
1,20E+04
Power density [W m -3 ]
1,00E+04
8,00E+03 lower edge
6,00E+03 upper_edge
middle
4,00E+03
2,00E+03
0,00E+00
0 100 200 300 400 500 600 700 800
[mm]
Fig.6.3.3. Power loss density in magnetic steel plate along full plate width (left) and detailed results at
approximately = 100 mm (right)
lower edge
1,60E+04
1,40E+04
Power density [W m -3 ]
1,20E+04
1,00E+04
8,00E+03
6,00E+03 lower edge

4,00E+03
2,00E+03
0,00E+00
0 100 200 300 400 500 600 700 800
[mm]
Page 22
Fig. 6.3.4. Power loss density in copper plate along full plate width (left) and detailed results at
x = 0 mm (right).
The above example has a shielding plate thickness of 100 mm, which would imply very high material cost. With a
more realistic plate thickness, the plate current distribution in Fig. 6.3.4 would be more or less identical along the
upper and lower edges of a conductive plate. The shielding factor would decrease.
Examples quantifying thermal impact/derating due to metallic shielding plates are presented in section 9.3.
6.4 Ferromagnetic raceways

Inside the closed ferromagnetic shield the field lines have a complicated behaviour that depends on the
combination in space and time of the three power line currents and of the induced currents, each with its own
phase shift and geometrical distribution. The magnetic field and consequently induced losses cannot be
expressed with analytical formulae. Experience confirmed that loss factor λ2r derived from IEC 60287-1-1 § 2.4.3
formulae used for pipe type cables, are still valid, provided that the power cables are in a closed arrangement and
that the raceway is made of selected commercial steel. Special and highly costly materials that have better
shielding performance have not been considered due to the fact that they are rarely used. In the ferromagnetic
raceways the cables are normally arranged in trefoil and their configuration can be controlled and adjusted,
because the raceway is open: the cement mortar finally locks the cables in position, before the cover is fixed (Fig.
6.4.1). Mechanical fixing of the power cables is essential for the functionality of the circuit. For other configuration
or different steel, specific FEM computation is advised.
X
Fig. 6.4.1 Raceway dimensions: base width X and height Y.
The equivalent radius Re can be computed according to the formula derived from IEC 60287-2-1 § 2.2.7.3, valid
for X/Y< 3:
1 x 4 x  y2 
ln Re  
x
      ln 1  2   ln (6.4.1)
2 y  y   x  2
For X=Y, as it is in most cases, Re is equal to 0.549 X.

It is straightforward to define the equivalent diameter De  2  Re .
Page 23
For 3 cables equally loaded, in trefoil formation, the following formula for the losses λ 2rt can be applied:
 0.0115  s  0.001485  De   5
2 rt  0.76     10 (6.4.2)
 R 
Y
where:
s is the cable interaxial distance [mm],

De is the raceway equivalent diameter [mm],
R is the conductor resistance at maximum operating temperature [Ohm/m].
X
The formula is valid up to cable diameter of about 200 mm, due to unknown accuracy above this physical size. For
other configurations (e.g. centred cables in large raceway) and larger diameters, FEM computation is advised.
The factor of 0.76 in the previous formula (6.4.2) extends the range of validity to 50 Hz.
For 3 cables equally loaded, in flat formation, the following formula for the losses λ 2rf can be applied:
 0.00438  s  0.00226  De 
2 rf  0.76     10 5 (6.4.3)
 R 
Y
where:

De is the raceway equivalent diameter [mm],
R is the conductor resistance at maximum operating temperature [Ohm/m],
X
The formula is valid up to cable diameter of about 200 mm.
For 3 cables equally loaded, in trefoil formation in special steel raceway, the following formula for the total losses
W r [W/m] can be used:
Wr  3  I 2  2 r  R (6.4.4)
where:
I is the cable current [A],

λ2r is the raceway losses parameter (either in trefoil or flat configuration),
R is the conductor resistance [Ohm/m],
In formula (6.4.4), the coefficient 3 accounts for the presence of three cables (Note that number of cables
different from 3 requires special computation).
As the losses are generated into a thin layer, in a nearly uniform way because the cables are close to the
centre of the steel structure, temperature rise  r of the area inside is uniform and can be computed according
to:
Page 24
t  4 L 
 r  Wr  T4  Wr   ln  
 (6.4.5)
2  De 
where
W r is the power dissipated into the raceway [W/m],
T4 is the external thermal resistance of the raceway [K*m/W],
ρt is the thermal resistivity of the soil [K*m/W],
L is the laying depth at the centre of the raceway [mm],
De is the “equivalent radius” of the raceway [mm].
For shallow burying the extended formula in IEC 60287-2-1 § 2.2.2 can be used in eq. (6.4.5)
T4  Wr 
t
2

ln u  u2  1 
(6.4.6)
where
T4 is the external thermal resistance of the raceway [K*m/W],
2L
u is the ratio between the laying depth L [mm] and the “equivalent radius” D e [mm].
De
In case of large raceways, FEM computation is advised, either for losses or for temperature rise.
Circulating and eddy current losses in screens and sheaths shall be computed according to IEC 60287-1-1 §
2.3.11. For skin and proximity effects in the conductor refer to IEC 60287-1-1 § 2.1.5, where loss factors are
increased by the so called “in pipe factor” by 50%. These corrections are due to the presence of the raceway
that modifies the magnetic field lines into the cables.
For cables with configurations different from trefoil or for raceways of significantly elongated shape of the
transverse section, a specific FEM computation is needed to determine the losses. In fact, the cable
configuration and relative height of the box that contains the cables are of fundamental importance in
determining the dissipated power.
6.5 Steel pipes

It is a well-known mitigation technique to encapsulate the three cables of a three-phase system within a closed
ferromagnetic structure like a steel pipe. Commercial available steel pipes normally have a high relative
permeability in a range of 500 up to 1300. They offer a low reluctance magnetic path to the magnetic flux, thus
decreasing the magnetic field outside the steel pipe. Shielding factors of more than 100 can be achieved.
It is essential for the shielding to be effective that the three-phase system of currents must be inside the steel pipe,
not necessarily symmetrical, but which in any case must add to zero: any residual zero-sequence current is not
shielded and may induce large losses.
Standard IEC 60287 contains an empirical formula for the losses induced into ferromagnetic pipes in the two
cases of cables laid cradle or in trefoil formation. For trefoil configuration and an operating frequency of 50 Hz
the loss factor λ2pt is given by:
 0.0115  s  0.001485  De 
2 pt  0.76     10 5 (6.5.1)
 R 
Page 25
where
d is the pipe internal diameter [mm],
R is the AC conductor resistance at maximum working temperature [Ohm/m].
The factor of 0.76 in the previous formula (6.5.1) accounts for the range of validity to 50 Hz.
For cradle configuration the loss factor λ 2c is given by: by:
 0.00438  s  0.00226  d  5
2 pc  0.76     10 (6.5.2)
 R 
the parameters having the same meaning as in the previous formula and the factor of 0.76 in the previous formula
(6.5.2) accounts for the range of validity to 50 Hz.
When a ferromagnetic pipe is used to shield the magnetic field, cables are usually installed in trefoil formation,
to minimize the magnetic field and the losses. Nevertheless, during laying operations, the cables can assume a
configuration midway between trefoil and cradle due installing procedure or due to snaking induced by thermal
elongation of the cable: average of the values obtained from equations (6.5.1) and (6.5.2) can be a better
approximation. The total losses, W p [W/m], generated into the pipe by three cables arranged inside the pipe are
given by
W p  3  R  I 2  2 (6.5.3),
where
I is the cable current [A],
λ2 is the loss factor (either in trefoil or flat configuration),
R is the AC conductor resistance at maximum working temperature [Ohm/m].
In formula (6.5.3), the coefficient 3 accounts for the presence of three cables (Note that number of cables
different from 3, requires special computation). Both the above empirical equations have been adopted by IEC
60287; they have been derived for pipe sizes and types of steel that have been extensively used in the US.
Cases with other pipe sizes, with other materials and especially with other laying arrangements, should be
treated very cautiously. As shown in the case studies, for big pipes and enlarged cables distances, extremely
high pipe losses may arise, which may deviate from the results of the above equations. It essential to carry out
further analysis of the eddy current and hysteretic losses of the steel pipe as well as the increased cables losses
(eddy currents and circulating currents in conductors, sheaths, screens) because of the higher magnetic field
concentration inside the pipe. Such analysis can be done by means of numerical methods, e.g. by the FEM,
which must ensure a coupled calculation of the thermal and the electromagnetic fields, with consideration of the
nonlinear and field dependent material properties (e.g. magnetic permeability).
The losses in screen and sheaths have to be computed according to IEC 60287-1-1 § 2.3.11, that presents an
additional factor of 1,5 times relative to formula presented in § 2.3.1.
For ferromagnetic pipes the losses are generated into a circular structure, in nearly uniform way, if the cables
are close to the centre. It is known that power losses uniformly generated along a circumference create a
uniform temperature rise in the area inside it. The exact value, Δθ p, can be easily computed with the well-known
Kennelly formula, normally used to compute cable temperatures when the laying depth is not much greater of
the diameter, as can happen for pipes that are close to the surface to improve heat dissipation:
ρt 2 L  2 L 
2  ρ  4L 
Δθ p  W p   ln      1   W p  t  ln   (6.5.4)
2π  d  d   2π  d 
 
Page 26
where
W p is the power dissipated into the pipe [W/m],
L is the laying depth at the axis of the pipe [mm],
d is the internal diameter of the pipe [mm].
d
Fig. 6.5.1: Steel pipe with three cables: pipe diameter d and laying depth L
The right-hand side of eq. (6.5.4) holds true in good approximation, if L >> d is fulfilled. For large raceways or
shallow burying the extended formula can be used in eq. (6.5.4). For the sake of precision also the thermal
resistance of the anticorrosion protection should be added, with its proper thermal resistivity.
Equation (6.5.4) describes the ambient temperature rise of the cable due to the losses in the steel pipe. This can
be considered in the rating equation in IEC-publication 60287 simply by increasing the ambient temperature by
p.
6.6 Special cable design
6.6.1 THREE-CORE CABLE

Three core cables are included in this document because of the similarity to cables in trefoil arrangement in steel
pipes or raceways.
6.6.1.1 Without magnetic tape
The three core high voltage cable design is equivalent to a standard high voltage cable arrangement in trefoil. From
construction point-of-view each of the three separate cable cores has its own metal sheath and PE outer covering.
The three individual cores are “laid” together and are provided with an overall steel armour for cable pulling and
protection purposes.
At medium and low voltages three phased circuits are normally installed using twisted three core cables and form a
helix with a pitch ranging from 1 to 2 m. This conductor arrangement greatly reduces the field at distances larger or
of the same order of magnitude as the pitch. This is highlighted in TB 373 (p 104). For high voltage cables, normal
practice is to use single core cable for land installation; for submarine installation HV three core cables are usually
installed, but are beyond the scope of this TB.
Page 27
The overall construction from an electromagnetic field perspective is similar to three parallel cables in touching
trefoil. Typically these kinds of cables are installed in a duct system, of either plastic or steel pipes.
For further reference to the behaviour of the emitted electromagnetic field when installed in non-metallic duct and
plastic pipes see paragraph 6.1.1 and 6.1.2. For installation in a steel pipe reference is made to section 6.5
6.6.1.2 With magnetic tape
Also referring to the three core design mentioned in paragraph 6.6.3.1 a three core high voltage cable was
developed with an integrated electromagnetic shield [1].
By applying two or three overlapping high permeable shielding tapes, very high shielding factors are attained even
when compared to standard steel pipe cables. Due to sensitivity to deformations, special diligence must be applied
in the handling of the shielding tapes during cable production.
The measurement results are complemented by means of an adapted finite elements analysis. The results shows
that the cable screen losses can be significantly reduced by the integrated shielding.
This design is also advantageous when the cable is laid in a steel pipe. The losses in the magnetic pipe and armor
are almost eliminated while the losses in the shielding tapes are very small. As a result the total cable losses are
lower and consequently the current capacity of the cable can be increased. Integrated electromagnetic shielded,
three-core cables up to the voltage level of 220 kV are available.
[1] D. Bielefeld, H. Brakelmann, J. Brüggmann, V. Waschk: IES Cables – Power Cables with Integrated
Electromagnetic Shielding, CIGRE-report B1-305, Cigré-Conf. 2008
6.6.2 SINGLE-CORE CABLES WITH LARGE SOLID-BONDED SCREENS
If the metallic sheaths or screens of single-core cables are bonded at both ends, currents will be induced which
tend to flow in phase opposition to the conductor currents and, for very large cross sections of the sheaths/screens,
approach more or less exactly the value of a return current.
Such coaxial constructions with very large cross sections of sheaths/screens are normally not used as a shielding
measure, since their shielding factor of approximately 20 to 30 is not enough high when compared with the
disadvantage of very high additional losses.
A new technique of integrating high permeability shielding tapes (see paragraph 6.6.1) offers the possibility to
achieve higher shielding factors by applying the coaxial principle. This is reported in paragraph 6.6.2 and 9.6.2.
This paragraph reports an example of a cable in which the screen has a very large cross section, allowing a
relatively low value of the total losses and cable derating when used in a solid bonding configuration. The
construction of such a cable needs some discussion to justify the use of a large section for the screen; no
application has been described so far.
2
For a range of sheath sizes up to 500 mm the screen losses, for a 110 kV XLPE cable with an
2
2500 mm aluminium conductor, for a laying depth of 1.2 m at a distance from axes of 0.5 m are reported in Fig.
6.6.2.1.1. The same Fig. 6.6.2.1.1 elucidates that for such a laying arrangement:
Page 28
 a loss maximum loss factor 1 is given for a sheath cross section As of approximately from
2
100 to 200 mm with a loss factor of more than 6, and that
 beyond this maximum the sheath losses will decrease with farther enlarged sheath cross
sections.
800
600
y 400
1
200
0
0 500 1000 1500 2000 mm2 2500
AS
Fig. 6.6.2.1.1 Sheath loss factor 1 (middle cable) as a function of the cross section AS
This phenomenon can be explained by the fact that sheath currents increase rapidly with increasing sheath cross
2
sections. In the range from 500 to 1000 mm sheath currents reach the magnitude of the conductor currents, so
that further enlargement of the sheath cross section will no longer result in larger sheath currents, but in decreasing
sheath losses.
By increasing the screen cross-section, the losses decrease from more than 600 per cent of the conductor losses
down to about 45 per cent, with a 2500 mm² copper screen. In comparison to systems with cross-bonded or single
point grounded systems, this might appear unfavourable, but the return currents strongly shield the emitted
magnetic field of the complete cable system with shielding factors of approximately 20 to 30.
The above described shielding effect can be greatly increased, with an integrated electromagnetic shielded cable
as described in Fig. 6.6.2.2.1. In addition, magnetic shielding tapes are added outside a large copper or aluminium
sheath: these tapes have a relative permeability of some tens of thousands with negligible eddy current and
hysteretic losses. These tapes have the effect that the sheath current magnitude reaches the conductor current
Page 29
value, and the phase angle between conductor and sheath current is nearly exactly 180 degrees, thus forming an
ideal coaxial cable.
As shown later in paragraph 9.6.2.2, such an integrated cable will fulfil the highest shielding demands. This cable
shows the behaviour of a coaxial transmission line with lowest magnetic fields outside the cable trench and a
complete decoupling of the phases.
PE-Corrosion-Protection
APL-Sheath
Shielding Tapes
Copper-Armouring
XLPE-Insulation & Semi-Conducting Layers

Conductor
Fig. 6.6.2.2.1: Single-core cable with integrated electromagnetic shielding
Page 30
7 INTERACTION WITH NON THERMAL CABLE SYSTEM DESIGN
The use of MF mitigation techniques may affect design parameters non-linked to thermal impact. This interaction is
discussed in this section, considering the reviewed available techniques.
The following table reports possible interactions, which are detailed in further sections. The influence of the design
parameters are assessed with a positive impact (▲= advantage), or negative impact (▼= drawback).
Mitigation Change of Increase of Cable Metallic

Phase splitting
technique cable route laying depth management components
Design parameter (§7.1) (§7.2) (§7.3) (§7.4) (§7.5)
Induced voltages ▲ ▲/▼ ▲/▼
Impedances ▼ ▲ ▲/▼ =
Electrodynamic
stresses and fault ▼ ▲ ▼
current
Fault containment ▲ ▼
Installation ▼ ▲/▼ ▲ ▲/▼ ▲/▼
Mechanical
▼
protection
Stray currents ▼
Operation ▲/▼
Fault location ▼
Repair ▼ ▼ ▼ ▼
Maintenance of
▼
mitigation device
Table 7.1 Summary of interaction of non thermal effects
7.1 Change of cable route
The simplest mitigation technique is to move the magnetic field source away from buildings and people. It means to
modify the cable route in order to avoid the sensitive areas.
7.1.1 IMPEDANCES
▼ The cable route altered from the original project may lead to a significant length increase. The
resulting impedance is increased proportionally.
7.1.2 INSTALLATION
▼ Cable route may be altered from the original project, with unexpected consequences: increased
length and cost, additional obstacles, etc.
The increased length and extra curves (leaving the straight track) cause further pulling efforts.
Page 31
7.2 Increase depth of laying
Another way to move the magnetic source away from people without changing the cable route is to increase the
laying depth.
7.2.1 FAULT CONTAINMENT
▲ In the unlikely event of a breakdown, possible projection of materials or gas are contained by the
soil thickness above the circuits [4]
7.2.2 INSTALLATION
▲ Increased depth of laying would be related to a resulting better protection against external damages
▼ A deeper trench means an increasing volume of excavated soil, longer duration of civil works and
extra cost. It may involve a more powerful trenching machine, plank lining to avoid crumbling, more
difficult handling operations at the bottom of the trench and safety issues.
7.2.3 REPAIR OF THE POWER SYSTEM
▼ A deeper trench leads to more significant digging works, with similar consequences to the previous
item relative to installation.
Cable management implies a position change of the cables: geometry of the circuits (triangular or flat formation)
and axial distance of cables. The closer the conductors, the lower the magnetic field.
7.3.1. INDUCED VOLTAGE
Page 32
Any conductor P, lying in parallel with a set of 3

cables carrying balanced 3-phase circuits will
have a voltage gradient induced along its
length, given by an expression involving
P
several sij (where sij is the axial separation of

conductors i & j).
Ic
 Electra Publications #28 [1] and #128 [2] provide

calculation data to assess induced voltage on
cable systems according the main geometricV induced
configurations (trefoil, flat formation, …) and
Fig. 7.3.1.1 Installation layout and grounding
operating modes (steady-state, short-circuits,
…).
The magnitude of the induced voltage is obviously linked to the geometry of the circuit and especially to the
spacing between the conductors.
▲ All expressions are based on the axial separation s between two conductors. It means that if the
clearance between two conductors is reduced in order to lower MF values, the value of the induced
voltage follows the same trend: compaction is favorable and enables increasing length of
elementary sections, with fewer joints and installation cost benefits.
Moreover, cable management involves also phase rotation or phase sequence, which influences
the induced voltage values. In case of multiple circuits, the optimal combination of phase
arrangement will minimize induced voltages, for a cost free improvement.
7.3.2. IMPEDANCES
The expression of the mutual impedance between two conductors i and j is given by:
  0   0 D
Zmij  j ln( )
8 2  d ij
Where D is the distance to the fictitious return current in the soil.
Once again, values of interest are directly linked to the axial distance dij between two conductors.
▲ Inductances are decreasing with compaction.

Theoretical power transfer capacity is improved.
7.3.3. ELECTRODYNAMIC STRESSES

When a fault occurs, electrodynamic forces may damage parallel cables with major vibration phenomena,
especially for cables installed in tunnel or shafts.
Page 33
Two parallel conductors, with an inter-axis distance s (m), transmitting currents i1 and i2 (A), generate an
electromagnetic force F’ per unit length (N/m):
i1  i2
F '  0
2   s
The magnitude of the force is inversely proportional to the axial separation between cables.
F’ is a force of attraction if the conductors are transmitting currents of same direction, and a force of repulsion if the
currents are of opposite directions.
i1 i2 i2
i1
s s
Fig. 7.3.3.1 Scheme for force interaction
The alternative current is not established instantaneously and a margin safety coefficient (value ≈ 2) is applied to
take into account the aperiodic component of the transient operation.
The choice of cable spacing may affect the

installation design (number and size of collars for
cables in tunnels).
Picture from Cigré 1996 paper (21-

205)
Fig. 7.3.3.2 Installation layout
▼ Compaction increases vibration and mechanical effects in case of short-circuit.
7.3.4 FAULT CONTAINMENT
Page 34
In the unlikely event of a breakdown, closely spaced cables, and especially in free air or tunnels, may be damaged
by the failure (projection of materials or gas). Although touching cables are very efficient to reduce the MF value,
the risk of potential damage propagation is increased.
▼ Compaction increases the risk of damage propagation in case of short-circuits.
7.3.5 INSTALLATION
▲ Compaction may decrease the width of the trench, the volume of excavated soil, the volume of
eventual backfill material. The use of narrow trench leads to possible interesting installation
techniques without wood planking and mechanical laying. Finally, the installation cost may
decrease.
For cables in tunnel, touching cables strapped and hung are widely used.
▼ The configuration in trefoil formation generates lower MF values than flat formation, but
lower cables are less accessible.
7.4 Phase splitting
7.4.1 INDUCED VOLTAGE

▲ All expressions are based on the axial separation s between two conductors. It means that if the
clearance between two conductors is reduced in order to lower MF values, the value of the induced
voltage follows the same trend: compaction is favorable and enables increasing length of
elementary sections, with fewer joints and installation cost benefits.
Moreover, cable management involves also phase rotation or phase sequence, which influences
the induced voltage values. In case of multiple circuits, the optimal combination of phase
arrangement will minimize induced voltages, for a cost free improvement.
▼ Possible hazard due to induced voltages when working on parallel circuits must be taken into
account with strong safety rules.
7.4.2 IMPEDANCES
▼ In replacing large conductors by several conductors of smaller size, the impedances of the circuits
are obviously modified. Smaller size conductors lead to an increased resistive component and
increased self-inductance as well (since self-inductance expression includes a term inversely
proportional to conductor diameter).
The temperature of the conductor has a strong impact on electrical resistivity of the material. The
influence of lower currents shared in several parallel conductors must be evaluated.
▲ Mutual impedances are generally lowered because the mean axial distance between cables is
increased (multiple circuits spread in a larger trench).
Page 35
7.4.3 ELECTRODYNAMIC STRESSES
▲ The magnitude of the stress is inversely proportional to the axial separation between cables (i.e.
compaction is favorable). If a large conductor circuit is replaced by a double circuit with the same
kind of configuration (plastic ducts for example), the axial distance between conductors shall not be
greater than the original one.
The magnitude of the stress is proportional to the square of the fault current.
7.4.4 INSTALLATION
▼ Replace a large conductor circuit by a double circuit of smaller conductor size often implies a wider
trench, more soil to excavate, doubled supply (plastic ducts for example), etc.
The duration of civil work, cable pulling out and accessories mounting will be significantly extended,
with increased cost, nuisance to residents.
More cables mean more terminations, and more space required in a substation or on OHL/UGL
tower.
▲ Cables of smaller conductor size lead to lower pulling efforts, longer delivery length on drums (less
joints), potential less demanding logistics (in terms of maximum constraints, but in accepting a
double volume).
7.4.5 OPERATION
▲ The conversion of one large conductor circuit to multiple circuits of smaller conductors increases
the availability of the global system (n-1 or more contingency).
▼ If joints are considered as potential points of failure (regarding in service statistics), the increase of
the number of joints to connect more cables in parallel will result in higher cumulated risk of
breakdown, with however the convenience of n-1 contingency quoted above .
▼ More conductors may increase the complexity to access cables of parallel circuits.
7.5 Mitigation techniques with metallic components
Beyond methods based on cable management (circuit geometry and phase sequence), mitigation techniques
require metallic component:
 Aluminium or copper conductors for compensation passive loops,
 Metallic plates with high electrical conductivity (eddy currents generating opposite field to the initial field)
 Metallic structures (mostly closed as raceways or steel pipes) with high ferromagnetic permeability to
concentrate and deflect the initial field.
Page 36
The vicinity of these metallic components has to be taken into account for installation and operating.
7.5.1 INDUCED VOLTAGES
Due to the currents circulating in the power cable system, a voltage gradient appears along any metallic
component.
▲ When metallic plates are connected in such a way to make a loop, a circulating current improves
the efficiency of the compensating system.
▼ However, this circulating current generates additional losses.
▼ Special care must be brought during installation or repair for essential safety reasons. Parallel
circuits (overhead or underground lines) may induce hazardous voltages on the MF mitigation
device, even if the circuit of interest is out of service.
▼ A mitigation device may interfere with the installation of a transposed earth conductor. Its position
relative to cables is of high consequence on induced voltage and elementary section design for
bonding. A possible solution is described in Paper [5]
7.5.2 IMPEDANCES
The vicinity of parallel conductors or conduit (earthing conductor, railways, pipes) may affect the value of the power
line impedance. A previous CIGRE study [3] describes the expression of modified impedances.
7.5.3 MECHANICAL PROTECTION
▲ Metallic plates and especially closed structures as steel pipes and raceways bring a visual warning
device and an additional protection against third party damage.
7.5.4 STRAY CURRENTS
▼ Metallic MF mitigation devices may become a path of unexpected stray or short circuit currents.
Special attention must be paid to prevent any risk of corrosion.
7.5.5 ELECTRODYNAMIC STRESSES AND SHORT-CIRCUIT CURRENTS
▼ The possible path of metallic MF mitigation devices for short circuit currents has to be taken into
account with an adapted risk assessment.
7.5.6 INSTALLATION
Page 37
▼ Additional conductors in loop require methodical installation but no specific additional skills, while
the installation of welded plates may be more complex.
The efficiency of the mitigation device depends upon the installation rules: passive loop cable
position compared with power cable system position, electrical continuity of the plates, quality of the
connections and welding, etc.
Hence, the quality of the mitigation device installation requires care and attentive workmanship.
Welding operations may be difficult in bad weather conditions (water in the trench), lack of space
for workers, etc.
Special care must be taken during the installation of horizontal metallic plates to avoid any air
pockets that may result in thermal degradation effect.
When steel pipes are used, plastic pipes may be pulled inside (no skid wires) to allow removal of
the cables at a later date for upgrading or after a failure (fault containment, possibility to remove
one phase, ability of future retrofitting).
▲ General advantages of laying in conduits can be applied to steel pipes by decoupling civil works
and later cable pulling. Only short sections of open trench are required.
▲ Trenchless techniques are possible.
7.5.7 FAULT LOCATION
▼ In case of metallic plates or steel pipes, the mitigation device may interfere with the cable sheath if
the fault location method is based on collection of current leakage.
7.5.8 REPAIR OF MITIGATION DEVICE AND POWER SYSTEM

▼ If the mitigation device must be removed to make the cable system repair possible, plates and
compensation conductors must be placed again with care (to recover the specified initial position
and to comply with the connections for electrical continuity).
▼ The risk of material theft (copper or aluminium) exists during the installation or repair.
7.5.9 MAINTENANCE OF THE MITIGATION DEVICE
▼ The electrical continuity of the mitigation device is essential. Any split or gap between plates (failure
of the welded connection) interrupts the circulating current between plates.
Any damaged ferromagnetic closed structure (raceway or pipe) may have the opposite
consequence of the expected solution: a hole or crack would concentrate the magnetic field outside
(e.g. magnetic leakage of field concentrated in the thickness of steel pipes).
A specific method must be applied to prevent any risk of corrosion of the directly buried equipment
(e.g. cathodic protection).
Page 38
▼ Metallic welded constructions may be sensitive to vibration and fatigue.
A check list for the system operator can be derived from rules applied to cross bonded systems (visual inspection
along the cable route, measurements above the joint bay).
7.5.10 SPECIFIC FEATURES OF METALLIC PLATES
▲ The presence of metallic plates may smooth hot spots due to longitudinal thermal conduction.
▼ Welded plates could prevent natural soil moisture migration.
▼ Special attention must to be paid during the handling and installation of the plates not to damage
the cables.
▼ Air filled pockets, especially underneath horizontal plates, may affect the current rating of the cable
system.
Color Shade Results

Quantity : Temperature Deg. Celsius
Time (s.) : 1E9 Phase (Deg): 0

Scale / Color
15 / 19,97457
19,97457 / 24,94913
24,94913 / 29,9237
29,9237 / 34,89827
34,89827 / 39,87284
39,87284 / 44,8474
44,8474 / 49,82197
49,82197 / 54,79654
54,79654 / 59,77111
59,77111 / 64,74568
64,74568 / 69,72025
69,72025 / 74,69481
74,69481 / 79,66937
79,66937 / 84,64395
84,64395 / 89,61852
89,61852 / 94,59308
Fig. 7.5.10.1 Example: 10-mm air pocket, centered and covering 50% of the plate,
yields 5 °C temperature rise at less than 1 % increase in power loss.
7.5.11 SPECIFIC FEATURES OF STEEL PIPES
▲ The installation of cables into conduits has been managed for decades, thus minimizing the risk of
damage during pulling.
▲ The mechanical protection offered by steel pipes may allow a shallow laying depth.
▼ Three cores must be installed into the same pipe to avoid excessive losses.
▼ Induced metallic sheath voltages and currents may be enlarged.
References
[1] Electra #28, Cigré WG 21.07, “The design of specially bonded cable systems”, 1973.
Page 39
[2] Electra #128, Cigré WG 21.07, “Guide to the protection of specially bonded cable systems against sheath
overvoltages”, 1990.
[3] Cigré TB #95, Cigré WG 36.02, “Guide on the influence of HVAC power systems on metallic pipelines”,
1995.
[4] Cigré Paper 2010 B1-202, “Study of direct burial of high voltage underground cables”, P. Hondaa, N.
Boudinet.
[5] H. Brakelmann: “ Optimisation of compensation conductor systems for magnetic field mitigation near to
high-voltage single-core cables”; 7th International Workshop on LSI Windpower, Madrid (2008), pp. 367
Page 40
8 ECONOMICAL EVALUATION OF MITIGATION MEASURES
For the evaluation and comparison of the costs of mitigation measures, the present value method can be used,
which relates all incurring expenses during the operation lifetime to the present point of time. These expenses may
be the investment costs, the costs of losses and maintenance and possibly the costs for re-investment. Investment
costs may not only be the costs of the mitigation device itself but also for additional measures, as greater conductor
cross sections etc., which are caused by the mitigation measure.
8.1 Economical evaluation
So for example, regarding an internal rate of interest p of the present value method and a rate of price increase tv of
the losses, the present value method valuates the costs of 1 kWh, which appears after x years of operation and
which charges today with E0 , by
(1  t v ) x 1
E x  E0   E0  (8.1)
(1  p) x
(1  a) x
The right-hand side of eq. (8.1) means, that these loss costs are discounted by means of a resultant discounting
rate a. In eq. (8.1) are
tv annual rate of price increase of the losses, and

p internal rate of interest of the present value method.
For example, with an internal rate of interest p = 4.00 % and a rate of price increase of the losses of tv = 2.0 %, the
resultant discounting rate a becomes
 a = 1.961% .
Related to the date of investment (the present point of time; begin of operation), annually recurring loss costs,
which charge today with E0 , accumulate over N years to the present value E:
N
E0 q N 1
E  E   E0  rB where q=1+a. (8.2)
q N  (q - 1)
0
i 1 qi
rB is the present value factor, which must be multiplied by the annual costs E0 to get the present value E.
The method described above is not the only way to evaluate the cost of mitigation measures; other methods can be
adopted according to local practice and experience.
Bibliography
[1] IEC 60853-2 Calculation of the cyclic and emergency current rating of cables. Part 2: Cyclic rating of cables
greater than 18/30 (36) kV and emergency ratings of all voltages
[2] IEC 60287-3-2 Electric cables – Part 3: Section on operating conditions – Section 2: Economic optimization of
power cable size.
Page 41
9 CASE STUDIES / EXAMPLES OF INSTALLATIONS

This chapter reports some examples of implementation of MF mitigation techniques described previously.
Application to different cases will help the designer to form a basic knowledge useful for many conditions. The
examples are not exhaustive but only representative of an expanding technology.
From the point-of-view of comparison of computation and measurement of derating values, it is not easy to obtain
data from a working installation. It is not in the scope of work of this TB to compare computed and measured
values of the circuit rating. Simulations have been carried out with different methods of thermal field analysis of
heat transfer in the soil, but the most of the assumptions are the same as in IEC 60287 or alternatively FEM
methods have been used.
9.1.1 SINGLE CIRCUIT
2
A 110 kV XLPE single-core cable system with a copper conductor of 1600 mm is considered. The cables have an
2
insulation thickness of 18 mm and a copper screen of 120 mm , which is covered by corrosion protection with an
integrated aluminium foil. The external diameter of the cable is 100 mm. Current ratings are calculated for
continuous load. Screen losses because of circulating currents are avoided by means of a cross-bonding of the
copper screens. The cable trench is thermally stabilized by means of a material with a thermal conductivity, which
holds the value of the surrounding wet soil of 1.0 W/(K*m) even for elevated temperatures (dry soil: 0.4 W/(K*m);
critical temperature rise for drying-out onset: 15 K). The ambient temperature is 15°C.
B (x, l) B (x, l)
l l
x x
y y
s s s
Page 42
Fig. 9.1.1.1: Laying arrangements of a single-core cable system

left: triangle formation right: flat formation
In the following, symmetrical arrangements of the three cores of the cable system in a triangle formation (Fig.
9.1.1.1, left side) as well in a flat formation (Fig. 9.1.1.1, right side) are considered. Since complexity and costs of a
cable trench in urban areas depend decisively on the trench depth, the laying depth h of the triangle formation is
defined here by the axes of the both lower cables, - and by this comparable with the flat formation. The laying
depth of the upper cable must be 0.8 m at least. The maximum magnetic induction Bmax is considered for a level in
a height l above ground.
Without restriction of the magnetic field, generally the flat formation, because of its better heat dissipation, is
chosen to achieve high current ratings. For this case, the thin characteristic lines in Fig. 9.1.1.2 and Fig. 9.1.1.3
show the current ratings , as function of the clearance Δs of the cables and their laying depth h, for triangle and for
flat formation, respectively, both without any restriction of the magnetic field.
2
The copper conductor has a section of 1600 mm ;
thin curves: no restriction;
thick curves: restriction of the magnetic induction to B1.0 = 20 μT at a level l = 1.0 m above ground;
dotted, red line: optimal laying depth hopt; optimum for sopt ≈ 0.2 m; hopt ≈ 1.2 m: Iopt ≈ 1650 A.
2000
A h = 1.2 m 2.0 m
hopt 3.0 m
1600
4.0 m
I
1200
1.2 m 2.0 m 3.0 m
800
400
B1.0 = 20 T
0
0 0,2 0,4 0,6 0,8 m 1
s
2
Fig. 9.1.1.2 Continuous current ratings of a 110-kV single-core XLPE-cable, 1600 mm copper conductor, for
the triangle formation of Fig. 9.1.1.1
Page 43
2000
A
h = 1,2 m
2,0 m
hopt
1600
I 3,0 m 4,0 m
1200
800
400
B1.0 = 20 T
0
0 0,2 0,4 0,6 0,8 m 1
s
Fig. 9.1.1.3: As Fig. 9.1.1.2, but for the flat formation of Fig. 9.1.1.1
A comparison between Fig. 9.1.1.2 and Fig. 9.1.1.3 elucidates, that the flat formation has essential thermal
advantages only for small laying depths, where the height of the upper cable is restricted, as for h = 1.2 m: for
greater laying distances the current ratings approaches a value of 1950 A, whereas the triangle formation is
restricted for Δsmax = 0.3 m to approximately 1760 A. For greater laying depths and especially for greater
clearances the triangle formation shows higher current ratings as the flat formation.
The related maximum magnetic inductions are shown in Fig. 9.1.1.4 for the triangle formation and in Fig. 9.1.1.5 for
the flat formation, both for a height of l = 1.0 m above ground (B1.0) and at ground level (B0), respectively. But keep
in mind that the magnetic inductions here refer to varying currents of the full-loaded cables with ratings following
Fig. 9.1.1.2 and Fig. 9.1.1.3
A comparison of the maximum magnetic inductions B1.0 for both arrangements is given in Fig. 9.1.1.6. For close
arrangements, the well-known relation of B for flat to triangle formation of approximately √2 (if the laying depth is
defined for the system axes) can be found. In general, it can be seen that the magnetic inductions of both
formations are quite similar but with the background, that the inducing currents are somewhat greater in the case of
the triangle formation.
Introducing now a restriction of the magnetic field by a limit value of B1.0 = 20 μT in a height of l = 1.0 m above
ground, leads to other results. For the triangle formation, Fig. 9.1.1.3 shows for laying depths up to 3 m, that
enlarging the clearance from close formation first leads to increased current ratings (for thermal reasons). But when
reaching the limit value of 20T for Δs ≈ 0.1 m, for farther enlarged clearances the current ratings decrease rapidly.
This means, that for any laying depth h a maximum current rating is achieved for an optimum clearance Δs of the
cables.
Connecting these optimum points lead to a characteristic line (in red), which for any clearance Δs of the cable
shows the optimum laying depth hopt with respect to current rating.
Similar results are achieved in Fig. 9.1.1.3 for the flat formation. For both formations the results is a maximum
current rating, which is given for the least laying depth of 1.2 m with approximately 1640 A (Δs ≈ 0.13 m) for the
triangle formation and with 1570 A (Δs ≈ 0.08 m) for the flat formation.
Page 44
200
T 2,0 m
h = 1,2 m
150
B
100
B0 B1.0 2,0 m
50
4,0 m
1,2 m
20 T
0
0 0,2 0,4 0,6 0,8 m 1
s
Fig. 9.1.1.4: Maximum magnetic induction B as function of clearance Δs and laying depth h for triangle
formation; B0 induction at ground, B1.0 induction 1.0 m above ground.
200
h = 1,2 m
T
150
2,0 m
B 1,2 m
100
B0 B1.0 2,0 m
50
20 T
4,0 m
0
0 0,2 0,4 0,6 0,8 m 1
s
Fig. 9.1.1.5: Maximum magnetic induction B as function of clearance Δs and laying depth h for flat
formation: B0 induction at ground; B1.0 induction 1.0 m above ground.
Page 45
200
T
150
B1.0
100
2,0 m
50
h = 1,2 m
4,0 m
20 T
0
0 0,2 0,4 0,6 0,8 m 1
s
Fig. 9.1.1.6: Comparison of the maximum magnetic inductions B1.0 (1.0 m above ground) as function of
clearance s and laying depth h for flat and triangle formation; the values of B1.0 are based on the current
ratings in Fig. 9.1.1.2 and Fig. 9.1.1.3 (thin lines).
The described results may change, if the limit value of the magnetic induction or the plane of reference is changed:
this is shown in Fig. 9.1.1.7 (triangle formation) for the case that the limit value of B = 20 T is referred now to the
ground level (l = 0 m: B0).
Now for the minimum laying depth even for a close trefoil arrangement of the cables, the current rating is reduced
because of the limit value B0. The red curve for the optimum laying depth hopt reveals an (weak) absolute optimum
of the current rating of Iopt ≈ 1550 A for Δsopt ≈ 0.5 m and hopt ≈ 3.2 m, but which is already nearly reached for
Δs ≈ 0.13 m and h ≈ 2.0 m with I ≈ 1480 A. Thus, enlarging the laying depth may lead to greater current ratings.
Page 46
2000
A h = 1,2 m 2,0 m
hopt 3,0 m
1600
4,0 m
I
1200
800
3,0 m
400
1,2 m 2,0 m
Bo = 20 T
0
0 0,2 0,4 0,6 0,8 m 1
s
Fig. 9.1.1.7: As in Fig. 9.1.1.2, but: restriction of the magnetic induction by B0 = 20 T at ground
level for the triangle formation in Fig. 9.1.1.1; dotted, red line: optimal laying depth hopt; absolute optimum
for sopt ≈ 0.5 m; hopt ≈ 3.20 m: Iopt ≈ 1505 A
This is further elucidated by Fig. 9.1.1.8, which shows the current rating as a function of the laying depth h with the
clearance s as a parameter. Again, for each fixed clearance s, with enlarged laying depths the current rating is
first rising and then, after arriving at the thermally defined (thin) characteristic line, falling because of the worsening
thermal conditions. Again, connecting the optimum points gives the optimal clearance Δsopt as function of the laying
depth, thus leading to the absolute optimum as before.
Page 47
2000
0.5 m 1.0 m
A
0.2 m
1600
sopt
I s = 0.0 m
1200 0.0 m
0.2 m 0.5 m 1.0 m

800
400
B0 = 20 T
0
0,0 1,0 2,0 3,0 m 4,0
h
Fig. 9.1.1.8 As in Fig. 9.1.1.7, but: continuous current rating for the triangle formation as function of the
clearance Δs; parameter: laying depth h ;dotted, red line: optimal clearance Δsopt
Fig. 9.1.1.9 shows the corresponding results for the flat formation: here the current rating for optimized clearances
Δsopt is continuously rising with increased laying depth h, finding the highest value of I = 1480 for the greatest
discussed laying depth of h = 4.0 m and for a clearance of Δs ≈ 0.6 m.
Finally, Fig. 9.1.1.10 discusses the example of a designated current rating of I ≈ 1500 A, without (a, b) and with (c,
d) a restriction of the magnetic induction of B0 = 20 μT at ground level, comparing the triangle and the flat
formation. First, without the restriction, the triangle formation gives a somewhat smaller cable trench. With the
above mentioned restriction, for the triangle formation the trench becomes broader and essentially deeper. For the
flat formation, the restriction leads to unacceptable dimensions of the cable trench.
Page 48
2000
1.0 m
A 0.5 m
0.2 m
1600
s = 0.0 m sopt
I
1200
0.0 m 0.2 m 0.5 m
800
1.0 m
400
B0 = 20 T
0
0,0 1,0 2,0 3,0 m 4,0
h
Fig. 9.1.1.9 As in Fig. 9.1.1.8, but for flat formation; highest current rating for h = 4.0 m; s ≈ 0.6 m;
I = 1480 A
Bo = 20 T Bo = 20 T
1500 A 1500 A 1505 A 1480 A
1.00 m 0.85 m 1.00 m 2.00 m
a) b) c) d)
1.50 m
2.30 m
0.05 m 0.05 m
4.30 m
0.13 m
0.60 m 0.60 m
Fig. 9.1.1.10: Trench dimension for I ≈ 1500 A; a), b) : no restrictions; c), d) : B0 = 20 μT
Page 49
9.1.2 DOUBLE CIRCUIT
High voltage AC single-core cables for bulk power transmission are frequently designed as twin systems, so that
electrically in each phase two cables are working in parallel. Hereby, the partitioning of the total phase current will
be strongly dependent on the phase sequence as well as on the geometrical arrangement of the six cables with
strong influence on both, the current rating as well as the magnetic field of the twin system.
In the following, for the two most important and common arrangements, the flat formation of Fig. 9.1.2.1a and the
trefoil formation of Fig. 9.1.2.1b, some possibilities of an optimization with respect to current rating and low EMF
shall be systematically investigated.
9.1.2.1 Double circuit in flat formation
2 2
Considered are 380 kV XLPE cables with segmental copper conductors of 2500 mm , 120 mm copper screens
and an outer diameter of approximately 137 mm. The copper screens are cross-bonded or single-point bonded, so
that longitudinal screen currents will be suppressed.
In the laying formation of Fig. 9.1.2.1a, the laying depth of the cables is 1.5 m. The spacing sc between the cables
cores and the spacing sS between the two cable systems will be varied for a pretended total spacing W, which
represents the sum of all spacing W = 4* sc + sS).
The cables are laid in a zone of special backfill, which even for elevated temperatures ensures a thermal
conductivity of 1.0 W/(K m). The bottom level of this zone is yb = 1.8 m, the top level yt = 1.2 m; the side
coordinates of the zone (xl , xr) are adopted to the actual width of the arrangement. The special backfill is
surrounded by normal backfill/soil, which is assumed to have a thermal conductivity of 1.0 W/(K m) in the moist
case and 0.4 W/(K m) in the case of drying-out (two-zone model, comp. [2]). The ambient temperature of the
undisturbed soil is 15°C. Assumed is a daily load cycle with a load factor of m = 0.85.
In a first step, a total spacing of W = 3.5 m (i.e. a total width of the twin system of W ≈ 4.3 m and a trench width of
approximately 4.7 m) shall be pretended. In Fig. 9.1.2.2, the spacing sc between the cable cores is varied
between 0 m (i.e. system wisely close formation) and 0.8 m, thus varying the system spacing sS m
and 0.3 m.
In Fig. 9.1.2.2a the current rating of the twin system is shown as a function of the core spacing sc. Parameter is
the phase sequence of the six cores with:
1: T-S-R—R-S-T (black curves) and 2: R-S-T—R-S-T (blue curves).
The first of these phase sequences ensures (for the symmetrical arrangement of Fig. 9.1.2.1a) a symmetrical split
of the phase currents. The second sequence leads to unsymmetrical currents, but which for smaller core spacing
sc only slightly diverge from the symmetrical case, thus leading to only small rating reductions.
Page 50
a)
x
h
yt
x xr
l
l 2 3 4 5 6
yb
sc ss sc
b)
x
yt
x
l
2 5 xr
l 3 4 6
yb
sc ss sc
Fig. 9.1.2.1 Considered arrangements of a twin system of 380 kV XLPE single-core cables
a) Flat formation b) Trefoil formation
This can be recognised in Fig. 9.1.2.2a for the region up to sc < 0.4 m without any deviation of the two current
rating curves (the maximum deviation from the 50 %- current split remains less than 2 %). For greater cable core
spacing, the split of the phase current for phase sequence 2 becomes more and more unsymmetrical, thus leading
to growing rating reductions.
Fig. 9.1.2.2b shows the resulting maximum magnetic induction Bmax of the twin system in a height of 1.0 m above
ground. First it can be seen, that for both phase sequences the highest magnetic inductions appear for the case
that the core spacing sc and the system spacing sS are nearly equal (≈ 0.7 m). 100 T -the legal limit value in
many countries – is exceeded in this case.
The maximum rating of 3,898 A (2,566 MVA) is reached in Fig. 9.1.2.2a for phase sequence 1 and a core spacing
of sc = 0.55 m, i.e. for a system spacing of ss = 1.3 m. This rating is defined to 100 %. Corresponding to this
rating is a maximum magnetic induction of B100 = 90.5 T, which is correlated in Fig. 9.1.2.2b to a “shielding factor
SF” of 1.0. For other magnetic inductions Bmax the shielding factor may, as usual [e.g. 3...5], be defined by
SF = B100 / Bmax.
Fig. 9.1.2.2b makes clear, that one possible “shielding measure” is the choice of the phase sequence 2 instead of
sequence 1, combined with an adequate combination of the cable spacings. One example:
Choosing a core spacing of sc = 0.40 m, a system spacing of ss = 1.9 m as well as the phase sequence 2, leads
to
 a reduction of the current rating of only 1.3 % (3,849 A/2,533 MVA instead of 3,898 A/2,566 MVA),
accompanied by
Page 51
 a reduction of the losses by 2.0 %, thus leading to

 a reduction of the maximum magnetic induction of 46.1 T, which means a shielding factor of SF = 2.0,
compared with the point of the maximum current rating:

phase sequence 1, core spacing sc = 0.55 m, system spacing ss = 1.30 m with:
3,898 A/2,566 MVA and 90.5 T.
4000
98,7% 100 %
A a) 97,6%
1
3800
92,4%
3,636 A
3600
I 2
3400
84,0%
3200
0,55 m
3000
120
T b) 1
100
SF = 1,0
80
B max 60
40 2,0
2
6,1 2,4
20
4,6 0,55 m
8,8
0
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 m 0,8
sc
Fig. 9.1.2.2 Current rating I (a) of the twin system in flat formation and maximum magnetic induction Bmax
(b) 1.0 m above ground as functions of the spacing sc between the cable cores
t ot a l s p a c i ng :  W = 3 . 5 m ; i n ( a) t h e r e l a t iv e c u r r e nt r a t i ng i s s h ow n a s a p a r a m e t e r;
the shielding factor SF in Fig. 9.1.2.2b is defined for the maximum current rating in Fig. 9.1.2.2a
parameter: phase sequence of the six cores (1: T -S-R—R-S-T ; 2: R-S-T—R-S-T)
Page 52
For this combination of cable spacings and for both phase sequences, Fig. 9.1.2.3 shows the horizontal
distributions of the magnetic induction B in a height of 1.0 m above ground. The strong reduction of the maximum
magnetic induction because of the change to phase sequence 2 can be recognized. In many cases the fact may
not be decisive, that the curve for phase sequence 2 is for some meters directly beside the cable twin system
somewhat higher than the curve of phase sequence 1.
80
T 1
60
B 40
2
20
0
-2 0 2 4 6 8 m 10
x
Fig. 9.1.2.3 Horizontal distribution of the magnetic induction B in a height of 1.0 m above ground for a core
spacing of sc = 0.40 m and a system spacing of ss = 1.9 m phase sequence 1 and 2 (blue curve), total
current I = 3,849 A (2,533 MVA)
In Fig. 9.1.2.4 the question of optimized cable spacings with respect to current ratings is generalized. Fig. 9.1.2.4a
shows the maximum current rating of the twin system for optimized cable spacings as a function of the total
spacing W (i.e. the sum of all spacings). Fig. 9.1.2.4b shows the corresponding optimized relation vopt of the
spacings
vc
vopt 
with
vs .
It becomes obvious, that for very wide cable trenches, this optimal relation tends to vopt = 100 %, whereas for
smaller cable trenches it tends to vopt ≈ 32 %.
The current rating of the twin system can be increased, starting with
 W = 0 m (all cables close together, without spacing): 2,500 A/1,645 MVA to
 W = 3,5 m; sc = 0.55 m; ss = 1.30 m, vopt = 42.3 %: 3,898 A/2,566 MVA and
 W = 7,0 m; sc = 1.32 m; ss = 1.72 m; vopt = 76.7 %: 4,487 A/2,953 MVA .
From the relation factor vopt in Fig. 9.1.2.4 the concrete spacings can be calculated for any total spacing of W by
W
vs 
1  4  vopt vc  vopt  vs
and (9.1.2.1)
Page 53
6000
A
5000
1
4000
I
3000
2000
1000
0
0 4 8 12 16 m 20
100
%
80
vopt 60
1
40
31,8%
20
0
0 4 8 12 16 m 20
W
Fig. 9.1.2.4 Current rating of the twin system in flat formation for phase sequence 1 (a) and optimal relation
vopt of spacings as functions of the total spacing W
9.1.2.2 Double circuit in triangular formation

2
For the same twin system of 380 kV XLPE cables (copper conductor 2500 mm ), now the triangle formation of
Fig. 9.1.2.2.1b shall be discussed. The laying depth h (again 1.5 m) is defined here by the axis of the upper cables
cores. Again the spacing sc between the cables cores and the spacing sS between the two cable systems will be
varied for a pretended total spacing W, which represents the sum of all spacings (W = 2* sc + sS). The zone of
special backfill has again a top level of yt = 1.2 m, whereas its bottom level yb as well as the side coordinates (xl ,
xr) are adopted to the actual width of the arrangement. All the rest of the parameters are identical with those of
chapter 1.
Again in a first step, a total spacing of W = 3.5 m (i.e. a total width of the twin system of W ≈ 3.9 m and a trench
width of approximately 4.3 m) shall be pretended. In Fig. 9.1.2.6, the spacing sc between the cable cores is varied
Page 54
between 0 m (i.e. system wisely close formation) and 0.8 m, thus varying the system spacing sS m
and 1.9 m.
In Fig. 9.1.2.2.1a the current rating of the twin system is shown as a function of the core spacing sc. Parameter is
again the phase sequence of the six cores with: 1: T-S-R—R-S-T (black), 2: R-S-T—R-S-T (blue) and 3: S-T-
R—R-S-T (red). In contrast to the flat formation, here another phase sequence, namely the phase sequence 3,
shows the lowest magnetic fields, whereas its current ratings are nearly identical with those for phase sequence 1.
The maximum of the current rating is very flat, compare Fig. 9.1.2.2.1a. The maximum rating is given with
3,799 A/2,500 MVA for phase sequences 1 and 3, a core spacings of sc = 0.65 m and a system spacing of
ss = 2.20 m. But the maximum magnetic inductions in this case become 45.4 T for phase sequence 1 and
33.3 T for phase sequence 3, comp. Fig. 9.1.2.2.1b. As compared with the flat formation and phase sequence 1,
this means a reduction of the maximum current rating of 2.5 %, but accompanied by a magnetic field reduction by a
factor of 2.7, if phase sequence 3 is chosen.
Whereas the magnetic inductions are strongly reduced, compared with the flat formation of the twin system, Fig.
9.1.2.2.1b shows only small differences for the phase sequences 1 and 2, but sensibly lowers magnetic inductions
for phase sequence 3, especially near to the maximum of the current rating. Again, the maximum rating of
3,799 A/2,500 MVA is defined to 100 %, with a corresponding maximum magnetic induction of B100 = 45.4 T for
phase sequence 1, which is correlated in Fig. 9.1.2.2.2 to a “shielding factor SF” of 1.0.
4000
A a) 1, 3
99,2 % 100 %
3800
3,636 A 98,8% 2
3600
I
3400
85,4%
3200
0,65 m
3000
120 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8
T b)
100
80
B max
60
SF = 1,0
40
2 1
1,8
1,4
20
2,0
3 0,65 m
0
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 m 0,8
sc
Fig. 9.1.2.2.1b Current rating I (a) of the twin system in triangle formation and maximum magnetic induction
Bmax (b) 1.0 m above ground as functions of the spacing sc between the cable cores total spacing:
Page 55
W = 3.5 m; in (a) the relative current rating is shown as a parameter; the shielding factor SF in Fig.
9.1.2.2.1b is defined for the maximum current rating in Fig. 9.1.2.2.1a parameter: phase sequence of the six
cores: 1: R-S-T—T-S-R; 2: R-S-T—R-S-T 3: S-T-R—R-S-T
Changing sc = 0.65 m, ss = 2.2 m; 3,799 A/2,500 MVA;
44.5 T and 33.3 T, respectively) to phase sequence 3, a core spacing of sc = 0.30 m and a system spacing of
ss = 2.9 m, leads to
 a reduction of the current rating of only 0.8 % (3,768 A/2,480 MVA), accompanied by
 a reduction of the losses by 2.7 %, thus leading to
 a reduction of the maximum magnetic induction of 22.5 T, which means a shielding factor of SF = 2.0.
Comparing this solution with the rating optimum of the flat formation, see chapter 1 (phase sequence 1;
sc = 0.55 m; ss = 1.3 m; 3,898 A/2,566 MVA; B100 = 90.5 T), and correlating the latter to a “shielding factor of
SF = 1.0, will lead to the conclusion, that by means of the following measures:
 trefoil instead of flat formation,

 phase sequence 3 and
 combination of spacings sc = 0.30 m, ss = 2.9 m
a shielding factor of SF = 4.0 can be achieved, accompanied by a reduction of the current rating by 3.3 % and a
reduction of the losses by 7.1 %.
In Fig. 9.1.2.2.2 the question of optimized cable spacings in the trefoil formation of Fig. 9.1.2.2b with respect to
current ratings is generalized. Fig. 9.1.2.2.2a shows the maximum current rating of the twin system for optimized
cable spacings as a function of the total spacing W (i.e. the sum of all spacings), and Fig. 9.1.2.2.2b shows the
corresponding optimized relation vopt of the spacings.
Fig. 9.1.2.2.2
Current rating of the twin system in

triangle formation for phase sequence
1 or 3 (a) and optimal relation vopt of
spacings as functions of the total
spacing W.
Page 56
6000
A a)
5000
1
4000
I
3000
2000
1000
0
0 4 8 12 16 20
100
% b)
80
v opt 60
40
1
20
0
0 4 8 12 16 m 20
W
It becomes obvious, that for very wide cable trenches up to 8.0 m, this optimal relation tends to a value of
approximately vopt = 32 % (system spacing three time larger than core spacing), whereas for small cable trenches it
tends to vopt ≈ 100 % (system spacing equals core spacing).
An enlargement of the total spacing W (i.e. the trench width) as a means to enlarge the current rating seems only
to be interesting up to W = 6...8 m within a rating range of 2,800 A/1,840 MVA up to approximately4,200 A/2,760
MVA.
9.1.2.3 Conclusions
High voltage AC single-core cables for bulk power transmission are frequently designed as twin systems, for which
the influence of the two most important parameters, phase sequence and laying geometry, on the current ratings
as well as on the magnetic fields have been discussed. For the most important laying arrangements and phase
sequences, possibilities of an optimized design with respect to current rating and low EMF are elucidated.
It can be shown, that an optimized design of twin systems may provide reductions of the magnetic field by
“shielding factor” of approximately SF = 2.0...4.0, accompanied by a reduction of the current ratings in a range of
1...4 %, but also with reductions of the system losses by 2...7 %. For both optimization goals, the correct
combination of core spacing, system spacing and phase sequence are turning out to be decisive. Table 9.1.2.3.1
gives a summarizing comparison.
Page 57
triangle triangle compared with flat

phase v=
sequence sc sS sc sS I/Imax P´/P´max SF I/Imax P´/P´max SF
m m % % % % %
1, 3 0.65 2.30 28.3 (= vopt, 1,3) 100.0 100.0 1.0 97,5 95.0 2.0
1 0.30 2.90 9.7 99.2 95.1 1.6 96.7 93.5 3.2
2 0.30 2.90 9.7 98.8 97.4 1.8 96.3 92.8 3.6
3 0.30 2.90 9.7 99.2 98.4 2.0 96.7 93.5 4.0
Table 9.1.2.3.1: Relative currents, shielding factors SF and related losses P´/P´max turning from flat to trefoil,
sc = 0.30 sS = 2.90 m leads to: SF = 4.0; current reduction by 3.3 %; loss
reduction by 6.5 %, but the trench becomes deeper.
9.1.2.4 References
[1] CIGRE, Characterisation of ELF magnetic fields, TF C4.205, CIGRE-publication 320, Paris, April 2007
[2] IEC 60287, Calculation of the continuous current ratings of cables (100 % load factor), IEC 60287, 1982
[3] H. Brakelmann, Magnetfeldreduktion durch Zusatzleiter in Energiekabeltrassen, Elektrizitätswirtschaft (1996), S.

274 - 279
[4] P. Maioli, E. Zaccone, Passive Loops Technique for Electromagnetic Field Mitigation: Applications and
Theoretical Considerations Jicable-Conf. 2007, Versailles, pp. 231-236
th
[5] H. Brakelmann, Optimisation of Compensation Conductor Systems for Magnetic Field Mitigation, 7 Intern.
Workshop on Large-Scale Integration of Wind Power, Madrid, 2008, pp. 367-374
[6] H. Brakelmann, J. Hoeffelman, “Optimization of HVAC cable twin systems with respect to current rating and
EMF”, CIGRE-Conf. On EMF, Sarajevo, Juni 2009
9.1.3 SPLIT PHASES
An example configuration for quantification of de-rating effect of some alternative shielding techniques was
defined by the WG. Coupled electro-thermal analyses of a number of comparative cases were performed using
two-dimensional FEA, and the results are presented in this section. Although the focus is on conductor
management, plate shielding and passive loop shielding cases are also included for direct comparison of (de)rating
effects.
Cable conductors are in principle modelled as copper bolts, but with an enforced uniform current density. Each bolt
is assigned maximum DC resistance in accordance with IEC 60228, and the diameter obtained by taking the cross-
sectional designation in IEC 60228 literally. The resulting model should be representative, while still on the
conservative side, for a Milliken-type conductor, although there is obviously room for some refinement. (For
example, a real life, stranded conductor will have a larger diameter, and thus a lower heat flux density at its
surface.)
FLAT, UNSHIELDED
Page 58
The first case is used as reference for the subsequent configurations, addressing alternative field mitigation
methods. A flat configuration with thermal conditions as illustrated in Fig. 9.1.3.1 is defined as a base case for
comparison, with geometrical data as shown in Fig. 9.1.3.2. The indicated shielding plate has identical properties
as the surrounding soil for this first case. Note that Table 9.1.3.1 at the end of this section summarizes all case
results and compares impact of different mitigation approaches, both on EMF and on current rating.
-2 -1
Heat convection coefficient at ground surface: 25 W m K
Color Shade Results


Scale / Color
15 / 19,68007
19,68007 / 24,36014
24,36014 / 29,04021
29,04021 / 33,72028
33,72028 / 38,40034
38,40034 / 43,08041
43,08041 / 47,76048
47,76048 / 52,44055
52,44055 / 57,12062
57,12062 / 61,80069
61,80069 / 66,48076
66,48076 / 71,16083
71,16083 / 75,8409
75,8409 / 80,52097
80,52097 / 85,20103
85,20103 / 89,8811
Thermal resistivity of soil: 1.0 K m W -1 (uniform)
Fig. 9.1.3.1 Thermal steady state results for the reference case at I = 1125 A (left), and conductors
9 3
temperature (right). 1E9 s = 10 seconds = 12·10 days = 32 years
Based on the results of the coupled electromagnetic-thermal analysis, the base case obtains a current rating of
Location of shielding plate (inactive for this case)
1125 A. The time needed to reach thermal steady state at continuous/constant current is approximately one year,
as illustrated by Fig. 9.1.3.1 (right). Identical thermal properties and boundary conditions as specified in Fig. 9.1.3.1
are also assigned to all subsequent cases.
The corresponding flux density (BR) above ground surface is shown at the left-hand side of Fig. 9.1.3.2, which also
includes a snapshot of the magnetic flux lines with north-south indicated by the purple line. (Note that the flux line
Thermal resistivity of polymers: 3.5 K m W -1 (PE)
colours are not associated with the colour legend which applies for the colour contour plot of BR.) The magnetic flux
lines are included for improved visualisation of the conceptual working of the alternative mitigation techniques
applied in subsequent cases.
↓ Thermal boundary: 15 °C (constant) at 10 m depth
The right-hand side of Fig. 9.1.3.2 shows the magnetic field (BR) above the unshielded cable circuit as calculated
by (from top)
a) the far-field approximation (as given in TB 373) at soil surface (0 m elevation)
b) FEA results at soil surface
c) FEA results 1 m above soil surface
Graphs a) and b) are overlapping such that the former is barely visible near x=0. The difference between the two
methods at this location is 1 µT (2 %), showing that the near-field effects are negligible at soil surface and above
for this case.
Page 59
↑ Ground surface Color Shade Results

BR [µT]
Quantity : User

Scale / Color
235,48402E-12 / Isovalues
2,83033E-6 Results
2,83033E-6 / Quantity
5,66042E-6
: Equi flux Weber
5,66042E-6 / Time8,49051E-6
Depth of shielding plate = 1300 mm 8,49051E-6 / Line
(s.) : 1E9 Phase (Deg): 60
11,3206E-6
/ Value
11,3206E-6 / 114,15069E-6
/ -1,03452E-3
Width of shielding plate = 800 mm 14,15069E-6 / 23 16,98078E-6
/ -905,20497E-6
/ -775,88996E-6
16,98078E-6 / 4 19,81086E-6
/ -646,575E-6
Thickness of shielding plate = 3 mm 19,81086E-6 / 5 22,64096E-6
/ -517,25999E-6
6 / -387,94498E-6
22,64096E-6 / 7 25,47105E-6
/ -258,6291E-6
25,47105E-6 / 8 28,30114E-6
/ -129,3141E-6
28,30114E-6 / 9 31,13123E-6
/ 0
10 / 129,3141E-6
31,13123E-6 /11 33,96132E-6
/ 258,6291E-6
33,96132E-6 /12 36,79141E-6
/ 387,94498E-6
36,79141E-6 /13 39,6215E-6
/ 517,25999E-6
14 / 646,575E-6
39,6215E-6 / 42,45159E-6
Depth of cable axes = 1450 mm 42,45159E-6 / 45,28168E-6
Axial phase distance = 250 mm

Gap between cables = 170 mm
Fig. 9.1.3.2: Unshielded reference case for the efficiency evaluation of plate shielding, trefoil and split-
phase conductor management: I = 1125 A.
2
FLAT, SHIELDED
Conductor size = 800 mm (Cu)
Polymer thickness = 24 mm
As a first, and purely theoretical/informative, case modification, the shielding plate is assigned thermal properties of
copper, while its electric conductivity is kept at zero. That is, a lossless, thermally conductive plate having zero
impact on EMF, as magnetic induction is unable to generate circulating currents in the electrically non-conductive
plate.
The expected net effect of the plate is then improved cooling of the cables, and in particular the centre cable,
assuming that a thermally highly conductive shielding plate will smooth out thermal gradients. The computed
temperature distribution at identical current as the above reference case is shown at the left-hand side of Fig.
9.1..3.2. The results clearly illustrate that the plate does impact heat flow and temperature distribution in soil, and
that the temperature of the centre cable is reduced – albeit by a modest 2 °C. The two outermost cables both see a
temperature reduction less than 1 °C.
Color Shade Results

Color Shade Results Quantity : Temperature Deg. Celsius
Time (s.) : 1E9 Phase (Deg): 0 Scale / Color
Scale / Color 15 / 19,68231
15 / 19,53816 19,68231 / 24,36462
19,53816 / 24,07633 24,36462 / 29,04692
24,07633 / 28,61449 29,04692 / 33,72923
28,61449 / 33,15266 33,72923 / 38,41154
33,15266 / 37,69082 38,41154 / 43,09385
37,69082 / 42,22899 43,09385 / 47,77616
42,22899 / 46,76715
47,77616 / 52,45847
46,76715 / 51,30531
51,30531 / 55,84348
52,45847 / 57,14077
55,84348 / 60,38165 57,14077 / 61,82308
60,38165 / 64,91981 61,82308 / 66,50539
64,91981 / 69,45797 66,50539 / 71,1877
69,45797 / 73,99614 71,1877 / 75,87001
73,99614 / 78,5343 75,87001 / 80,55231
78,5343 / 83,07246 80,55231 / 85,23462
83,07246 / 87,61063 85,23462 / 89,91693
Fig. 9.1.3.3. Thermal results after introduction of a lossless (left, at 1125 A) and a lossy (right, at 1093 A)
copper plate for EMF shielding.
The next step is to also assign the electromagnetic properties of copper to the shielding plate. This introduces the
EMF shielding effect originating from induced currents, but also active power loss in the plate. For this case, refer
to Fig. 9.1.3.3 (right), cable current must be reduced to 1093 A (97 % of the reference case) in order to limit the
steady state conductor temperature to 90 °C. Comparison with the previous cases reveals that, although the
thermal conductivity of the shielding plate does yield a reduced thermal resistance of soil, its inherent power loss
yields a net increase in cable temperatures. For some larger distance between shielding plate and cables the net
effect might be the opposite, but at the same time the plate must be close to the cables if it is to provide any
significant reduction of the magnetic field.
Page 60
Fig. 9.1.3.4 illustrates the shielding effect produced by induced, circulating currents in the shielding plate.
Comparing flux line with those in Fig. 9.1.3.2 reveals a “field repelling effect” above the cable circuit. The power
loss density shown in the left-hand part of Fig. 9.1.3.4 also shows where the induced, circulating current flows, and
reveals that the shielding effect is approaching that of a passive loop – refer to section 6.2 and section 9.2.
Color Shade Results

Quantity : Power density W/(cubic m)

Scale / Color
2,12224E3 / 2,84259E3 Isovalues Results
2,84259E3 / Quantity
3,56294E3
: Equi flux Weber
3,56294E3 / Line 4,28329E3

/ Value
1 / -945,79899E-6
4,28329E3 / 32 5,00364E3
/ -827,57411E-6
/ -709,34923E-6
5,00364E3 / 54 5,72398E3
/ -591,12435E-6
/ -472,8995E-6
6 / -354,67462E-6
5,72398E3 / 7 6,44433E3
/ -236,44975E-6
8 / -118,22487E-6
6,44433E3 / 109 7,16468E3
/ 0
/ 118,22487E-6
7,16468E3 / 1211 7,88503E3
/ 236,44975E-6
/ 354,67462E-6
7,88503E3 / 13 8,60538E3
/ 472,8995E-6
8,60538E3 / 9,32572E3
9,32572E3 / 10,04607E3
10,04607E3 / 10,76642E3
Color Shade Results
10,76642E3 / 11,48677E3 BR [µT]
Quantity : User
11,48677E3 / 12,20712E3
12,20712E3 / 12,92747E3 Time (s.) : 0 Phase (Deg): 60
12,92747E3 / 13,64781E3 Scale / Color
Isovalues Results
139,5971E-12 / 2,83024E-6 Quantity : Equi flux Weber
2,83024E-6 / 5,66033E-6Line Time (s.) : 0 Phase (Deg): 60

/ Value
5,66033E-6 / 8,49043E-621 // -492,62238E-6
-562,99701E-6
8,49043E-6 / 11,32052E-643 // -351,87314E-6

-422,24774E-6
5 / -281,49851E-6
11,32052E-6 / 14,15062E-6 6 / -211,12387E-6
7 / -140,74925E-6
14,15062E-6 / 16,98072E-6 8 / -70,37463E-6
9 / 0
10 / 70,37463E-6
16,98072E-6 / 19,81081E-6 11 / 140,74925E-6
12 / 211,12387E-6
19,81081E-6 / 22,64091E-6 13 / 281,49851E-6
14 / 351,87314E-6
22,64091E-6 / 25,471E-61516 // 422,24774E-6
492,62238E-6
25,471E-6 / 28,3011E-6 17 / 562,99701E-6
28,3011E-6 / 31,1312E-6
31,1312E-6 / 33,9613E-6
33,9613E-6 / 36,79139E-6
36,79139E-6 / 39,62149E-6
39,62149E-6 / 42,45158E-6
42,45158E-6 / 45,28168E-6
Fig. 9.1.3.4: Power loss density in the 3-mm thick copper shielding plate (left), and magnetic field (BR)
above soil surface (right). Magnetic flux lines (snapshot) are not related to colour legends.
The (thermally based) 3 % reduction of cable current will instantly yield 3 % reduction of magnetic field, when all
relevant material properties are exhibiting linear behaviour. At ground surface directly above the centre cable a SF
of 3.0 is obtained, as illustrated by Fig. 9.1.3.5 (right). Note that this value also includes the effect of the 3%
current derating.
Fig. 9.1.3.5 Resultant magnetic field (BR) along ground surface and at 1 m elevation with 3 mm copper plate
(left), and shielding factor obtained when including current derating effect (right); I = 1093 A.
Within ±10 m from the cable trench centre shielding factors between 2.3 and 3.0 are obtained by this very basic
mitigation device. The highest SF is obtained directly above the trench. Comparison with Fig. 9.1.3.2 (right) reveals
that the corridor width for which BR > 1 µT is reduced from roughly 20 m to 13 m.
One final observation related to Fig. 9.1.3.5 is the presence of a small horizontal shift/offset of curves, yielding
graphs that are slightly asymmetrical on the y-axis. This effect is correct, and originates from the non-symmetrical
induction effects of a rotating magnetic field sweeping across a flat plate at a fixed location. Magnetic flux lines
shown in Fig. 9.1.3.4 illustrate that the net effect on magnetic field should not be expected to be symmetrical on x =
0.
Page 61
Results at 1 m elevation are similar to those at ground level, but for the near-field achieved SFs are somewhat less
than at ground surface. The near-field shielding effect of the plate decreases with distance, while the far field
shielding effect is the same.
Table 9.1.3.1 provides some additional results, and also compare results obtained with 5 mm thick aluminium and
copper plates, plus trefoil and split-phase configurations.
TREFOIL
Shifting from a flat formation to a trefoil arrangement as illustrated by the left-hand side of Fig. 9.1.3.6 is an
example of EMF mitigation by conductor (re)arrangement. Keeping the same trench depth and inter-axial distance
as the reference case, there is no derating associated with the unshielded trefoil configuration, but only a very
modest SF = 1.2 is achieved. (The predicted SF = 1.41 is not achieved because the centre cable is shifted upwards
relative to the flat configuration.)
SFs obtained by introducing a conductive shielding plate are smaller than for the flat configuration. The shielding
plate is shifted upwards to ensure the same distance from it to the nearest cable as for the reference case, while
shielding plate dimensions are kept unchanged.
Color Shade Results

Quantity : User
Time (s.) : 1E9 Phase (Deg): -90 Color Shade Results

Scale / Color Isovalues Results Quantity : User
Quantity : Equi flux Weber
38,0701E-9 / 2,8658E-6 Line Time (s.) : 1E9 Phase (Deg): -90

/ Value Time (s.) : 1E9 Phase (Deg): 90
2,8658E-6 / 5,69352E-6 32 // -621,0111E-6
-828,01597E-6
Scale / Color Isovalues Results
5,69352E-6 / 8,52125E-6 54 // -207,00399E-6
-414,00798E-6 Quantity : Equi flux Weber
354,76401E-9 / 3,1627E-6
8,52125E-6 / 11,34897E-667 // 0207,00399E-6 3,1627E-6 / 5,97063E-6
Line / Value
8 / 414,00798E-6 1 / -838,46803E-6
11,34897E-6 / 14,1767E-6109 // 621,0111E-6
828,01597E-6 5,97063E-6 / 8,77856E-6 2 / -733,65954E-6
3 / -628,8501E-6
14,1767E-6 / 17,00442E-6 8,77856E-6 / 11,58649E-64 / -524,04252E-6
5 / -419,23401E-6
17,00442E-6 / 19,83215E-6 11,58649E-6 / 14,39443E-6
6 / -314,4255E-6
7 / -209,61701E-6
8 / -104,8085E-6
19,83215E-6 / 22,65987E-6 14,39443E-6 / 17,20236E-6
9 / 0
10 / 104,8085E-6
22,65987E-6 / 25,4876E-6 17,20236E-6 / 20,01029E-6
11 / 209,61701E-6
12 / 314,4255E-6
25,4876E-6 / 28,31533E-6 20,01029E-6 / 22,81822E-6
13 / 419,23401E-6
14 / 524,04252E-6
28,31533E-6 / 31,14305E-6 22,81822E-6 / 25,62615E-6
15 / 628,8501E-6
16 / 733,65954E-6
25,62615E-6 / 28,43408E-6
17 / 838,46803E-6
31,14305E-6 / 33,97078E-6
28,43408E-6 / 31,24202E-6
33,97078E-6 / 36,7985E-6 31,24202E-6 / 34,04995E-6
36,7985E-6 / 39,62623E-6 34,04995E-6 / 36,85788E-6
39,62623E-6 / 42,45395E-6 36,85788E-6 / 39,66581E-6
42,45395E-6 / 45,28168E-6 39,66581E-6 / 42,47374E-6
42,47374E-6 / 45,28168E-6
Fig. 9.1.3.6 BR for unshielded trefoil (left) and shielded trefoil (right).
The same colour scales as for the flat case have been applied in the figures, and again reference is made to Table
9.1.3.1 for comparison with the other cases.
SPLIT-PHASE
Before moving on to split-phase it is necessary to include a comment related to the term “cross-sectional area”. As
specified in IEC 60228, clause 2.2, the “nominal cross-sectional area” is not subject to direct measurement – it is in
fact the maximum resistance value which defines “conductor size”. The “nominal cross-sectional area” must thus
not be taken literarily, but be considered as a name only. As a consequence, replacing a single conductor by two
parallel “half-size” conductors will generally not result in identical per-phase DC resistance.
2
The specific relevance to these computational examples is the fact that the DC resistance of two parallel 400 mm
2
copper conductors is 6% higher than the DC resistance of a single 800 mm conductor (according to IEC 60228).
Consequently, power loss and temperature rise will be significantly worse for a split-phase configuration, but, to
obtain a most realistic comparison, maximum resistance values defined in IEC 60228 are nevertheless applied.
The geometry of two split-phase examples is shown with thermal steady state results in Fig. 9.1.3.8. The depth of
the trench is identical for all cases in this section. Obtained shielding factor values for these split-phase
arrangements are shown in Fig. 9.1.3.9 (jagged curves imply that a finer mesh is needed for optimum result quality,
but values are still representative).
Page 62
Fig. 9.1.3.8 Soil temperature plots for split-phase, hexagonal (left) and “low-reactance” (right)
arrangements at same trench depth as base case.
Fig. 9.1.3.9 Shielding factor for split-phase hexagonal (left) and “low-reactance” (right) configurations.
As results in Table 9.1.3.1 show, the built-in cancellation principle of the split-phase configuration results in a much
reduced magnetic field. Compared to derating and shielding factors obtained through (very basic) plate shielding,
phase splitting seems extremely effective:
1. Reducing the current per conductor to approximately half immediately halves the maximum of the magnetic
field close to any conductor (near-field).
2. Connecting geometrically opposing cables may be viewed as creating two trefoils that are rotated 180°
relative to each other. Each trefoil will generate a rotating magnetic field of about half the magnitude of the
unshielded trefoil case above. Given the 180° rotation, the two fields cancel each other out very effectively
beyond a certain distance (far-field).
3. As described in TB 373 the magnetic field from a split-phase configuration will decay with distance cubed.
This implies that the sensitivity with respect to burial depth is much higher for a split-phase arrangement
than for conventional flat or trefoil configurations.
4. EMF mitigation is achieved by means of high-voltage cables alone, i.e., there are no added metallic
elements that might lose integrity due to corrosion.
5. The current rating is slightly higher for the hexagonal split-phase configuration, despite the more than 6%
higher DC resistance, due to Joule losses being more diffused/dispersed.
6. In case of non-Milliken conductors and/or non-cross-bonded metallic screens, the split-phase arrangement
implies reduced power loss due to smoother conductor density and lower induced screen current.
Practical downsides of phase splitting are mostly related to installation and repair. It is not feasible to determine on
a general basis whether phase splitting or an alternative method will be more cost-effective. However, if a shielding
Page 63
factor above 10 is required, it is believed that phase splitting may have significant benefits, including zero (or even
negative) de-rating for the same amount of conductor material.
CASE SUMMARY
Table 9.1.3.1 De-rating case simulation results.
As the results in Table 9.1.3.1 show, the two split-phase configurations are by far the most efficient methods of
reducing the magnetic field. Furthermore, the hexagonal configuration introduces an up-rating rather than a
derating, although the per-phase DC resistance is 6 % higher (in accordance with IEC 60228).
Plate shielding efficiency can be improved by adding more plates, but it seems challenging to reach a shielding
factor significantly larger than 10. Furthermore, additional derating is likely to result if more plates are added, given
the introduction of more additional power loss. There is also a risk of air pockets forming underneath a horizontal
plate (soil shifting). In order for plate shielding to be effective, plates must be located close to the cables, and this
location also implies significant power dissipation and derating.
One of the main questions to be asked with respect to EMF mitigation is where shielding required, or should be
most effective:
a) directly above the cables, or
b) at a certain distance on both sides (corridor).
Some mitigation techniques can be more effective for one, but less effective for the other.
9.1.4 SIX-PHASE SYSTEMS
The proposed transmission system uses six single core cables connected to two conventional AC systems with
opposite polarity. Those six-phase bipolar AC systems offer the opportunity to lay one pair of cables simultaneously
in one trench. Because the sum of the currents in one pair of cables is
Zero, the resulting magnetic field is very low and, hence, the three pairs of cables are magnetically decoupled. The
concept uses conventional components.
As shown in [5, 6] there is the possibility (by means of a Scott circuit) to switch a faulty six phase system to a
symmetric four phase system, to continue the transmission of 75.5% of the produced power. All unsymmetrical
interim states are shown to be acceptable for the high voltage grid for a certain time, corresponding to the
standards. Owing to this, in the case of a cable failure it is possible to transmit16% more power than using a
double-circuit-line. This can be done until the restoration of the 6-phase-system. The higher investment for this are:
six extra connections with one circuit breaker and two switches each, transformers with two extra connections and
one extra bus (for the Scott circuit).
Page 64
As an example, a (n-1)-sure transmission power of 1500 MVA (load factor m = 0.7) shall be realized by means of
two 380-kV-XLPE-cable systems. The geometry of the cable trench is shown in Fig. 9.1.4.1. The cable trench has
thermal conductivities of 1.0 W/(K m) in the wet situation and of 0.4 W/(K m), if the critical temperature rise for a
partial drying-out is exceeded. The cable are laid in a thermally stabilized region with a thermal conductivity of
1.0 W/(K m). The ambient temperature is 15°C. The cable screen are cross-bonded.
2
With these parameters, the 380-kV-XLPE-cables must be equipped with copper conductors of 2500 mm cross-
sections, comp. Fig. 9.1.4.1. In the case of a cable failure, one cable system is left with a transmission capacity of
1,515 MVA.
6*2500 mm2 Cu
2*1364 MVA
1*1515 MVA
2
Fig. 9.1.4.1 Considered installation with two 380-kV-XLPE-cable systems (2500 mm copper conductors,
load factor 0.7; with cross-bonding and thermal stabilisation trench width 3.0 m; laying depth 1.5 m;
distance of axes 0.5 m between the cores and of 1.0 m between the systems
As an alternative, Fig. 9.1.4.2 shows a six-phase system with six cable cores, which are arranged as three string
with two cables cores, side-by-side, in each string. Now in the case of a cable failure, two cable strings are left with
four cables cores. This faulty six phase system will be switched to a symmetric four phase system by means of a
Scott-circuiting of the transformers. The four remaining cables cores enable a transmission capacity of 1,606 MVA,
2
which would allow to reduce the conductor cross section to 1,600 mm or, alternatively, to choose aluminium
2
conductors of 2,500 mm . In this case, the transmission power of the undisturbed six-phase system is 2.208 MVA.
6*1600 mm2 Cu (6*2500 mm 2 Al)
2*1104 MVA
1*1606 MVA
2
Fig. 9.1.4.2 As in Fig. 9.1.4.1, but: six-phase system with six 380-kV-XLPE-c able cores with 1,600 mm
2
copper conductors or, alternatively, with 2500 mm aluminium conductors (right hand side)
Page 65
Cost savings because of the conductor material are enabled in a magnitude up to 700 T€ (for aluminium
conductors) per kilometre. In addition, lower eddy losses will result from reduced skin- and proximity effects.
Further advantages of such six-phase transmission systems is the simple cross-bonding (only between two cores
of a string), the electromagnetic decoupling of the strings with very low and symmetrical transmission reactances,
as well as the possibility of bulk power transmission accompanied by extremely low magnetic fields, independent
on the laying distance between the three cable strings.
Because of the large laying distances, comp. Fig. 9.1.4.1, the described conventional three-phase twin system
causes high magnetic inductions near to the cable trench. This is shown in Fig. 9.1.4.3 for the (n-1)-operation with
1500 MVA. The magnetic induction in a height of 1.0 m above ground reaches a value of 56 T and of 125 T in a
height of 0.2 m. For comparison, the corresponding distributions of a 380-kV-overhead line is shown (in red).
Obviously, the highest values, direct under the overhead line, are lower than those directly above the cable. On the
other hand, for distances of more than some meters, the magnetic field of the overhead line becomes by far larger
than that of the cable system.
The corresponding distributions for the (n-1)-operating four-phase system is shown in Fig. 9.1.4.3, too. It can be
seen, that because of the adjacent cable cores of each string (with the total current of zero) the magnetic field is by
far lower than that of the normal three-phase system. This is independent of the large laying distances of the
strings. Magnetic inductions of less than e.g. 1 T can be hold for horizontal distances of less than 5 m. For normal
three-phase systems, comparable values can only be hold for trefoil arrangements, but with the consequence of
sensibly reduced current ratings.
150
(n-1)-operation:
T
B0,2m 1*1500 MVA
125
100 x
75
B
50
x
B1,0m
25
0
0 5 10 15 m 20
x
Fig. 9.1.4.3 Magnetic induction during (n-1)-operation with 1,500 MVA in a height of 1.0 m above ground
(thick characteristic curves) and in 0.2 m (thin curves) for comparison: overhead line (in red, [5])
Page 66
References to 6.1.4 and 9.1.4
[1] H. Brakelmann, M. Jensen, “Neues sechsphasiges Übertragungssystem für VPE-isolierte HVAC- See- und
Landkabel hoher Übertragungsleistung,“ Ew (2006) , H. 4, S. 34-43
[2] H. Brakelmann, C. Burges, M. Jensen, Th. Schütte, “Bipolar Transmission Systems with XLPE HVAC
Submarine Cables,” 6. Int. Workshop on Large Scale Integration of Wind Power and Cables transmission
Networks for Offshore Wind farms October 2006, Delft, pp. 165-169
[3] H. Brakelmann, M. Jensen, “Aufwandsminimierte, magnetfeldarme Hochleistungsübertragung mit bipolar

betriebenen Drehstromkabeln e&i, 2008, Austria, 2008,H. 1/2, pp. 42-47
[4] H. Brakelmann, M. Jensen, Th. Schütte, “Transformatorschaltungen und Redundanzen bipolarer HVAC-
Übertragungssysteme,“ Ew (2007), H. 25-26, pp. 68-73
[5] H. Brakelmann, I. Erlich, R. van de Sandt, “Analysis of bipolar-6-phase-transmission system

with option of 4-phase-operation for large scale wind farms“, Power and Energy Conference, 2008.
PECon 2008. IEEE 2nd International/ , pp.127-132, Dec. 2008
[6] H. Brakelmann, J. Brüggmann, J. Stammen, “Connection of Wind Energy to the Grid by an Optimized HVAC
th
Cable Concept,” 7 Intern. Workshop on Large-Scale Integration of Wind Power, Madrid, 2008, pp. 353-35
[7] H. Brakelmann, J. Brüggmann, A. Jensen, J. Stammen, “Bulk Power Transmission with HVAC Four-Core
Submarine Cables,” DEWEK-Conference, Bremen, 2008, pp. 1-6
[8] H. Brakelmann, J. Brüggmann, J. Stammen, “Bipolar Cable Systems and their Benefits in Bulk Power
Transmission”, PowerGrid Europe 2009, Köln
[9] H. Brakelmann, J. Brüggmann, J. Stammen, “Onshore continuation of Bipolar Cable Systems for Bulk Power
th
Transmission”, 8 Intern. Workshop on Large-Scale Integration of Windpower, Bremen 2009
[10] R. Brown et a., “Six-Phase Successfully Applied to Utility Transmission System” CIGRE-report, JWG
22/33/36-01, 1998 ,Report of the New York State Electric and Gas Corporation and the Power Technologies,
Further Co-Funders are: Empire State Electric Research Corporation, U.S. Statement of Energy, State Energy
Research and Development Authority, Electric Power Research Institute (EPRI)
9.2 Passive loops
9.2.1 PASSIVE LOOPS ON A JOINT BAY
This paragraph reports the magnetic shielding of an existing joint bay with a description of the design, the
installation and final measurement. Thermal impact is an important part of this example.
Scope of work
The owner of a 132 kV underground line commissioned the execution of the shielding of magnetic field emitted
by line, in the city of Bologna (Italy), as a gentleman agreement between with owner of a building in close
proximity to a joint bay, as a consequence of a change in the use of the internal space, becoming regularly lived
by people. The job had to be done without removing the joints or repositioning the power cables previously
installed.
Measure of the magnetic field in the sensitive position determined that the magnetic field had to be reduced by a
factor of 5, at the interior of a building located approximately 6.5 m from the longitudinal axis of the joint bay, at
1 m above ground.
Page 67
For the recovering operation of the magnetic field the owner of the HV circuit agreed in allowing a thermal
margin of 5K at the nominal current of 860 A.
Design of the passive loop structure
The solution proposed was to use the passive loops technique, for the first time on an existing joint bay a with
SF of 5.
The design was made with 2D modelling, according to the previous description of chapter 6.3. In fact it is
possible to adjust the design as a work in progress, it is not a “do it right first time” job, because the current is
disconnected during the working operation but is given again during night time, so that it is possible to measure
the SF during the progress of global work and compare it with intermediate computations
Magnetic field has to be computed at 360 A which is the mean value of the current circulating into the circuit
but derating has to be respected at 860 A: SF is not dependent on the current for this technology.
To obtain the required field reduction it is necessary not only to shield the area of the joint where the cables
have a lager interaxial distance, but also the two parts of the circuit immediately adjacent to the joint bay, for a
length of about ten meters per part.
2
The solution adopted consists of passive loops, 185 mm Al conductor cables, placed around and on top of the
HV cables as described in Fig. 9.2.1.1.
For the two parts of HV cables a similar structure of 18 loops was installed, with two trefoils on each side of
power cables and 3 loops equally spaced, 200 mm above. For both side a length of about 10 meters was
shielded.
Description of the job.
Fig. 9.2.1.1 describes the solution adopted, with passive loops. According to the design, the joints had been
installed with an interaxial distance of 650 mm, at a depth of 500 mm measured at the axis.
The passive loops were installed on the sides and above the joints, with a deep excavation work that
necessitated to operate carefully to avoid joint movement or other mechanical injuring.
650 mm
1900 mm
Page 68
Fig. 9.2.1.1 Section of the passive loops into the joint bay.
First of all the position of the joints had to be recovered with a careful dig out of the sand around the joints care
to avoid any slipping of the walls of the bay according. The work was done without current in the HV circuit,
current that was circulated during the night with people not working
The cables were firstly positioned on the bottom of the trench, jointed to form loops and protected with heat-
shrinkable sheaths.
The trench was partially backfilled and other loops were installed, again grouped in at number of three and
some sand was added. Measure of the partial SF was made during the progression of the global shielding
structure and compared with theoretical computations; some more loops were added to the original design to
increase the foreseen SF.
Finally the structure of upper layer of passive loops on top of the trench was assembled and the magnetic field
measured again.
Measured values of the magnetic field were done at 277 A and then reported linearly to the nominal value of
360 A for the magnetic field: the minimum value of the SF at requested position was computed in 5.1.
The trench was completely backfilled and a document reported the job done and value of the magnetic field
than had been reached.
Fig. 9.2.1.2 shows the upper layer of the passive loops arranged on top of the compacted backfill.
The positioning of the passive cables and the choice of the conductor section and a small increasing allow
limiting the increase of ambient temperature to a 2 K, evaluated at 860 A.
Fig. 9.2.1.2 Part of the passive loops installed on top of the joint bay.
Page 69
9.2.2 PASSIVE LOOPS IN A JOINT CHAMBER
This paragraph reports the magnetic shielding of a joint chamber with a description of the design, the
laboratory tests and final installation. Thermal impact is also discussed in the example.
Scope of work
The scope of work in the installation reported in this example is the reduction of the magnetic field emitted by a
joint chamber at soil level, with a current of 1500 A. The minimum requested SF is 1.5 times. In the remaining
part of the circuit, magnetic field mitigation is achieved with triangular arrangement and burying of the cables in
a deep trench.
The owner of the circuit allowed a thermal increase of the temperature of the joint chamber of maximum 3K,
because that is not the hottest part of the circuit.
Design of the passive loop structure
Tests were done on slightly different configuration for simplicity reason, in order to test the model and verify the
correctness of the hypothesis, with the two loops at the same level of the joints. Once verified the correctness of
the circulating currents and of the measured magnetic field, the loop position was optimized with simulation
software. Best solution was found with the loops on a different level. Passive loops have been installed
according to the schematic section reported in Fig. 9.2.2.1.
When passive cables are installed inside a joint chamber to mitigate the magnetic field, the heat dissipated by
the cables does not flow directly into the soil but is conveyed by air to the walls of the room. Ambient
temperature rise can be easily computed, assuming a 2D simplification, with the Kennelly formula, on the base
of the losses of all the passive loops, as presented in previous paragraph 6.2.
Description of the job.
In the example reported here, to shield the EMF of a joint chamber with a cross section of 2.5 m x 2 m and
buried at a depth of about 1 m measured at the roof, installed in Vienna, two passive loops have been installed
according the scheme of Fig. 9.2.2.1. Assuming a soil resistivity of 1 K*m/W, the external thermal resistance T4t
of the chamber can be estimated in 0.20 ÷ 0.25 K*m/W, with the Kennelly formula. In this case (Fig. 9.2.2.2) the
two loops carry 350 A and 80 A.
Passive loops
50
300 mm2 Cu
750
70
200 300 300 250
Page
Fig. 9.2.2.1 Schematic disposition inside the joint chamber.
The thermal resistance T 4w at the walls of the chamber is about 0.037 K*m/W (see eq. 6.4.6), giving a global
thermal resistance T4 of 0.24 ÷ 0.29 K*m/W.
The corresponding losses are respectively 5 W/m and 0.27 W/m, for a total value of about 10.5 W/m, for the
four cables. The ambient temperature rise is in the order of 2.5 ÷ 3 K, with no effect on the global rating of the
circuit due to fact that the joint are not a hot point of the circuit as they are installed in air.
For more stringent limits on the total thermal rise in the chamber, larger conductor sections can be used, giving
rise to correspondingly less heat dissipated in the passive loops.
It can be concluded that for a joint chamber the total thermal rise is limited or negligible, due to the very low
thermal resistance of the chamber towards air temperature.
Measurement of the magnetic field on site reported the correctness of the installation done and of the Shielding
Factor achieved.
External
loop
Internal
loop
Fig. 9.2.2.2 Passive loops installed in Vienna
9.3 Metallic plates
9.3.1 ALUMINIUM PLATE SHIELDING FOR UNDERGROUND CABLES IN BELGIUM
The shielding design presented in Fig. 9.3.1.1 is an illustration of the technique described in TB 373 and in section
6.3. It has been incorporated in 2003 in the construction of a 6 km section of 150 kV double circuit underground
2
cable link which is 30 km long. Each circuit involved three 2000 mm (Al) XLPE cables laid in a flat formation 20 cm
Page 71
above the bottom of a the 150 cm deep cable trench. The horizontal separation distance between the axes of the
cables is 25 cm. The horizontal separation distance between the centres of each cable circuit is nominally about 2
m, but in can vary between 1.5 and 5 m.
The screen is built with a large number of 200 × 80 cm aluminium (EN-AW 1050A alloy, i.e. 99,5 % Al without
copper) plates of 3 mm thickness. The vertical plates are installed to line up both sides of the trench for the entire
12 km length of the cable section. The distance between the vertical plates on either side of the cable circuit is 100
cm. The horizontal plates are installed at 40 cm above the bottom of the trench.
The vertical plates are installed with a small overlap and welded together by MIG method under argon atmosphere,
starting from the top, over 2/3 of the width of the vertical plates (i.e. the lowest parts of the plates are not welded
together).
At both ends of any shielded section, the two vertical walls, formed from welded aluminium plates, were electrically
connected together by U-shaped aluminium sections of the same thickness and width as all other plates. This
ensures the necessary electrical linkage between the two aluminium walls installed on either side of the cable
circuit (see dotted lines in Fig. 9.3.1.1). For safety reasons, additional bridges are regularly added and also before
and after crossing a road.
Using this technique, it has been possible to achieve a shielding factor of about 10. Concerning the losses, the
magnitude of the current flowing in the vertical plates is about 1/3 of that in the cables; the eddy currents in the
horizontal plates are even higher (but they could not been measured) and are responsible for the main losses ( 20
W/m @ 1 kA, i.e. nearly one half of the losses of the cables). These losses, however, have practically no influence
on the soil temperature and, hence, on the cable rating.
100
150
welding
bridge
80
20
25 20
80
Fig. 9.3.1.1 : Practical layout for flat formation (H shape shielding).
Fig. 9.3.1.1 illustrates the welding process and shows how easy it is, using aluminium plates, to follow the
unevenness of the terrain (H shape layout).
Page 72
a) b)
Fig. 9.3.1.2 a) Sight of the vertical and horizontal plates before refilling the trench
b) Welding of a bridge at the extremities of a shielded area
Calculation of shielding factor and derating
It is noted that the analysed cross-section is not the thermal bottleneck of the cable circuit – other parts of the total
route are limiting the circuit’s rating. Coupled electro-thermal analysis has been done using both two- and three-
dimensional commercial FEA software, and with the two yielding more or less identical results for temperature and
magnetic field. Selected results are shown in Fig. 9.3.1.3 and Fig. 9.3.1.4. The obtained SF some meters away
from the cable trench is approximately 10, while higher values apply directly above.
Fig. 9.3.1.3 Flux lines and BR above coil surface (left) and SF (right) for H shape shielded flat formation
Page 73
BELGIUM_SHLDD_T
Color Shade Results

Quantity : Temperature degrees C.
Time (s.) : 1E9

Scale / Color
40 / 43,11149
43,11149 / 46,22297
46,22297 / 49,33446
49,33446 / 52,44595
52,44595 / 55,55743
55,55743 / 58,66892
58,66892 / 61,78041
61,78041 / 64,8919
64,8919 / 68,00339
68,00339 / 71,11488
71,11488 / 74,22636
74,22636 / 77,33784
77,33784 / 80,44933
80,44933 / 83,56082
83,56082 / 86,67231
86,67231 / 89,7838
Fig. 9.3.1.4 Steady state temperature (left) and temperature rise with time (right)
From right-hand part of Fig. 9.1.3.4 it is seen that it takes approximately 100 days (at constant, rated current) to
reach thermal steady state and 90 °C on the hottest (centre) conductor.
When the H shape shielding is added to a reference (unshielded) flat formation, current rating drops from 1500 A to
1450 A. This is due to additional heating from power loss in the shielding plates as shown in Fig. 9.3.1.5. MF
mitigation by means of H shaped plates causes a (local) derating of 3.3 %. In this case, the derated capacity
remains above that of the cable circuit.
Page 74
BELGIUM_SHLDD_M
Color Shade Results
Quantity : Power density W/(cubic m)
Isovalues Results
Time (s.) : 1E9 Quantity
Phase: Equi (Deg): 0
flux Weber
Scale / Color Time (s.) : 1E9 Phase (Deg): 0

Line / Value
6,36945 / 262,28992
5 / -496,70099E-6
6 / -372,52574E-6
262,28992 / 518,21033
7 / -248,35049E-6
518,21033 / 774,13074
8 / -124,17525E-6
9 / 0
774,13074 / 1,03005E3
10 / 124,17525E-6
11 / 248,35049E-6
1,03005E3 / 1,28597E3
12 / 372,52574E-6
1,28597E3 / 1,54189E3
13 / 496,70099E-6
14 / 620,87621E-6
1,54189E3 / 1,79781E3
15 / 745,05148E-6
16 / 869,22676E-6
1,79781E3 / 2,05373E3
17 / 993,40198E-6
2,05373E3 / 2,30965E3
2,30965E3 / 2,56557E3
2,56557E3 / 2,82149E3
2,82149E3 / 3,07741E3
3,07741E3 / 3,33333E3
3,33333E3 / 3,58926E3
3,58926E3 / 3,84518E3
3,84518E3 / 4,1011E3
Fig. 9.3.1.5 Flux lines (not related to colour legend) and power loss intensity in vertical plates
Maintenance and repair
Six years after installation, control measurements have shown that some shielded sections were inefficient. This
was due to the break of some welds between vertical plates. The analysis of the failures showed that the breaks
were due to bad welds and not to some aging process. Fortunately, when a break occurs, the induced voltage
concentrates on the edges of the broken plates and it becomes possible to locate the failure by measuring the soil
surface potential in the vicinity (a few tenths of a volt). The increase of the induced voltage has also been
responsible for important leakage currents leading to local AC corrosion (cf. TB 373, chapter 2.3.4).
References
[ 1] A. Gille, V. Beghin, G. Geerts, J. Hoeffelman, D. Liémans, K. Van Gucht - Double 150 kV link, 32 km long, in
Belgium: Design and construction - CIGRE report B1-305 – Paris, 2004 session.
[2] A method for applying a magnetic shielding along an AC power line - European - patent application EP 1598911
A1.
Page 75
9.4 Ferromagnetic raceways
9.4.1 INSTALLATION IN PEDESTRIAN TUNNEL
As an example of application of rating computation with application of the ferromagnetic raceway, it is reported
here the description of a particular shielding case installed into a pedestrian tunnel (Fig. 9.4.1.1a). The
installation is requested to carry a current of 1025 A at HV.
Due to the particular installation, the three single-core cables are to be laid flat in vertical configuration fixed to the
wall of the tunnel, to minimize space occupancy. The tunnel is 27 m long, plus two portions at the extremities of the
tunnel for a total installation of 67 m. The cables have an outside diameter of about 100 mm and are inserted into
PEHD pipes with external diameter of 160 mm, placed with an interaxial distance of nearly 190 mm. A Shielding
Factor of 13 is required to reduce the magnetic field below the requested value, at the external surface of the brick
wall protecting the cables.
Specific FEM computation, due to this unusual thin geometry has been used to determine the EMF and the total
losses generated into the raceway: the computed value of the extra losses is 61 W/m, at full load. Also a copper
plate placed in front of the pipes has been studied that would have given lower induced losses but insufficient
magnetic shielding.
A gap of some centimetres has been left all around the raceway, to guaranty that natural air convection between
the raceway and the protecting wall efficiently extracts most of the heat generated inside, so that full rating of
the circuit can be maintained (Fig. 9.4.1.1b). Pedestrian walking was allowed during installation.
After laying measurements of the magnetic field confirmed the effectiveness of the solution designed achieves the
required SF.
Fig. 9.4.1.1a Raceway during installation. Fig. 9.4.1.1b Natural convection cooling of the raceway.
Thermal solution: air convection cooling.
When the cables pass from buried conditions to installation in open air, inside a trough, the thermal conditions are
better and there are no derating problems. In this case the heat induced by the raceway can be evacuated
exploiting natural air convection circulating in the gap between raceway cover and brick wall.
Page 76
To be on the safe side, the natural cooling has been dimensioned to evacuate all the 61 W/m at full load of 1025 A.
The maximum difference between inlet and outlet air temperature has been limited to 10K. Air inlet windows have
been designed on the bottom part of the front panel and air outlet windows on top of the same panel, to exploit
2
maximum height difference. A couple of holes, each with a section of more than 200 cm have been designed
every 1 meter of circuit. A simple calculation gives and air speed at inlet of 0.57 m/sec, computed as (2 x 9.8 x 1 m
1/2 3
x 5°C/ 300 °C) . Assuming air specific heat of 1000 J/K/m , the total heat flux evacuated by air in the above
conditions is in excess of 100 W per each meter of raceway.
This installation demonstrates an example of magnetic shielding in which the losses induces in the mitigation
device do not reduce the current carrying capacity of the circuit, through a careful thermal design.
9.4.2 UNDERGROUND INSTALLATION
The following example describes the shielding of underground cables using the ferromagnetic raceway in the city
of Naples (Italy). It is a widely experimented mitigation technique used when the most stringent SF.
Circuit description
The 3-phase cable link to be shielded is 5 km long and follows a tortuous path numerous underground crossing
services between the towns of Lettere and Castellammare, near Naples. The circuit is operated at 132 kV at a
2
nominal rating of 860 A. with XLPE and. The cable has an aluminium conductor of 1000 mm , XLPE insulation and
an external diameter of 103 mm. Laying depth is 110 mm.
Shielding Factor and design
In view of the special characteristics of the path, the shield design posed several challenges and the ferromagnetic
raceway revealed to be the best solution. The raceways are light, easy to install and are designed in such a way
that the cables automatically assume the trefoil configuration. The particular shape allows the raceways to follow
the small curves of the trench and lateral and vertical variations of direction; for large curves, special shaped
elements are used.
The ferromagnetic raceway comprises different elements: the base, the cover and the closing clips, curved
elements. It is constructed with high permeability ferromagnetic steel, galvanised to avoid corrosion problems.
The raceway has a base width of 220 mm, a height of 210 mm: the cover is 350 mm wide (Fig. 9.4.2.1).
Page 77
Fig. 9.4.2.1 Picture of the shielding raceway containing three cables.

The dimensions of the raceway are adapted to the diameter of the cables, so that any high voltage connection can
be shielded. Each section of the raceway has a conical shape so that it can be partially overlapped to the next one,
to form a continuous shield, in such a way that any welding is avoided. For a better mechanical and electrical
contact galvanized steel closing clips are applied between the base and the cover, with regular intervals.
Installation
The bottom of the trench is prepared with a bed of compacted sand and the bases of the raceway are installed,
with an overlapping of about 200 mm. Then the rollers are placed on the top of the bases and the cable is pulled
with a simple operation and deposited into the raceway simply by lifting the roller. The same operation is repeated
for the other two cables; the cables automatically assume the trefoil configuration. Subsequently, the raceway is
filled with a weak mix of cement and closed with the cover, which is finally fixed by the clips.
Page 78
Fig. 9.4.2.2 Installation of two parallel raceways in Bologna (Italy).
Results and conclusions
It was possible to reduce the peak value of the external magnetic field at 1 meter above ground below the
requested value, at the nominal rating current of 860 A, achieving an effective Shielding Factor greater than 18.
The ambient thermal rise due to the losses into the raceway was of 5 K, corresponding to a reduction in the current
rating of less than 4%.
Clear advantages of the raceway shielding technique are the simple design without welding and the ease of
installation with cables that automatically assume the trefoil configuration. The use of standard materials makes
this method cost-effective when large SF are required.
References
[1] P. Argaut, J. Y. Daurelle, F. Protal, K. Savina and C.A. Wallaert, Shielding technique to reduce magnetic fields
from buried cables, JICABLE, A10.5, pp. 331 Fig. 9.1.4.-338, 1999.
Page 79
[2] P. Maioli: "Environmental constraints: electromagnetic shielding with ferromagnetic raceway”, communication
presented at 2004 CIGRE’ Conference.
9.5 Steel pipes
9.5.1 INFLUENCE OF GEOMETRY ON LOSSES AND SHIELDING FACTOR
For consideration of losses and shielding factors of a steel pipe the used arrangement of cables is shown in Fig.
9.5.1.1. In order to maintain a larger mutual distance between the cables plastic pipes are used inside the steel
pipe. Subsequently the free space around the plastic pipes is filled with concrete. The mitigation will be in the
middle of the steel pipe and therefore the end effects has no influence on the shielded part.
The use of plastic pipes within the steel pipe makes it possible, in case of cable failure, to pull out the cable from a
joint bay without any additional trench digging.
Steel pipe
Fig. 9.5.1.1 Single-core cable

system in a steel pipe
PE duct Cable
Concret
e
2
To consider the nowadays technical limits a 380 kV XLPE cable with a cross-section of 3200 mm copper is used
for this study. To minimize the induced metal sheath currents the metal sheaths are cross-bonded.
The steel pipe has an outer diameter of 1000 mm with a thickness of 10 mm. The steel pipe itself is covered with
an outer plastic serving of 10 mm which gives an overall diameter of 1200 mm. Concerning the thermal aspects of
the inside filling of the pipe is conducted with high thermal conductive specialized concrete.
Specialized concrete has a 2-4 times higher conductivity than ordinary concrete. In case of a horizontal drilling the
steel pipes is surrounded by bentonite which is used during drilling for securing an open drill hole.
When installed directly buried the surrounding of the steel pipe can be of a soil type with good thermal properties.
For the performed calculations the standardized properties of soil are used in accordance with the IE/VDE; i.e.:
 Thermal conductivity wet soil λw = 1,0 W/Km
 Thermal conductivity dry soil λw = 0,4 W/Km
 Dry out temperature Δθd = 15 K
 Undisturbed soil temperature θa = 15 °C
The permeability of ferromagnetic materials is to a large extend depending on the effective magnetic field. The, by
the cable configuration emitted, magnetic field changes the permeability of the screening material and as a
consequence influences the skin depth of the magnetic field into the material. The effect influences directly the
Page 80
metal cable screen and conductor losses of the cables. Also additional losses occur within the magnetic screening
material due to the hysteretic losses by the hysteretic of the magnetic field properties of the screening material. In
order to analyse the non-linear ferromagnetic behaviour of the material a finite element method is used.
In Fig. 9.5.1.2, the current rating as a function of installation depth of the cables in triangular formation are shown.
The trefoil arrangement could, for example, be chosen for impedance regulation or reduce the magnetic emitted
field from the cables. For line 2 whereas the inter axial cable distance 300 mm is the current rating varies from
2100 A to 1650 A for the accompanying laying depth; i.e. 1400 MVA and 1100 MVA at 380 kV.
2200
A 2
2000
I
1800
1
1650 A
1600
1400
0 1 2 3 m 4
h
Fig. 9.5.1.2 Continuous current rating I of 380-kV-XLPE-cables as a function of the laying depth h at the
2
axis of the upper cable; copper conductor 3200 mm ; thermal stabilized region with St = 4,0 W/(K m); line
1: trefoil formation (s = 0 m); line 2: triangle with distance of axes of s = 0,3 m.
The aim is to install the cable system within the steel pipe such that the magnetic shielding is optimal. The optimal
spatial distance between the cables in reference to the additional losses is part of the study, also in respect to
current rating and metal sheath losses. Due to the installed plastic pipes within the steel pipe the minimum spatial
distance is 200 mm. In Fig. 9.5.1.3 are the correlations given between above mentioned aspects.
In Fig. 9.5.1.2 can be seen that losses of cables without steel pipe varies between 66 W/m and 85 W/m for a
distance of 200mm to 600mm between the plastic pipes. When installed in the steel pipe the losses already start to
increase significantly from 200 to 300mm; i.e. ~75% of the overall cable losses. For distances bigger than 300mm
the steel pipe losses are growing towards 550W/m at a spatial distance of 600mm; i.e. a factor 8 higher than the
overall cable losses.
The shielding factor SF, in Fig. 9.5.1.3, is the relationship of the magnetic shielding of the unshielded cables
divided by the shielded cables. The given SF is on earth surface. In the Fig. 9.5.1.3 can be seen that for a distance
a = 300 mm a maximum SF factor of 131 is reached. For the biggest spatial distance of the cables the SF is 74; i.e.
the plastic pipes are against the inner wall of the steel pipe.
Page 81
60 15
0
W/m 0
50 12
0 SF 5
40 10
PR 0 0 SF
´ 30
0
7
5
20 5
0 PR Without steel
0
10 pipe
2
´
0
0
Pcable 5
0
0 10
0
20
0
30
0
40
0
´ 50 m 60
0 m 0
a
Fig. 9.5.1.3 Pipe losses p.u.l. PR´ and shielding factor SF as a function of the interaxial distance a of the
cable cores in a steel pipe d; P´cable are the cable losses p.u.l. without steel pipe 380-kV-XLPE-single-core
2
cable; copper conductors 3200 mm ; conductor current 1650 A pipe diameter 1000 mm; wall
thickness 10 mm
In Fig. 9.5.1.4 and Fig. 9.5.1.5 the permeability spread is given for two spatial distances of the cables. Herewith the
effect of the steel pipe on the magnetic field reduction can be analysed. From Fig. 9.5.1.4 can be seen that for a
distance a=300mm the relative permeability can grow up to 600 at some points of the steel pipe. For a distance
a=589mm, the cables are only 50mm from the inner wall of the steel pipe, the relative permeability can be 1200. As
a consequence the circulating currents and the hysteretic losses will be significantly higher in the steel pipe.
Furthermore shows Fig. 9.5.1.6 the relationship between the conductor temperature at a specific current of 1600A
as a function of the distance “a”. For installation of cable without steel pipe, the conductor temperature decrease
with increasing distance a. If a steel pipe is present the changing distance a gives a temperature minimum and at
this distance the biggest conductor current.
μr
600
0.4 550
500
450
400
350
0.2 300
250
200
150
0
-0.4 0.4
-0.2
-0.2 0.2
0 0
-0.4 0.2 -0.2
0.4 -0.4
-0.4 -0.2 0 0.2 0.4
Page 82
Fig. 9.5.1.4 Distribution of the relative permeability in the steel pipe (right hand); distance of axes
a = 300 mm; conductor current 1650 A
0.4
μr
1200
0.2 1100
1000
900
800
700
600
500
0 400
300
200
100
-0.2
-0.4 0.4
-0.2 0.2
-0.4 0 0
0.2 -0.2
0.4 -0.4
-0.4 -0.2 0 0.2 0.4
Fig. 9.5.1.5 similar to Fig. 9.5.1.4, only distance of axes a = 589 mm
25
°C
0
20
0

c 15
0
10

0 c
5
Without steel
0 pipe
0
0 10 20 30 40 50 m 60
m
0 0 0 a0 0 0
Fig. 9.5.1.6 Maximum conductor temperature c as a function of the distance of axes a,
with and without steel pipe
Based on above it is of great importance to determine in which way the losses in the steel pipe and the SF are
dominated by the thickness of the wall of the steel pipe. In Fig. 9.5.1.7 is the optimum given at a distance of
a=300mm. Variation of the wall thickness between 10 en 5mm shows that the losses are almost constant at 53
W/m. Lowering the wall thickness from 5 downwards gives a maximum loss at a thickness of 3mm of 58 W/m and a
lowest loss at a thickness of 1mm of 44 W/m.
In respect to other shielding solutions, such as compensation cables or aluminium plates with SF of 10, the wall
thickness must be thicker than 5mm. SF of 50 can be reached with a thickness in the range of 9mm; i.e. 131 as a
SF for a thickness of 10mm.
Page 83
60 150
W/m P´
R
50 125
40 100
P´
R SF
30 75
20 50
SF
10 25
0 0
0 2 4 6 8 mm 10
s
Fig. 9.5.1.7 As Fig. 9.5.1.3, but: Pipe losses p.u.l. PR´ and shielding factor SF as a function of the wall
thickness s of the steel pipe; distance of axes a = 300 mm
9.5.1.1 Current derating
The losses in the steel pipe causes additional warming up of the steel pipe and as a consequence the cable
surroundings. Therefore the current rating of cables without steel pipe (I without) reduces when installed in a steel pipe
(Iwith). Taking into account the given equations of the current in [6] the current derating factor νi results to:
I  max   d   R   x
vI   (9.5.1.1.1)
I oR  max   d   x
with d conductor heating due to dielectric losses (°C),

R conductor heating due to steel pipe losses (°C),
max = max - a maximum conductor heating (°C),
 x   g  (F / T  1) correction factor for partial drying out of soil.
For the values given in Fig. 9.5.1.3 and Fig. 9.5.1.7, i.e. distance a = 300, wall thickness s = 10 mm, are in Fig.
9.5.1.8 the relative heating of the pipe ΔθR and the derating factor νi as a function of laying depth h given. As can
be seen the growth of the relative heating of the pipe enlarges with bigger laying depths.
Page 84
100
%
°C
80
vI
vI 60
R R
40
20
0
0 1 2 3 4 m 5
h
Fig. 9.5.1.8 Temperature rise R because of the pipe losses as well as derating factor vI as functions of
the laying depth h; interaxial distance a = 300 mm; wall thickness s = 10 mm
BIBLIOGRAPHY
[1] H. Brakelmann, J. Dietrich, A. Märten, H.P. May: „Hochwärmeleitfähiger Spezialbeton für den
Kabelgraben“; El.wirtschaft ew, H. 14/15, Juli 2009, pp. 66-72.
[2] P. Maioli, E. Zaccone: “Passive Loops Technique for Electromagnetic Field Mitigation: Applications
and Theoretical Considerations”; Jicable-Conf. 2007, Versailles, pp. 231-236.
[3] H. Brakelmann: “ Optimisation of compensation conductor systems for magnetic field mitigation near to high-
voltage single-core cables”; 7th International Workshop on LSI Windpower, Madrid (2008), pp. 367 -374.
[4] IEC 60287: “Current ratings of cables for continuous load (100 % load factor)“; IEC-Publikation, 1995.
[5] H. Brakelmann:; „Belastbarkeiten der Energiekabel – Berechnungsmethoden und Parameteranalysen“; VDE-

Verlag (1985), pp. 1-400.
[6] G.J. Anders: “Rating of electric power cables”; IEEE press, New York, 1997.
9.5.2 STEEL PIPE IN GENOA AT 132 KV
This example reports the implementation of a mitigation technique for a three-phase system of underground
cables. The technique is based on the use of ferromagnetic shields with tubular geometry. As a result very effective
shielding factors of up to two orders of magnitude were obtained [1], [2].
Shielding of underground cables using ferromagnetic pipes
Page 85
The methodology was applied to HV underground cables in Italy. Some of the circuits of these systems cross
central and populated areas of cities such as Genoa and Trieste. The present report is the work made in the city of
2
Genoa where about 2 km of circuit of 150 kV 1x1000 mm XLPE cable were shielded with this technology.
Fig. 9.5.2.1 Disposition of cables within the steel shielding pipe.
Technique and materials
The applied technique consists of enclosing the three cables inside a tubular “closed section” shield of a relatively
good ferromagnetic material Fig. 9.5.2.1. The effect of this operation is the trapping of the magnetic flux lines inside
the magnetic circuit, preventing propagation to the outside region. The ferromagnetic pipes were realized with low
carbon steel, with an external diameter of 508 mm and a thickness of 9.5 mm: the laying depth was between 1.6
and 1.0 meters, according to interference with existing services.
The pipes are protected with a 5 mm thick polyethylene sheath, to prevent corrosion. The extremities of each tube,
8 meters long, were spherically bickered to follow small changes in the path direction and for a perfect welding (Fig.
9.5.2.2). The overlapping zone was protected with a special primer and a polyethylene sheath.
Three plastic pipes, external diameter 160 mm, were pulled inside the shielding tube and blocked with cement
injected from an extremity. The trench was backfilled and each cable was pulled in when a section of the entire link
was prepared.
Results and observations
The measured magnetic field at 1 meter above ground was found to be always below 0.2 microTesla on the axis of
the link, with a current of 800 A. This corresponds to shielding factors larger than 50. It was not possible to use the
ferromagnetic pipes in the main curves, due to mechanical pulling problems. In these cases, copper plates, 3 mm
Page 86
thick and 1 m wide, were installed, below and above the cables: a higher field value was accepted for these
sections.
Fig. 9.5.2.2 Head of the installation of a steel pipe inside the trench.
In spite of its effectiveness, the pipe system has demonstrated some difficulties in the laying operations and in the
successive pull-in of the cables, mainly for very tortuous paths. These two are the main reasons for the subsequent
introduction of the raceway system, combining the closed perimeter shielding efficiency with an open shape and
the absence of welding.
References
[1] P. Argaut, J. Y. Daurelle, F. Protal, K. Savina and C.A. Wallaert. Shielding technique to reduce magnetic fields
from buried cables, JICABLE, A10.5, pp. 331-338, 1999.
[2] R. Conti, F. Donazzi, P. Maioli, R. Rendina, E.A. Sena. Some Italian experiences in the utilization oh HV
underground cable systems to solve local problems due to magnetic field and other environmental issues,
Cigré 2006 session, paper C4 -303.
9.6 Special cable design
Page 87
9.6.1 THREE-CORE CABLE
Reference for case studies concerning three core cables without magnetic tapes installed in non-metallic ducts or
plastic pipes, see the described methods in paragraph 9.1.1 and 9.1.2. For reference of case studies with three
core cables without magnetic shielding installed in steel pipe, see the described method in paragraph 9.5.
First experience with integrated electromagnetic shielding was gained with a three core XLPE cable, type
A2X(FL)2Y 3x1x240RM 12/20kV. Special magnetically soft shielding tapes have been wrapped around the cable in
two layers. These shielding tapes are characterized by an extremely high permeability and very low hysteretic
losses. For protection reasons a plastic foil has been wrapped around the ferromagnetic tapes.
For the test approximately 100m of the three core cable with the ferromagnetic tapes has been laid straight and
free from other magnetic disturbing effects. Subsequently currents were induced into the cable with a maximum of
350A and simultaneously magnetic field measurements were performed.
From the magnetic field measurements, depending on the distance towards the cable, a shielding factor could be
obtained for more over 80 compared to measured magnetic field of an unshielded cable.
From this test is can be concluded that the shielding factor of two overlapping ferromagnetic shielding tapes, with a
thickness of 0.3mm, is more than five times higher than the shielding factor of a standard steel pipe.
Based on this the next step was to verify these results for a cable with an Integrated Electromagnetic Shield (IES)
and higher current capacity. This resulted in the first development and production of a so called IES cable.
The first produced IES Cable is shown in Fig. 9.6.1.1 is identified as 2X(FL)2YVMC2YF 3x500RM (3x70)(FO)
64/110kV. The cable consists of three single cores with a conductor cross section of 500 mm². The insulation
thickness is of 9.5mm XLPE. The three single cores have each an aluminium-plastic-laminated sheath for water
protection and for carrying the charging current. Three tinned copper conductors are laid into the interstices of the
single cores as earthing conductors.
For the magnetic shielding, two special shielding tapes with a thickness of 0.3mm each have been wrapped around
the cable with an overlap. As a corrosion protection a HDPE-sheath has been extruded around the magnetic tapes.
Finally a flat steel wire armour has been applied to enable the cable pulling.
Fig. 9.6.1.1 110 kV power cable with integrated electromagnetic shielding

²
1 Copper conductor 500 mm stranded 2 XLPE-Insulation
Page 88
3 Earthing conductor 4 Electromagnetic shielding

5 Steel armour 6 Plastic pipe
The above described three core cable has been subjected to a field test in order to demonstrate the magnetic
shielding characteristics of the IES cable. For this field test 110 m of the IES cable has been laid straight in a
magnetic field free desolated area (see Fig. 9.6.1.3 and Fig. 9.6.1.4). The three cables phases were short circuited
at both ends. An emergency generator has been used to inject currents up to 1000 A. It has been ascertained that
the induced currents are sinusoidal without any harmonics of different frequencies.
During the measurements of the magnetic induction, the earthing conductors were insulated at one end of the
cable to measure the shielding effect of the magnetic tapes alone. The solution to minimise the induced currents in
operation is a simple cyclically permutation of the position of the earthing conductors in the joints.
The tests have been performed with a measuring sensor type W&G, EFA 2. This indicator measures the magnetic
induction in three spatial coordinates. Fig. 9.6.1.3 and Fig. 9.6.1.4 show photos of the measurements.
Fig. 9.6.1.2 IES cable test arrangement Fig. 9.6.1.3 IES cable and magnetic field measurement
The measurements were taken at two different positions (MP1 and MP2) in longitudinal direction of the cable. MP1
was 36.5m and MP2 was 56m away from the emergency generator. At both positions the magnetic induction was
measured left from (Pos. 1), right from (Pos. 2) and above the cable (Pos. 3).
It is shown in Fig. 9.6.1.4 that for great distances, the measured field strengths in all directions are the same. For
short distances “s” in the direct vicinity of the cable Fig. 9.6.1.4 shows varying field strengths depending on the
direction of the measurement.
There are two reasons for these deviations:
a) Close to the cable a magnetic induction in longitudinal direction of the cable can be measured. This is
caused by the stranding of the cores: the conductors of the three cores do not lead the current exactly in
longitudinal direction of the cable. In addition the shielding tapes and flat steel wires partly lead the
magnetic flux in longitudinal direction.
b) Due to the stranding the surface of the IES Cable is not completely circular. The distance from the centre of
the cable is not exactly the same for the measurements in different directions.
Page 89
Fig. 9.6.1.4:
Magnetic induction B
as function of distance
s positions MP1 (black
curve) and MP2 (red
curve).
For both positions MP1

and MP2 the values of
the magnetic induction
B in all three directions
(Pos.1, Pos.2 and
Pos.3) are depicted.
Complementary to the measurements, the given IES Cable has been analysed by finite elements software. A
special algorithm for the calculations has been applied, taking the non-linear characteristics of the shielding tapes
into account. Fig. 9.6.1.5 shows the chosen discretisation of the arrangement.
Fig. 9.6.1.6 demonstrates the spatial distribution of the relative permeabilityr of the shielding tapes as well as of
the steel armour.
Fig. 9.6.1.5: FEM-discretisation Fig. 9.6.1.6: Relative permittivity r (I = 800 A)
For even higher requirements or for currents larger than 1000 A, the effective thickness of the shielding can
approximately be doubled by adding a third layer. The resulting shielding factors then increase by a factor of 3
compared to the given IES Cable (Fig. 9.6.1.7). Due to the thickness of 0.3 mm per layer shielding tape, the
increase of the outer diameter of the cable can be neglected. The additional weight per layer shielding tape is
approximately 1 kg/m.
In case two shielding tapes are applied, the thermal losses generated in the electromagnetic shielding are maximal
0.7 W/m, and for three shielding tapes 1.2 W/m, thus reaching approximately 1 % of the total losses of the cable.
On the other hand the magnetic losses in the armoring or if present in the steel pipe are reduced almost to zero, as
the magnetic induction outside the shielding is extremely reduced.
Page 90
Altogether the total losses of the given IES Cable in a plastic pipe are approximately 7 % lower than the losses in
an unshielded cable. Therefore the application of IES Cables in pipes should be considered from an economical
perspective, too, irrespective of any given limit values for the magnetic induction.
Fig. 9.6.1.7:
Shielding factors SF of
steel pipe cable (1) wall-thickness:

4.5 mm
IES Cables with two (2)
or three (3) shielding tapes

thickness: 0.3 mm
Using IES Cables, a cable system design with installation in a plastic pipe can be realized. This design offers
advantages in cable laying, fitting, corrosion protection, current carrying capacity, losses and costs respectively.
Simultaneously the IES Cable in a plastic pipe is characterized by extremely low magnetic inductions and lower
losses compared to a steel pipe cable.
From losses point of view the IES design is also advantageous when the cable is laid in a steel pipe. The losses in
the magnetic pipe and armor are almost eliminated while the losses in the shielding tapes are negligibly small. As a
result the total cable losses are lower and consequently the current carrying rating is higher.
BIBLIOGRAPHY
[1] H. Brakelmann, M. Kirchner, W. Rasquin, V. Waschk: Retrofitting von 110-kV-Druckkabelanlagen mit 110-kV-
VPE-Kabeln. Elektrizitätswirtschaft 96 (1997), pp. 116…119
[2] H. Brakelmann: Measurements of magnetic induction on a shielded cable. Internal report University of Duisburg-
Essen 26.02.2006
[3] J. Brüggmann: Numerische Untersuchungen zum elektromagnetischen und thermischen Verhalten von 110-kV-
VPE-Stadtkabeln. Diploma thesis at University of Duisburg-Essen, July 2007
[4] J. Stammen: Numerische Berechnung elektromagnetischer und thermischer Felder in

Hochspannungskabelanlagen. Dissertation at University of Duisburg-Essen, Shaker Verlag, Aachen, 2001
[5] H. Brakelmann, J. Stammen: Simulation magnetischer Schirmungen im Niederfrequenzbereich unter

Berücksichtigung feldabhängiger Parameter. EMV 2006, Düsseldorf, pp. 647...654
9.6.2 SINGLE-CORE CABLES WITH LARGE SOLID-BONDED SCREENS
Page 91
In CIGRÉ-paper B1-305, 2008 [1], a new three-core cable with an integrated electromagnetic shielding, the so
called IES-cable was presented, which enables highest shielding factors without negative influence on the current
capacity. In continuation of this work, this special shielding method has been applied to single-core cables. In the
following, the design and properties of such a SC-IES cable are described here and compared with single-core
cables with large solid-bonded screens, but without the additional magnetic shielding tapes.
For this, a XLPE-insulated cable is considered. Different to conventional cables, which might have copper-screens
of 50 mm² to 250 mm², the cable has a reinforced copper-screen of up to e.g. 2500 mm². Above this massive
screen, there may be a layer of high permeable tapes, as already used in the three-core IES-cable. These tapes
have a relative permeability of some ten thousands at negligible eddy-current- and hysteretic-losses. The outer
layers of the cable are an APL-sheath as well as PE-corrosion-protection.
To show up the effect of the tapes, copper armoured cables are compared with armoured and additionally shielded
SC-IES-cables. For this, a three-phase system of these cables, equipped with 2500 mm² aluminium conductors in
flat formation, buried in a depth of 1.5 m with a core-spacing of 0.5 m is analysed. The symmetrical three-phase
conductor current is 1100 A. The screens are solidly bonded at both ends, so that screen currents are induced.
Fig. 9.6.2.1 shows the horizontal distribution of the magnetic induction B at the earth’s surface for single-core
cables with different copper-screen cross-sections AS.
For reaching screen currents in magnitude of the conductor currents, copper-screens of more than 700 mm² are
needed, but nevertheless, noticeable angle deviation from the ideal 180° phase-shift are left, even for extreme
screen cross-sections.
2
For a cross section of the copper screens of 2500 mm a shielding factor of approximately 20 is reached,
accompanied by a loss factor of the screens of 45 %.
a) b)
80 4
T T
unscreened
60 3
2 2500 mm2 without shielding
A s = 100 mm
40 2
B 2 B
As = 500 mm
20 1
0.2 µT
with shielding
0 0
0 1 2 3 m 4 0 1 2 3 m 4
x x
Fig. 9.6.2.1 Horizontal distribution of the magnetic induction B at the earth’s surface
a) single-core cables with different copper-screen cross-sections AS
b) AS=2500 mm², optionally equipped with high-permeable tapes
Page 92
In the following, the same single-core cables with large solid-bonded screens are considered, but now as SC-IES
cables with integrated shielding tapes (Fig. 9.6.2.2).
PE-Corrosion-Protection
APL-Sheath
Shielding Tapes
Copper-Armouring
XLPE-Insulation & Semi-Conducting Layers

Conductor
Fig. 9.6.2.2 SC-IES cable: Single-Core Cable with Integrated Electromagnetic Shielding
For such armoured and shielded cables, the magnitude of the screen currents reaches the conductor current even
for rather small screen cross-sections, accompanied with negligible angle deviations. By this, the return currents in
the sheaths extremely decrease the outer magnetic field of the cable system even for enlarged laying distances.
This is illustrated in Fig. 9.6.2.1b, where the magnetic induction B is significantly decreased (compared with Fig.
9.6.2.1a, the scale is changed by a factor of 1/20). In spite of the huge laying distance, even challenging limits, e.g.
0.2 µT, are not exceeded directly above the cable system.
Page 93
1200
1000
800
shielded
600
SF 400
200
unshielded
0 2
0 500 1000 1500 2000 mm 2500
Fig. 9.6.2.3 Shielding factor SF of shielded (with tapes) and unshielded cables on the basis of the
screen cross-section
The resulting shielding factors SF are shown in Fig. 9.6.2.3 as functions of the screen cross section. This figure
underlines the high efficiency of the shielding system and its superiority to cables without the shielding tapes. The
system already works for small screens, so that the dimensioning of the cable can be done with respect to the
aspired losses or weight.
Both cable types have similar losses. In contrast to ferromagnetic shielding measures like steel-pipes, which in any
case must enclose the complete three-phase system, there is no so-called “in pipe-factor”, which describes the
increased eddy-current losses in the conductors and metallic sheaths/screens, resulting from a concentration of the
magnetic field inside the ferromagnetic structure.
The SC-IES cable is a solution for highest shielding demands. With an effective design, conductor – and screen
currents have the same magnitude and nearly opposing phases. This leads to the behaviour of a coaxial
transmission line with lowest magnetic fields outside the cable trench and a complete decoupling of the phases.
This decoupling enables large core distances for a better heat dissipation.
Obviously, the material usage is increased, But other shielding measures, e.g. steel enclosures, which enforce
small laying distances, require high additional costs, too. They will cause a stronger derating because of the
restricted laying arrangement, accompanied by additional internal losses (“in pipe effect” in the cables; eddy-
current, circulating-current and hysteretic losses in the steel enclosure).
[1] B1-305 “IES-cables – Power cables with Integrated Electromagnetic Shielding”; D. Bielefeld, H. Brakelmann, J.
Brüggmann, V. Waschk; Cigré-Session 2008 Paris
[2] IEC-Publication 60287 “Electric cables - Calculation of the current rating (100% Load factor)”
[3] “SC-IES-Cables: Single-Core Cables with Integrated Electromagnetic Shielding”, H. Brakelmann, J. Brüggmann,
V. Waschk; Contribution to B1-PS2-Q1, Cigré-Session 2010, Paris
9.7 Magnetic field mitigation techniques for a 330kV cable
Circuit Description
A 28 km long single-circuit cable was installed in 2002 from the outer suburbs to Sydney. The circuit consisted of
three single-phase cables operating at 330 kV. The cable was designed to be operated at 330 kV with a continuous
cyclic rating of 750 MVA (1312 A), a 3 day cyclic overload rating of 900 MVA (1574 A) and a 15 minute emergency
Page 94
rating of 1100 MVA (1924 A). The cable was a Poly-Propylene Laminated Paper, Self-Contained Fluid Filled cable,
2
with a 1600 mm copper conductor, a diameter of 125 mm and was directly buried, generally in residential streets
at a nominal burial depth of 900 mm.
Whilst seeking approval for the project, the minimisation of Electromagnetic fields was one of many conditions to
be met. In order to achieve the targets, set during the approval process, there was a need to identify and minimise
the Magnetic Fields surrounding the cable.
Assessment of Mitigation Techniques
Several mitigation techniques were identified as being theoretically practical. In order to verify the practicality of
some of the techniques a field trial of the various configurations was undertaken
The field trial consisted of a length of cable installed in various cable management configurations and mitigation
devices. A series of thermal detectors provided information on various temperatures in and around the cable.
The trial was set up to verify the thermal model for the cable, measure and compare the resulting magnetic fields
and determine the effectiveness of the mitigation devices. The measured temperatures were then used to
determine any derating factors of the selected mitigation devices.
The mitigation techniques selected to be evaluated were:

 Cable Management
 Passive Loops
 Metallic Plates
 Ferromagnetic Reinforcement in Concrete Trough
Summary of outcomes of the trial
The trial provided some indications that the mitigation techniques provided some benefit but in practice it was
always less than the predicted benefit. This was thought to be due to the short nature of the mitigation installations
(being between 2 m and 11 m in length) and measurements taken 10 m from the cable would be influenced by
other parts of the cable whether shielded or not.
In general, the results showed that over the measured area (up to 10 m from the cable), the looped conductor and
copper plate both provided between 7% and 20% reduction in magnetic field. The reinforced concrete trough
provided the most consistent reduction in magnetic field with approximately a 20% reduction across the measuring
range. The high permeability material did not provide any additional significant degree of shielding over the
reinforced concrete trough.
The de-rating effect of the various methods indicated that metallic plates on top of the cable did not provide any
additional heating of the cable itself, and was therefore assessed as not de-rating the cable. It is possible that the
metallic plates acted as a heat sink, and provided a radiating surface to disperse any heat produced away from the
cable.
The passive loops were trialled in various loop dimensions and the most cost effective method (to use the loops in
the same dimensions as the excavated trench) resulted in a temperature rise of approximately 1°C in the cable.
Better results were obtained for loops that were larger than the trench, however these were considered to be
special cases and not practical for standard use in public roads. The temperature rise in this situation was very
localised and it was assessed that the propagation of the heat along the cable axis would distribute and disperse
the heat sufficiently so as not to derate the installation if used.
Page 95
Fig. 9.7.1 Example of use of Ferromagnetic reinforcement for cast in-situ trough
The reinforced concrete trough, while providing the most consistent reduction in Magnetic Field strength also
resulted in the greatest measured temperature rise in the cable, approximately 4°C. This resulted in concrete
troughs not being the preferred technique for shielding as the cables would necessarily be required to be installed
with greater spacing to compensate for the increased temperature, thus creating greater magnetic fields and
negating the benefit of the shielding properties.
Trough and reinforced trough shielding methods showed an estimated derating effect up to 5%, either in trefoil or
flat configuration. Metallic plates, non-magnetic, despite having some heat dissipation, did not show a thermal
increment on the cable temperatures. Looped conductors did not any derating effect.
Selected Mitigation Techniques
As a result of the trial, a prioritised list of the order in which the mitigation devices were to be implemented as part
of the project was created:
a) Circuit Management by moving the alignment of cable.

b) Assess the application of metallic plates or passive loops.
c) Cable Management by changing the cable configuration from flat to inverted trefoil if possible.
d) Assess application of Trough, Ducting etc.
Page 96
10 ASSESSMENT AND METHODS TO LIMIT DERATING
Measurement of (de)rating is very difficult because of the very long thermal time constant of buried cable circuits
(thermal inertia). Whereas measurement of the electromagnetic fields is very precise and mitigation techniques can
be immediately verified, their impact on rating is difficult to verify in a field situation.
The summary Table 10.1 below provides a comparison of the mitigation methods along with a quantification of the
main aspects involved in the MF mitigation design.
10 .1 Summary table
The main aspects involved in mitigation design are compared in the following Table 10.1. These aspects include
Derating [%], Shielding Factor (SF) [times], Increasing initial cost [$$$] and Increasing operational cost [$$$].
Shielding Increasing
Derating Increasing
Mitigation method Factor (SF) operational
[%] initial cost
[times] cost
Solid bonding
5 - 50 1-2 $1 $$$
(conventional screen)
Solid bonding
10 - 30 20 - 40 $$ 1 $$
(enlarged screen)
Solid bonding
(enlarged screen and 10 - 30 100 - 1000 2 $$$ 1 $
magnetic foil)
Trefoil from touching flat
0-5 1 - 1.5 0 0
(same depth)
Triangular from flat 0-5 1-2 $ 0
Split phase -5 - 0 3 20 - 200 4 $$ 0
Passive loops 0-2 2-8 $ $
Metallic plates 1-4 3 - 10 $$ $$
Ferromagnetic raceway 2 - 10 5 - 40 $$ $$
Steel pipe 4 - 20 10 - 100 $$$ $$$
Table 10.1 Summary table of the MF mitigation techniques.
Footnote 1: Cost comparison only and higher cost due to cable increased complexity.
Footnote 2: Calculated value only, not confirmed by measurements.
Page 97
Footnote 3: Negative value mean higher rating.

Footnote 4: Sensitive on distance from circuit.
Derating is relative to a conductor temperature of 90°C and unperturbed soil temperature of 20°C. Shielding Factor
is defined in Glossary and Definitions. Increasing initial cost refers to increased cost of the cable (if present), of the
shielding components and additional installation costs. Increasing operational cost comprises the capitalization of
the extra losses due to the mitigation technique and maintenance and repair of the installed components.
Cost comparisons in Table 10.1 are provided in multiple of $’s, $$$ being the most expensive. The reason for
taking this approach is the vast cost differences between countries worldwide. Final cost comparison has to be
carried out on project specific basis.
10 .2 Assessment of derating
Temperature measurements can provide important information about the actual thermal effect due to the various
mitigation devices installed in proximity of the cables. The measurement techniques are not new, but it can be easy
and not expensive to install optical fibres along the circuit to measure the temperature with a Distributed
Temperature Sensor (DTS). Also Thermocouples (TC) can give localized information of the thermal situation.
In the case of a mitigation device installed on a short part of a circuit, the effective thermal variations of laying
conditions can be practically measured only outside the cable in close proximity and not directly on the conductor.
The corresponding computed derating will comprise the effect of the mitigation device as well as the variability in
the ambient conditions, due for example to other heat sources or variation of thermal dissipation of the soil
surrounding the cables.
Many different solutions are discretely applied to enhance the rating of cables, such as special backfills, reduced
laying depth, heat pipes or forced cooling. Detailed solutions and proper designs have to be verified locally,
considering all the limiting aspects. This particular TB is based on steady state conditions and does not consider,
for the sake of simplicity, any cyclic loading or daily load factor, which can be a factor leading to a cooler cable
operation.
Page 98
11 CALCULATION METHODS AND SOFTWARE CHARACTERISTICS
Some of the mitigation techniques presented in this paper can be solved electromagnetically through application
of simple far-field approximations based on the Biot-Savart law. Examples include mitigation by passive loops (refer
to sections 6.2 and 9.2), reduced shield resistance (refer to sections 6.6.2 and 9.6.2), or by means of phase
splitting (refer to sections 6.1.3 and 9.1.3). For such cases, calculation of shielding factors and mutual
heating/derating may be feasible through application of methods and formulae given in IEC 60287 (steady state
conditions) or IEC 60853 (cyclic conditions).
Minimum complexity in thermal analysis applies for cases of single-core cables that are directly buried in uniform
soil, lying in a fixed, non-shallow configuration without any additional, nearby heat source. However, few real life
cases meet all these requirements. The introduction of IEC TR 62095 describes and discusses limitations in
application of IEC 60287 and IEC 60853, illustrating that numerical analysis will be required in many cases. While
the superposition principle may often be applied for electromagnetic problems, it is not equally applicable to
thermal problems where heat sources have overlapping thermal fields, if variation of temperatures causes
significant variations of electrical resistivity and consequent modification of heat generation.
Calculation of derating effects due to MF mitigation techniques will often require application of numerical tools.
There are numerous commercially available finite element software packages for two-dimensional (2D)
electromagnetic analysis, where conductor temperature must be estimated by other means. There are also a few
finite element packages capable of 2D multi-physics analysis of directly buried cables, for example coupled
electromagnetic/thermal analysis. The latter will also provide results on the thermal inertia of the configuration, as
exemplified in some of the preceding case studies.
As pointed out in IEC TR 62095, application of more sophisticated analysis tools make little sense if the thermal
properties of soil are unknown or associated with large uncertainties. However, it seems reasonable to assume that
a relative comparison of MF mitigation techniques can be made when using identical soil parameters (as done in
section 9.1.3).
Coupled electromagnetic/thermal FEA minimises the number of boundary conditions and parameters to be
assumed and/or specified, as computed temperature rise is iteratively converted into increased resistance and
power loss. Uncertainties are mostly related to thermal properties and thermal model boundaries. IEC TR 62095
provides guidance for the soil/air interface by defining a range for a convection coefficient, while geothermal
handbooks may be used to find the depth at which soil temperature is constant over the year.
Choosing between two- or three-dimensional (3D) analysis capabilities depends on the needs of the user.
Generally, 2D will allow significantly faster modelling and analysis, and will typically suffice for normal cable
configurations with or without MF mitigation devices. It is assumed that a typical application of MF mitigation will
apply only for some section(s) of a cable circuit. Unavoidable end effects of mitigation devices imply that they
should extend somewhat beyond the shielded section. If this is not possible, analysis of end effects will require 3D
numerical analysis. Other cases where 3D modelling might be needed include shield discontinuities (intended or
unintended), or change of configuration (for example trefoil to flat for joint bays).
Page 99
12 MF MEASUREMENT TECHNIQUES
The measurements of MF are covered by Cigré TB 375 [1], and for characterization of MF refer to Cigré TB 320
[2]. Cigré TBs deals with technical information and requirements for measuring probes and measuring procedures.
These guides focus on EMF emission from power equipment and advice is given on the good practices to
characterize EMF levels.
From the measurement point-of- view, background magnetic fields are very important because shielded circuits
are generally in urban areas where it is most probable to meet other magnetic sources: they must be identified,
reduced by switching off any sources where reasonably affordable or minimized. Some measurement devices
have three coils, measuring only the modulus of the magnetic field; other devices have single coils which may help
in distinguishing between the different sources from the field interesting in, from for example the cable or joint
chamber.
For laboratory installation, where components are tested, lateral and longitudinal profile should be tested at
various elevations, according to the use of the device. Care should be used to identify possible end effects.
For field measurement possible load variation with time should be considered as well as recording of the
measurement time, unperturbed field and field before the installation of the line.
12 .1 References
Here is a short, non-comprehensive list of relevant publications regarding MF measurement.
[1] Cigré TB 375, “Technical Guide for Measurement of Low Frequency Electric and Magnetic Fields near
Overhead Power Lines”, April 2009.
[2] Cigré TB 320, “Characterization of ELF Magnetic Fields”, April 2007.
[3] IEC standard 61786: “Measurement of low frequency magnetic and electric fields with regard to exposure of
human beings. Special requirements for instruments and guidance for measurements”, 1998.
[4] IEEE Std 1308-1994 (R2001), IEEE Recommended Practice for Instrumentation: Specifications for Magnetic
Flux Density and Electric Field Strength Meters—10 Hz to 3 kHz.
[5] IEEE Std 1460-1996 (R2002), IEEE Guide for the Measurement of Quasi-Static Magnetic and Electric Fields.
[6] IEEE Std 644-1994, IEEE Standard Procedures for Measurement of Power Frequency Electric and Magnetic
Fields From AC Power Lines.
Page 100
CONCLUSIONS
This TB presents guidelines for managing (de)rating of a HV electric system with Magnetic Field mitigation techniques,
such as cable management, passive loops, metallic plates, ferromagnetic raceways and steel pipes. The analysis is
based on IEC standards, with the addition of innovative applications and related case studies.
The presence of mitigation devices may modify the ambient conditions surrounding the cables due to increased losses
and a different capacity to dissipate heat.
The case studies reported here provide good technical solutions leading to practical installations. In fact, each technique
has a range of applications based on a Shielding Factor that can be easily reached and so it is important to fully exploit
each technique to limit the cost and the complexity of the mitigation apparatus.
Case studies have been provided in detail for theoretical, laboratory and practical installations. An overall assessment
and methods to limit the derating, plus a wise selection of the shielding methods are of great importance and are
summarized in a table.
Economical evaluation of investment costs, cost of additional losses and maintenance of the mitigation device is
discussed.
The overall conclusion is that magnetic fields from cable systems can be effectively shielded, with minimal influence on
the rating of the whole link: the designer has the understanding required to complete a sound solution to shield any HV
electric system.
Page 101
FURTHER READING
[1] IEC 60287 International Standard: Calculation of the current rating.
[2] A. Bolza, F. Donazzi, P. Maioli: "Campi elettrici e magnetici: possibilità offerte dagli elettrodotti in cavo". AEI
Conference “Elettrodotti e territorio”, Padova 22/11/2000.
[3] H. Orton: "Progress report of WGB1.23: Impact of EMF on current ratings and cable systems”, communication
presented at 2008 CIGRE’ Conference, Paris.
[4] J. Vavra, M. Wanda, “Vienna 400 kV North input”, 2006 CIGRE’ Conference, paper B1-101. [7] J. Vavra, M. Wanda,
“400 kV Vienna: the Vienna 400 kV north input”, 2006 CIGRE’ Conference, paper B1-101.
[5] R. Benato, M. Del Brenna, C. Di Mario, A. Lorenzoni, E. Zaccone, “A new procedure to compare the social costs
of EHV-HV overhead lines and underground XLPE cables”, 2006 CIGRE’ Conf., paper B1-301.
[6] P. Maioli: "Environmental constraints: electromagnetic shielding with ferromagnetic raceway”, communication
presented at 2004 CIGRE’ Conference.
[7] CIGRE TB N°320 “Characterization of ELF Magnetic Fields”, April 2007.
[8] CIGRE TB N°373 “Mitigation techniques of power-frequency magnetic fields originated from electric power systems”,
February 2009.
[9] CIGRE TB N°375 “Technical guide for measurement of low frequency electric and magnetic fields near overhead
power lines”, April 2009.
[10] P. Maioli, E. Zaccone: “Passive loops technique for electromagnetic field mitigation: applications and theoretical
th
considerations” 7 International Conference on Power Insulated Cables, JICABLE 07, 2007.
[11] Decree of the president of the council of ministers of Italy: “Establishment of exposure limits, attention values, and
quality goals to protect the population against power frequency (50 Hz) electric and magnetic fields generated by power
th
lines”, July 8 2003.
[12] P. Maioli, E. Zaccone: “Thermal design of HV electric systems with EMF mitigation devices” Cigré International
Colloquium on Power frequency Electromagnetic Fields ELF EMF, 3-4 June 2009 Sarajevo, Bosnia and Herzegovina.
[13] H. Orton, P. Maioli et al: “Impact of EMF on current rating and cable systems” 2nd International Conference on
EMF-ELF, 24-25 March 2011 Paris.
Page 102
View publication stats

CIGRE TB 559 - Impact of EMF On Current Ratings and Cable Systems

Uploaded by

Copyright:

Available Formats

You might also like

CIGRE TB 559 - Impact of EMF On Current Ratings and Cable Systems

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CIGRE TB 559 - Impact of EMF On Current Ratings and Cable Systems

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

CIGRE TB 559 - Impact of EMF on current ratings and cable systems

Technical Report · December 2013

Paolo Maioli Jean Hoeffelman

The user has requested enhancement of the downloaded file.

Impact of EMF on current ratings

7.4 Phase splitting .................................................................................. 35

Where mitigation is required, the first question to answer is:

a) to reduce the magnetic field directly above the cable circuit, or

1 GLOSSARY AND DEFINITIONS

When an RMS value is applied B is a scalar quantity.

B is related to H via permeability of the medium, that is B = µH.

Some keywords used in this TB:

Near-field and far-field:

Resultant magnetic field, BR

where Bmax and Bmin are as defined in IEC 61786.

o For the two-dimensional case:

General case: Special case #1: Special case #2:

Some abbreviations used in this TB:

ELF: Extremely Low Frequency

EMF: Electromagnetic Field

FEA: Finite Element Analysis

FEM: Finite Element Method

ICNIRP: International Commission on Non-Ionizing Radiation Protection

MF: Magnetic Field

SF: Shielding Factor

A more complete list of physical quantities [SI units]:

H: magnetic field vector (or magnetic field intensity) [A/m]

BR: resultant magnetic field [T]

t: time variable [s]

: angular frequency [s-1]

: magnetic permeability [H/m]

r: relative magnetic permeability [no unit]

: electrical conductivity [S/m]

δ: skin depth or penetration depth [m]

2 SCOPE OF WORK AND TERMS OF REFERENCE

2.1 Scope of work

2.2 Terms of reference

1 To define correct terminology for field management techniques.

3.1 Salient References

[3] Cigré TB 320, “Characterization of ELF Magnetic Fields”, April 2007.

4 CURRENT (DE)RATING PRINCIPLES

5 DERATING DEFINITION AND PRINCIPLES

Derated Operation (as defined by IEEE Dictionary).

6 EMF MITIGATION METHODS: IMPACT ON CABLE RATINGS

I is the conductor current [A],

r is the distance of the computation point from circuit centre [m].

d is the interaxial distance between the conductors [m],

k is the geometric coefficient (refer e.g. to TB 373, page 27)

6.1 Cable management

6.1.1 SINGLE CIRCUIT

Straight-forward methods to reduce the magnetic field are:

 to increase the cable laying depth,

6.1.2 DOUBLE CIRCUIT

6.1.3 SPECIAL CASES - OPERATIONAL MODES

6.1.3.1 Split phases

L1 I1 = I ej0° I3 = I/ 2 ej120° L3 L2 I2 = I/ 2 e-j120°

6.1.3.2 Six-phase system