Copper For Busbars All Sections

Copper for Busbars
Guidance for Design and Installation

Copper for Busbars
David Chapman & Professor Toby Norris
Copper Development Association Publication No 22
European Copper Institute Publication No Cu0201
Revised May 2014
First issued 1936 In this new edition the calculation of current-carrying capacity has been
2nd-3rd revisions 1936-1950 greatly simplified by the provision of exact formulae for some common busbar
4th revision 1950 configurations and graphical methods for others. Other sections have been
5th revision 1952 updated and modified to reflect current practice.
6th-10th revisions 1954-1959
11th revision 1960
12th revision 1962
Reprinted 1964
Reprinted 1965
13th revision 1984
14th revision 1996
Copper Development Association is a non-trading organisation that promotes

and supports the use of copper based on its superior technical performance
and its contribution to a higher quality of life. Its services, which include the
provision of technical advice and information, are available to those interested
in the utilisation of copper and copper alloys in all their aspects. The Association
also provides a link between research and the user industries and is part of an
international network of trade associations, the Copper Alliance™.
Founded in 1996, ECI is a joint venture between the International Copper

Association, Ltd (ICA), headquartered in New York, representing the majority of
the world’s leading mining companies, custom smelters and semi-fabricators,
and the European copper industry. ECI is also part of the Copper Alliance, an
international network of industry associations. Its shared mission is to work,
with its members, to defend and grow markets for copper based on its superior
technical performance and contributions to a higher quality of life.
David Chapman
David Chapman was the Electrical Programme Manager for Copper Development Association in the UK, where his main interests included power
quality and energy efficiency. He was an author and Chief Editor of the LPQI Power Quality Application Guide.
Professor Toby Norris

Toby Norris is an electrical engineer who has worked in industry and at university. A central interest has been electromagnetic theory, especially in
electric power plant and including superconductors.
Cover page picture acknowledgment:

High precision copper busbars
(Courtesy of H V Wooding Ltd)
2 | COPPER FOR BUSBARS

Contents
1.0 Introduction.................................................6 3.0 Life Cycle Costing..................................... 56

1.1 About this Guide ................................................................................. 6 3.1 Introduction.........................................................................................56
1.2 Materials for Busbars......................................................................... 6 3.1.1 The Future Value of Money –
1.2.1 Material Requirements...................................................6 Net Present Value............................................................56
1.2.2 Material Choice..................................................................7 3.1.1.1 Monthly Payments..........................................................59
1.2.2.1 High Conductivity.............................................................7 3.1.2 Loading................................................................................60
1.2.2.1.1 Effect of Temperature on Conductivity...................9 3.1.3 Lifetime................................................................................60
1.2.2.1.2 Effect of Cold Work on Conductivity........................9 3.1.4 Sensitivity Analysis.........................................................60
1.2.2.2 Mechanical Strength.......................................................9 3.2 Application to an Electrical Installation...................................60
1.2.2.2.1 Tensile Strength...............................................................10 3.2.1 Installation Design..........................................................61
1.2.2.2.2 Proof Strength..................................................................10 3.2.2 Installation Costs.............................................................61
1.2.2.2.3 Hardness.............................................................................11 3.2.3 Recurring Costs................................................................61
1.2.2.2.4 Resistance to Softening...............................................11 3.2.3.1 Maintenance Costs.........................................................61
1.2.2.2.5 Creep Resistance.............................................................11 3.2.3.2 Energy..................................................................................61
1.2.2.2.6 Fatigue Resistance..........................................................12 3.2.4 End of Life Costs..............................................................68
1.2.2.2.7 Bending and Forming....................................................13 3.3 Conclusion............................................................................................68
1.2.2.3 Connectivity......................................................................13
1.2.2.4 Maintenance.....................................................................13 4.0 Short-Circuit Effects................................ 69
1.2.3 Types of High Conductivity Copper Available.....14 4.1 Introduction..........................................................................................69
1.2.3.1 Tough Pitch Copper, CW004A and CW005A 4.2 Short-Circuit Heating of Bars........................................................69
(C101 and C102)..............................................................14 4.3 Electromagnetic Stresses.................................................................70
1.2.3.2 Oxygen-free High Conductivity Copper, 4.3.1 Estimating the Forces Between Parallel
CW008A (C103) .............................................................14 Sets of Bars........................................................................72
1.2.4 Available Forms................................................................14
4.3.1.1 Round Bars........................................................................72
4.3.1.1.1 Triangular Array...............................................................73
2.0 Current-Carrying Capacity of Busbars...........15 4.3.1.1.2 In-Line Array.....................................................................74
2.1 Design Philosophy.............................................................................15
4.3.1.2 Bars of Rectangular Section.......................................74
2.2 Calculation of Maximum Current-Carrying Capacity..........15
4.4 Mounting Arrangements................................................................76
2.2.1 Methods of Heat Loss....................................................15
4.4.1 Maximum Permissible Stress......................................76
2.2.1.1 Convection – Natural Air Cooling............................16
4.4.1.1 Moment of Inertia..........................................................77
2.2.1.2 Convection Heat Loss - Forced Air Cooling..........17
4.4.2 Deflection...........................................................................78
2.2.1.3 Radiation............................................................................18
4.4.3 Natural Frequency..........................................................78
2.2.2 Heat Generated by a Conductor...............................21
2.2.2.1 Alternating Current Effects – the Factor S...........22
2.2.2.1.1 Skin Effect and Skin Depth.........................................24 5.0 Busbar Profiles.......................................... 79
2.2.2.1.2 Proximity Factor, Sp........................................................27 5.1 Introduction.........................................................................................79
2.3 Conclusion............................................................................................27 5.2 Reasons for Using Profiles.............................................................79
2A Shape and Proximity Factors for Typical Configurations..28 5.2.1 Skin Effect Reduction....................................................79
2A.1 Skin and Proximity Factors for Common 5.2.2 Weight and Cost Saving...............................................79
Busbar Shapes..................................................................28 5.2.3 Integrated Fixings and Mountings...........................80
2A.1.1 Single Solid Rods.............................................................28 5.2.4 Retention of Intellectual Integrity...........................80
2A.1.1.1 Shape Factor for Single Solid Rods.........................28 5.3 Economics of Profiles.......................................................................80
2A.1.1.2 Proximity Factor for Single Solid Rods..................29 5.4 Practical Profiles.................................................................................80
2A.1.2 Single Tubes.......................................................................34 5.4.1 Manufacturing Process.................................................80
2A.1.2.1 Shape Factors for Single Tubes.................................34 5.4.1.1 EN 13605............................................................................80
2A.1.2.2 Proximity Factors for Tubes........................................37 5.4.1.1.1 Straightness, Flatness and Twist...............................82
2A.1.3 Single Rectangular Sections.......................................42 5.4.2 Design for Manufacturing...........................................84
2A.1.3.1 Shape Factor for Rectangular Bars ........................42 5.4.2.1 Wall Thickness..................................................................84
2A.1.3.2 Approximate Portmanteau Formula for 5.4.2.2 Avoid Sharp Corners......................................................84
Single Bars ........................................................................45 5.4.2.3 Symmetry...........................................................................84
2A.1.4 Parallel Bars.......................................................................45 5.4.2.4 Be Compact.......................................................................85
2A.1.4.1 Proximity Factor - Anti-parallel Currents.............45 5.4.2.5 Avoid Deep Narrow Channels....................................85
2A.1.4.2 Proximity Factor - Parallel Currents........................47 5.4.2.6 Avoid Hollow Chambers...............................................85
2A.1.5 Three-phase Configurations.......................................49 5.4.3 Functional Design...........................................................85
2A.1.5.1 Linear Plots........................................................................49
2A.1.5.2 Logarithmic Plots............................................................53
COPPER FOR BUSBARS | 3

Contents Tables & Figures
5.5 Electrical Design Considerations................................................85

5.5.1 Skin Effect.........................................................................85
5.5.2 Thermal Dissipation.......................................................86
Tables
5.5.3 Jointing and Mounting................................................87 Table 1 Properties of Typical Grades of Copper and Aluminium.7
5.5.4 Short Circuit Performance - Table 2 Properties of 100% IACS Copper.............................................. 7
Moment of Inertia.........................................................88 Table 3 Implied Properties of 100% IACS Copper.............................. 8
5.6 Calculation of Moment of Inertia of Complex Sections....88 Table 4 Comparison of Creep Properties of HC Copper and
Aluminium........................................................................................12
6.0 Jointing of Copper Busbars...................... 90 Table 5 Comparison of Fatigue Properties of HC Copper and
6.1 Introduction........................................................................................90 Aluminium........................................................................................12
6.2 Busbar Jointing Methods...............................................................90 Table 6 Minimum Bend Radius of HC Copper...................................13
6.3 Joint Resistance.................................................................................91 Table 7 Self-extinguishing Arcs in Copper and Aluminium
6.3.1 Streamline Effect............................................................92 Busbars..............................................................................................13
6.3.2 Contact Resistance........................................................94 Table 8 Discount Factors for Various Discount Rates....................57
6.3.2.1 Condition of Contact Surfaces.................................94 Table 9 Maximum Working Current for a Range of
6.3.2.2 Effect of Pressure on Contact Resistance............94 Busbar Sizes.....................................................................................62
6.4 Bolting Arrangements.....................................................................98 Table 10 Resistance and Power Loss at 500 Amps............................62
6.4.1 Joint Efficiency................................................................99 Table 11 Energy Cost Per Metre for Various Widths of
6.5 Clamped Joints.................................................................................101 Copper Bars at 500 A Load........................................................63
6.6 Degradation Mechanisms............................................................101 Table 12 Present Value (€) per Metre of Bar........................................67
6.6.1 Oxidation.........................................................................101 Table 13 Power Factor and Peak Current............................................... 71
6.6.2 Corrosion.........................................................................102 Table 14 Tolerances for Dimensions b and h for bmax
or hmax <20:1...................................................................................82
6.6.3 Fretting.............................................................................102
Table 15 Tolerances for Dimensions b and h for bmax
6.6.4 Creep and Stress Relaxation....................................102
or hmax ≥ 20:1..................................................................................82
6.6.5 Thermal Expansion.......................................................102
Table 16 Thickness Tolerances....................................................................82
6.7 Conclusion.........................................................................................102
Table 17 Coefficient for Twist Tolerance................................................84
Table 18 Maximum Sizes of Profile According to Two
Annex: Coatings............................................. 103 Manufacturers................................................................................84
A1.0 Introduction................................................................................... 103
Table 19 Nut Factors for Different States of Lubrication................96
A2.0 Reasons for Coating................................................................... 103
Table 20 Proof Strength and Coefficient of Thermal
A2.1 Coating to Provide Electrical Insulation............ 103
Expansion for Copper and Typical Bolt Materials............96
A2.2 Coating to Inhibit Corrosion................................... 103 Table 21 Typical Thread Characteristics..................................................97
A2.2.1 Metal Coatings............................................................. 103 Table 22 Typical Busbar Bolting Arrangements...................................99
A2.2.2 Non-Metallic Coatings.............................................. 103
A2.3 Coating to Increase Current Rating.................... 103
A2.4 Coating for Cosmetic Purposes............................. 104
A2.5 Coating to Improve Joint Performance.............. 104 Figures
A3.0 Methods of Coating.................................................................... 104
A3.1 Factory Application Methods................................. 104 Figure 1 Effect of small concentrations of impurities on the
A3.1.1 Extrusion......................................................................... 104 resistivity of copper.................................................................... 8
A3.1.2 Powder Coating............................................................ 105 Figure 2 Effect of cold rolling on mechanical properties and
A3.2 On-Site Application Methods................................. 105 hardness of high conductivity copper strips..................10
A3.2.1 Heat-Shrinkable Sleeve............................................. 105 Figure 3 Typical creep properties of commercially pure
A3.2.2 Painting........................................................................... 105 copper and aluminium.............................................................12
A3.2.2.1 Alkyd Paints................................................................... 105 Figure 4 Heat dissipation by convection from a vertical surface
A3.2.2.2 Two-Part Epoxy............................................................ 105 for various temperature rises above ambient................16
A4.0 Inspection and Maintenance................................................... 105 Figure 5 Convection loss from typical bar sections .....................17
Figure 6 Heat dissipation by radiation from a surface assuming
relative emissivity of 0.5 and surroundings at 30ºC...18
Figure 7 Radiation loss from typical bar sections..........................19
Figure 8 Convection and radiation losses at various
temperatures................................................................................20
Figure 9 Total heat losses for a single bar of various
heights against temperature rise........................................20
Figure 10 Total heat losses for each bar of a parallel pair of
various heights against temperature rise........................21

Figure 11 Parameter p versus cross-sectional area in mm2 Figure 42 Total cost per metre against current density for a
for typical copper at 80˚C.......................................................23 range of operation times........................................................66
Figure 12 Resistivity of typical HC copper (101.5% IACS) as a Figure 43 Estimated working temperature versus width
function of temperature.........................................................25 of bar (mm) for 500 A load....................................................66
Figure 13 Skin depth of typical HC copper (101.5% IACS) at Figure 44 Cost per metre (€) against bar width for a) 20 000
50 Hz, 60 Hz, 400 Hz as a function of temperature....25 hours operation without discount and b) 2000 hours
Figure 14 dc resistance of typical HC Copper (101.5% IACS) per year for 10 years at 5% discount................................67
versus area at 20ºC and 80ºC................................................26 Figure 45 Short-circuit current waveform.......................................... 71
Figure 15 Plots of shape factor versus γ..............................................29 Figure 46 Three phase system with spacings D (mm),
Figure 16 Factor A for round bars as a function of g......................30 (a) triangular array (b) inline array.....................................73
Figure 17 Proximity factor, SP, for single phase systems with Figure 47 Factor K for calculating the force between
parallel round bars....................................................................31 two bars of rectangular section...........................................75
Figure 18 Mean proximity factors for flat arrangement of round Figure 48 Form factor Kb. (a) low values of a/b (long sides
bars carrying balanced three phase currents.................32 facing each other) (b) high values of a/b.......................... 75
Figure 19 Mean proximity factors for delta arrangement of Figure 49 Typical copper profiles.............................................................79
round bars carrying balanced three phase currents...33 Figure 50 A non-typical profile indicating the dimensions
Figure 20 Shape factor for tubes.............................................................35 used in the standard..................................................................81
Figure 21 Shape factor for tubes with low values of Figure 51 Measurement of straightness................................................83
shape factor.................................................................................35 Figure 52 Measurement of flatness.........................................................83
Figure 22 Shape factor for tubes.............................................................36 Figure 53 Measurement of twist...............................................................83
Figure 23 The shape factor computed from the Bessel function Figure 54 Profile cross-section showing mounting lugs
formula using √(f/rdc) as the frequency parameter...36 and slots for bolt-head.............................................................85
Figure 24 Shape factor as a function of the ratio, ...37 Figure 55 Rac /Rdc as a function of the parameter √f /R dc
with f in Hz and Rdc in µΩ/m.................................................86
Figure 25 Factor A versus β = t/a for values of g = t/ δ.................38
Figure 56 Thermal image of two profiles under load.......................87
Figure 26 Proximity factor for single phase tubes as function
Figure 57 Hole-free joints............................................................................87
of the factor A for values of η=s/2a=s/d........................38
Figure 58 Calculation of moment of inertia of a complex
Figure 27 Proximity factor for single phase tubes...........................39
shape................................................................................................88
Figure 28 Factor A for proximity loss factor as a function of
Figure 59 Parameters used in the calculation of moment
g = t/ δ for various values of β = t/a................................40
of inertia of each element.......................................................89
Figure 29 Proximity factor, Sp , for round bars at various
Figure 60 A typical bolted joint.................................................................90
spacings designated by h.......................................................41
Figure 61 A simple clamped joint.............................................................90
Figure 30 Rac /Rdc as a function of the parameter √f /R dc.............. 43
Figure 62 A riveted joint...............................................................................91
Figure 31 Rac /Rdc as a function of the parameter √f /R dc ..............43
Figure 63 A soldered joint............................................................................91
Figure 32 Shape factor, Rac /Rdc, as a function of the parameter
Figure 64 A welded joint..............................................................................91
√f / R dc .............................................................................................44
Figure 65 An overlapped joint....................................................................91
Figure 33 Shape factor versus 2b/ d for various values of b/a..44
Figure 66 Streamline effect in overlapped joints...............................92
Figure 34 Plots of the shape ratio versus the ratio for Figure 67 Bolt placement in overlapped joints...................................93
various values of for anti-parallel currents.........47 Figure 68 Overlap joint between bars with angled ends................93
Figure 69 The effect of pressure on the contact resistance
Figure 35 Shape factor S = as a function of the separation
of a joint.........................................................................................95
s for various values of with Rdc in μΩ/m Figure 70 Possible bolting techniques for copper busbars............98
and f in Hz ...................................................................................49 Figure 71 Joint with a longitudinal slot.................................................98
Figure 36 Linear plots of Rac /Rdc versus p where p ≈ 1·5853
Disclaimer
with Rdc in μΩ/m and f in Hz ..............................................52

Figure 37 Quasi logarithmic plots of Rac /Rdc versus p where While this publication has been prepared with care, Copper
p ≈ 1·5853 with Rdc in μΩ/m and f in Hz.............55 Development Association, European Copper Institute and other
contributors provide no warranty with regards to the content
Figure 38 Power loss versus width for 6.3 mm thick and shall not be liable for any direct, incidental or consequential
copper bars...................................................................................63 damages that may result from the use of the information or the
Figure 39 Cost of energy loss per metre versus width for data contained.
6.3 mm thick copper bars.......................................................64
Figure 40 Total cost per metre versus bar width for Copyright© Copper Development Association and European Copper
5000 hours operation at 500 A............................................64 Institute.
Figure 41 Total cost per metre of bar versus bar width
for a range of operating times.............................................65

1.0 Introduction
David Chapman
1.1. About this Guide
Busbars are used within electrical installations for distributing power from a supply point to a number of output circuits. They may be used in
a variety of configurations ranging from vertical risers, carrying current to each floor of a multi-storey building, to bars used entirely within a
distribution panel or within an industrial process.
The issues that need to be addressed in the design of busbar systems are:
• Temperature rise due to energy losses

• Energy efficiency and lifetime cost
• Short-circuit current stresses and protection
• Jointing methods and performance
• Maintenance.
This book provides the information needed to design efficient, economic and reliable busbar systems.
In any electrical circuit some electrical energy is lost as heat which, if not kept within safe limits, may impair the long term performance or the
safety of the system. For busbar systems, the maximum working current is determined primarily by the maximum tolerable working temperature,
which is, in turn, determined by considerations such as safety, the retention of mechanical properties of the conductor, compatibility with
mounting structures and cable connections. ‘2.0 Current-Carrying Capacity of Busbars’ discusses how to estimate the working current and
temperature.
A higher working temperature means that energy is being wasted. Designing for lower energy loss requires the use of more conductor material but
results in more reliable operation due to the lower working temperature and, because the cost of lifetime energy losses is far greater than the cost
of first installation, lower lifetime costs. The process of assessing the life cycle cost of a busbar system is described in ‘3.0 Life Cycle Costing’.
Because of the large currents involved, short circuit protection of busbar systems needs careful consideration. The important issues are the
temperature rise of the busbar during the event and the magnitude of the forces generated by the high current, which may cause deformation of
the bars and the failure of mountings. The design of the mounting system is an important factor and one that is becoming more important with
the increase in harmonic currents, which can trigger mechanical resonances in the busbar. ‘4.0 Short-Circuit Effects’ discusses these issues.
It is usually necessary to joint busbars on site during installation and this is most easily accomplished by bolting bars together or by welding.
For long and reliable service, joints need to be carefully made with controlled torque applied to correctly sized bolts. A properly designed and
implemented joint can have a resistance lower than that of the same length of plain bar. The design of efficient joints is discussed in ‘6.0 Jointing’.
The remainder of this Introduction presents reference material giving mechanical and electrical properties of copper that are required for design
purposes.
1.2 Materials for Busbars
1.2.1 Material Requirements
To achieve a long and reliable service life at the lowest lifetime cost, the conductor material needs the following properties:
• Low electrical and thermal resistance

• High mechanical strength in tension, compression and shear
• High resistance to fatigue failure
• Low electrical resistance of surface films
• Ease of fabrication
• High resistance to corrosion
• Competitive first cost and high eventual recovery value.
This combination of properties is best met by copper. Aluminium is the main alternative material, but a comparison of the properties of the two
metals shows that, in nearly all respects, copper is the superior busbar material.

1.2.2 Material Choice
Busbars are generally made from either copper or aluminium. For a complete list of mechanical properties and compositions of copper used
for busbars, see BS EN 13601: 2013 Copper rod, bar and wire for electrical purposes. Table 1 below gives a comparison of some electrical and
mechanical properties. It can be seen that for conductivity and strength, high conductivity (HC) copper is far superior to aluminium. The only
disadvantage of copper is its higher density, which results in higher weight. The greater hardness of copper compared with aluminium gives it
better resistance to mechanical damage, both during erection and in service. Copper bars are also less likely to develop problems in clamped joints
due to cold metal flow under the prolonged application of a high contact pressure. The higher modulus of elasticity of copper gives it greater
beam stiffness compared with an aluminium conductor of the same dimensions. The temperature variations encountered under service conditions
require a certain amount of flexibility to be allowed for in the design. The lower coefficient of linear expansion of copper reduces the degree of
flexibility required.
Table 1 – Properties of Typical Grades of Copper and Aluminium

Property (at 20°C) Copper (C101) Aluminium (1350) Units
Electrical conductivity (annealed) 101 61 % IACS
Electrical resistance (annealed) 17.2 28.3 nΩ mm
Temperature coefficient of resistivity 0.0039 0.004 per °K
Thermal conductivity 397 230 W/m°K
Specific heat 385 900 J/kg °K
Coefficient of expansion 17 x 10-6 23 x 10-6 per °K
Tensile strength (annealed) 200-250 50-60 N/mm2
Tensile strength (half hard) 260-300 85-100 N/mm2
0.2% proof strength (annealed) 50-55 20-30 N/mm2
0.2% proof strength (half hard)) 170-200 60-65 N/mm2
Elastic modulus 116-130 70 kN/mm2
Density 8910 2700 kg/m2
Melting point 1083 660 °C
1.2.2.1 High Conductivity
The electrical properties of HC copper were standardised in 1913 by the International Electrotechnical Commission, which defined the International
Annealed Copper Standard (IACS) in terms of the following properties at 20°C:
Table 2 – Properties of 100% IACS Copper

Volume conductivity, sv 58 MS/m
Density, d 8890 kg/m³
Temperature coefficient of resistance, a 0.00393/°C
It follows from the first two of these values that:

Table 3 – Implied Properties of 100% IACS Copper
Volume resistivity, rv ( = 1/s v ) 1.7241 mΩcm
Mass conductivity, sm ( = sv /d ) 6524 Sm²/kg
Mass resistivity, rm ( = 1/sm) 153.28 μΩkg/m²
These values correspond to 100% IACS. Since that time, processing technology has improved to the point where routine day-to-day-production
HC copper can reach a conductivity of 102% IACS or even higher. It has become normal practice to express the conductivity of all electrical alloys,
including aluminium, in terms of IACS.
Aluminium has a lower conductivity than copper (61% IACS), so an equivalent resistance conductor would have a cross-sectional area which is
a factor of 1.6 times greater than one made from copper. Taking into account the difference in density, the aluminium conductor would have
half the weight of copper. Copper will therefore allow a more compact installation and, in many systems, space considerations are of greater
importance than weight. Greater weight may make installation slightly more difficult but, since the electromagnetic stresses set up in the bar are
usually more severe than the stress due to its weight, the mounting design is largely unaffected.
Although busbar systems should normally be designed for lowest lifetime cost – which means a lower working temperature to reduce waste
energy costs – the ability of copper to maintain its mechanical properties at higher temperatures provides extra capacity and safety during short
circuit conditions. Copper also allows the use of higher peak operating temperatures in special circumstances.
185
Ti P Co Fe As Si Mn Ge Cr Nb Y Sb Al Sn Mg
180 85
Ni
V Zn
175 Cd Pb
Mass resistivity (microhm.kg.m2)
Zr
170 Te 90
Conductivity (% IACS)
Hf
165
160 95
In
O
155
Ag 100
150
103
145
0 0.05 0.10 0.15 0.20 0.25 0.30

Concentration of impurity by weight (%)
Figure 1 – Effect of small concentrations of impurities on the resistivity of copper
The degree to which the electrical conductivity is affected by an impurity depends largely on the element present, as illustrated in Figure 1. For
example, the presence of only 0.04% phosphorus reduces the conductivity of HC copper to around 80% IACS. Normally, the purest copper is used
for bulk conductors. In particular, oxygen-free copper is usually specified, not only for its high conductivity but also because it can be welded
without the risk of hydrogen embrittlement.
When enhanced mechanical properties - such as increased strength or wear resistance - are required, a silver-bearing (up to 0.12%) copper can be
used at the expense of a small increase in resistivity. Many other electrical alloys are available, but they are intended for special purposes, such as
contacts or terminals, and are not used as bulk conductors.

1.2.2.1.1 Effect of Temperature on Conductivity
Conductivity varies with temperature so it is important that the correct value is used for the actual working temperature. The resistance of a
copper conductor at temperatures up to 200°C can be calculated from:

where:
R20 is the conductor resistance at a temperature of 20°C, in Ω
α20 is the temperature coefficient of resistance at 20°C, per K. α = 0.0039 for copper
ΔT = Tk –20 is the temperature difference, in degrees K
Tk is the final temperature, in K.
Note that α is temperature dependent; the value given will yield sufficiently accurate results up to 200°C. An alternative equation, which gives
sufficiently accurate results for engineering purposes over a wide temperature range, is:

where:
R is the resistance at temperature T
R20 is the resistance at 20°C (293°K)
T is the temperature in °K, i.e. temperature in °C + 273.
1.2.2.1.2 Effect of Cold Work on Conductivity
The conductivity of copper is decreased by cold working and may be 2 to 3% less in the hard drawn condition than when annealed. Thus standards
for hard drawn HC copper products should stipulate a minimum conductivity requirement of 97% IACS compared with 100% IACS for annealed
products.
An approximate relationship between tensile strength of cold worked copper and its increase in electrical resistivity is:
P = F/160
where:
P is % increase in electrical resistivity of cold worked copper over its resistivity when annealed
F is tensile strength, N/mm².
1.2.2.2 Mechanical Strength
The mechanical strength of the busbar material is important to ensure that the material is not deformed during transport and assembly, does not
sag over an extended working life at maximum temperature, does not creep under pressure leading to loosened joints and does not permanently
distort under short circuit loads.
The mechanical properties are influenced by the production processes employed. Most busbar materials will have been extruded and drawn
resulting, typically, in a ‘half-hard’ material. The effect on the mechanical properties of cold work by rolling (to reduce the thickness) is shown in
Figure 2. Cold working of the material has the effect of raising the tensile strength, proof strength and hardness, but reducing its elongation.

1.2.2.2.1 Tensile Strength
In the ‘as-cast’ condition, HC copper has a tensile strength of 150-170 N/mm². The changes in structure brought about by hot working raise the
tensile strength to the order of 200-220 N/mm².
The maximum tensile strength obtainable in practice depends on the shape and cross-sectional area of the conductor. The larger the cross-
sectional area of a conductor the lower its tensile strength, since the amount of cold work that can be applied is limited by the reduction in area
which can be achieved.
For the usual sizes of busbar conductors in the hard-condition, tensile strengths from 250 N/mm² up to 340 N/mm² can be obtained depending on
the cross-sectional area.
1.2.2.2.2 Proof Strength
The ‘proof strength’ is the stress required to produce a defined amount of permanent deformation in the metal and is a valuable guide to its
mechanical properties. Proof strength is defined as the stress at which a non-proportional elongation equal to a specified percentage (usually
0.2%) of the original gauge length occurs.
As with the tensile strength, the proof strength varies with the amount of cold work put into the material (see Figure 2).
400 80
350 70
Tensile Strength
300 60
0.2% Proof Strength

Tensile strength and 0.2% Proof strength (N/mm2)
50 Elongation (%)
250
200 40 120
Hardness
150 30 100
Hardness (HV)
100 20 80
50 10 60
Elongation
0 0 40
0 10 20 30 40 50 60 70
Reduction of thickness by rolling (%)
Figure 2 - Effect of cold rolling on mechanical properties and hardness of high conductivity copper strips

1.2.2.2.3 Hardness
Standards applicable to busbar conductors do not specify hardness measurement as part of the testing requirements. It can however be more
quickly and easily carried out than a tensile test and is therefore convenient as a guide to the strength of a conductor. The results have to be used
with discretion for two reasons:
• Unlike ferrous materials, the relationship between hardness and tensile strength is not constant (see Figure 2).
• A hardness test is usually only a measurement of the outer skin of the material tested. If the conductor is of large cross-sectional
area and has received a minimum amount of cold work, the skin will be harder than the underlying metal. Consequently, variations in
hardness may be obtained dependent on where the measurement is made in relation to its cross-section.
As a guide, typical hardness figures of the temper range of conductors supplied are:
Annealed (O) 60 HV max

Half-hard (½H) 70-95 HV
Hard (H) 90 HV min.
1.2.2.2.4 Resistance to Softening
It is well known that the exposure of cold worked copper to elevated temperatures results in softening and mechanical properties typical of those
of annealed material. Softening is time and temperature dependent and it is difficult to estimate precisely the time at which it starts and finishes.
It is usual therefore to consider the time to ‘half-softening’, i.e. the time taken for the hardness to fall by 50% of the original increase in hardness
caused by cold reduction.
In the case of HC copper, this softening occurs at temperatures above 150°C. It has been established experimentally that such copper would
operate successfully at a temperature of 105°C for periods of 20-25 years, and that it could withstand short-circuit conditions as high as 250°C for
a few seconds without any adverse effect.
If hard drawn conductors are required to retain strength under operating conditions higher than normal, the addition of small amounts of silver at
the melting and casting stage produces alloys with improved resistance to softening. The addition of 0.06% silver raises the softening temperature
by approximately 100°C without any significant effect on its conductivity, at the same time appreciably increasing its creep resistance.
1.2.2.2.5 Creep Resistance
Creep, another time and temperature dependent property, is the non-recoverable plastic deformation of a metal under prolonged stress. The ability
of a metal to resist creep is of prime importance to design engineers.
From published creep data, it can be seen that high conductivity aluminium exhibits evidence of significant creep at ambient temperature if heavily
stressed. At the same stress, a similar rate of creep is only shown by high conductivity copper at a temperature of 150°C, which is well above the
usual operating temperature of busbars.

0.06
0.05 ºC
1 50
Cu,
HC
led, 20ºC
50
ºC
0.04 ,1
ed
Min creep rate, % per 1000 hours
n eal
), annea
, an
OF
Cu
0.03
Al (1080
0.02
Al (1080), annealed 80ºC
0.01
0
0 10 20 30 40
Stress, N/mm2
Figure 3 - Typical creep properties of commercially pure copper and aluminium
Table 4 - Comparison of Creep Properties of High Conductivity Copper and Aluminium

Material Testing Temp (°C) Min Creep Rate (% per 1000h) Stress (N/mm²)
Al (1080) annealed 20 0.022 26 *
HC Cu annealed 150 0.022 26 *
Cu-0.086% Ag 50% c.w. 130 0.004 138
Cu-0.086% Ag 50% c.w. 225 0.029 96.5
* Interpolated from Figure 3
The creep resistance of oxygen-free HC copper is better than that of tough pitch HC copper. This is due to the very small amounts of impurities
which remain in solid solution in oxygen-free copper, but which are absorbed in the oxide particles in tough pitch copper. Some typical
observations are shown in Figure 3. Tough pitch copper creeps relatively rapidly under low stress at 220°C. The addition of silver to both oxygen-
free and tough pitch coppers results in a significant increase in creep resistance.
1.2.2.2.6 Fatigue Resistance
Fatigue is the mechanism leading to fracture under repeated or fluctuating stresses. Fatigue fractures are progressive, beginning as minute cracks
which grow under the action of the stress. As the crack propagates, the load bearing area is reduced and failure occurs by ductile fracture before
the crack develops across the full area.
Table 5 - Comparison of Fatigue Properties of High Conductivity Copper and Aluminium

Material Fatigue Strength (N/mm²) No of Cycles x 106
HC Aluminium annealed 20 50
half-hard (H8) 45 50
HC Copper annealed 62 300
half-hard 115 300

Conditions for such failures can be set up in a busbar system rigidly clamped for support and then subjected to vibrating conditions. Where higher
stresses or working temperatures are to be allowed for, copper containing small amounts of silver (about 0.1%) is used. The creep resistance and
softening resistance of copper-silver alloys increase with increasing silver content.
In the conditions in which high conductivity aluminium and copper are used, either annealed (or as-welded) or half-hard, the fatigue strength
of copper is approximately double that of aluminium. This gives a useful reserve of strength against failure initiated by mechanical or thermal
cycling.
1.2.2.2.7 Bending and Forming
The high conductivity coppers are ductile and, in the correct temper, will withstand severe bending and forming operations. As a general guide to
bending, copper in the half-hard or hard temper will bend satisfactorily round formers of the following radii:
Table 6 – Minimum Bend Radius of High Conductivity Copper

Thickness (t) Minimum Bend Radius
Up to 10 mm 1t
11-25 mm 1.5 t
26-50 mm 2t
Material of thicknesses greater than 50 mm is not normally bent; however, it is possible to do so by localised annealing prior to bending.
1.2.2.3 Connectivity
The surface of copper naturally oxidises, forming a thin hard layer on the surface which normally prevents further oxidation. This oxide film is
conductive (under busbar conditions where it is sandwiched between two copper electrodes) so it does not affect the quality of bolted or clamped
joints. On the other hand, exposed aluminium surfaces rapidly form a hard insulating film of aluminium oxide, which makes jointing very difficult
and can lead to long term reliability problems. Terminals, switch contacts and similar parts are nearly always produced from copper or a copper
alloy. The use of copper for the busbars to which these parts are connected therefore avoids contacts between dissimilar metals and the inherent
jointing and corrosion problems associated with them.
1.2.2.4 Maintenance
The higher melting point and thermal conductivity of copper reduce the possibility of damage resulting from hot spots or accidental flashovers
in service. If arcing occurs, copper busbars are less likely to support the arc than aluminium. Table 7 shows that copper can self-extinguish arcs
across smaller separations, and at higher busbar currents. This self-extinguishing behaviour is related to the much larger heat input required to
vaporise copper than aluminium.
Table 7 - Self-extinguishing Arcs in Copper and Aluminium Busbars

Copper Aluminium
Minimum busbar spacing, mm 50 100
Maximum current per busbar, A 4500 3220

Copper liberates considerably less heat during oxidation than aluminium and is therefore much less likely to sustain combustion in the case of
accidental ignition by an arc. The large amounts of heat liberated by the oxidation of aluminium in this event are sufficient to vaporise more metal
than was originally oxidised. This vaporised aluminium can itself rapidly oxidise, thus sustaining the reaction. The excess heat generated in this
way heats nearby materials, including the busbar itself, the air and any supporting fixtures. As the busbar and air temperatures rise, the rates of
the vaporisation and oxidation increase, so accelerating the whole process. As the air temperature is increased, the air expands and propels hot
oxide particles. The busbar may reach its melting point, further increasing the rate of oxidation and providing hot liquid to be propelled, while
other materials, such as wood panels, may be raised to their ignition temperatures. These dangers are avoided by the use of copper busbars.
Finally, copper is an economical conductor material. It gives long and reliable service at minimum maintenance costs and, when an installation
is eventually replaced, the copper will have a high recovery value. Because of its many advantages, copper is still used worldwide as an electrical
conductor material despite attempts at substitution.
1.2.3 Types of High Conductivity Copper Available
1.2.3.1 Tough Pitch Copper, CW004A and CW005A (C101 and C102)
Coppers of this type, produced by fire-refining or remelting of electrolytic cathode, contain a small, deliberate addition of oxygen which scavenges
impurities from the metal. The oxygen is present in the form of fine, well-distributed cuprous oxide particles only visible by microscopic
examination of a polished section of the metal. Typical oxygen contents of these coppers fall in the range 0.02-0.05%. Between these limits the
presence of the oxygen in this form has only a slight effect on the mechanical and electrical properties of the copper. It can, however, give rise to
porosity and intergranular cracks or fissures if the copper is heated in a reducing atmosphere, as can happen during welding or brazing. This is a
result of the reaction of the cuprous oxide particles with hydrogen and is known as ‘hydrogen embrittlement’. Provided a reducing atmosphere is
avoided, good welds and brazes can be readily achieved.
1.2.3.2 Oxygen-free High Conductivity Copper, CW008A (C103)
In view of the above remarks, if welding and brazing operations under reducing conditions are unavoidable, it is necessary to use a different (and
more expensive) grade of high conductivity copper which is specially produced for this purpose. This type of copper, known as ‘oxygen-free high
conductivity copper’, is normally produced by melting and casting under a protective atmosphere. To obtain the high conductivity required it is
necessary to select the best raw materials. The result is a high purity product containing 99.95% copper. This enables a conductivity of 100% IACS
to be specified even in the absence of the scavenging oxygen.
1.2.4 Available Forms
HC copper conductors are obtainable in bar, strip, rod or tube form. For busbar applications, the most common forms supplied are bar, rod or
tube and these are normally supplied in the hard condition. In this condition they offer greater stiffness, strength and hardness and have a better
surface finish. Because of the practical difficulty of straightening uncoiled hard material, it is normally supplied in straight lengths, coiled material
being limited to the smaller sizes.
The maximum length of material available with the advent of continuous casting methods is dependent on the capability of the supplier’s plant.

2.0 Current-Carrying Capacity of Busbars
2.1 Design Philosophy
The current-carrying capacity of a busbar is limited by the maximum acceptable working temperature of the system, taking into account the
properties of the conductor material, the materials used for mounting the bars and the limitations of any cables (including their insulation) or
devices connected to the bars.
There are two design limits; the maximum permitted temperature rise, as defined by switchgear standards, and the maximum temperature rise
consistent with lowest lifetime costs - in the vast majority of cases, the maximum temperature dictated by economic considerations will be rather
lower than that permitted by standards.
National and international standards, such as British Standard BS 159 and American Standard ANSI C37.20, give maximum temperature rises
as well as maximum ambient temperatures. For example, BS 159:1992 stipulates a maximum temperature rise of 50°C above a 24 hour mean
ambient temperature of up to 35°C, and a peak ambient temperature of 40°C. Alternatively, ANSI C37.20 permits a temperature rise of 65°C above
a maximum ambient of 40°C, provided that silver-plated (or acceptable alternative) bolted terminations are used. If not, a temperature rise of 30°C
is allowed.
These upper temperature limits were chosen to limit the potential for surface oxidation of conductor materials and to reduce the mechanical
stress at joints due to cyclic temperature variations. In practice these limitations on temperature rise may be relaxed for copper busbars if suitable
insulation materials are used. A nominal rise of 60°C or more above an ambient of 40°C is allowed by EN 60439-1:1994 provided that suitable
precautions, such as plating, are taken. EN 60439-1:1994 states that the temperature rise of busbars and conductors is limited by the mechanical
strength of the busbar material, the effect on adjacent equipment, the permissible temperature rise of insulating materials in contact with the bars
and the effect on apparatus connected to the busbars. In practice, this last consideration limits the maximum working temperature to that of the
insulation of cables connected to the bar – typically 90°C or 70°C.
Running busbars at a high working temperature allows the size of the bar to be minimised, saving material and initial cost. However, there are
good reasons to design for a lower working temperature.
• A higher maximum temperature implies a wider variation in temperature during the load cycle, which causes varying stress on joints and
supports as the bars expand and contract. Eventually, this may lead to poor joint performance and decreased reliability.
• In some cases it is difficult to remove heat from busbar compartments, resulting in ancillary equipment operating in an elevated ambient
temperature.
• High temperatures indicate high energy losses. Lowering the temperature by increasing the size of the conductor reduces energy losses
and thereby reduces the cost of ownership over the whole lifetime of the installation. If the installation is designed for lowest lifetime
cost, the working temperature will be far below the limit set by standards and the system will be much more reliable.
2.2 Calculation of Maximum Current-Carrying Capacity
In engineering terms, the current rating for a busbar depends on the choice of working temperature. The bar is heated by the power dissipated in
it by the load current flowing through the resistance, and cooled by radiation to its surroundings and convection from its surfaces. At the working
temperature, the heat generation and loss mechanisms, which are highly temperature and shape dependent, are balanced.
2.2.1 Methods of Heat Loss
The heat generated in a busbar can only be dissipated in the following ways:
• Convection
• Radiation
• Conduction
In most cases, convection and radiation heat losses determine the current-carrying capacity of a busbar system. In a simple busbar, conduction
plays no part since there is no heat flow along a bar of uniform temperature. Conduction need only be taken into account where a known amount
of heat can flow into a heat sink outside the busbar system, or where adjacent parts of the system have differing cooling capacities. Conduction
may be important in panel enclosures.

These cooling mechanisms are highly non-linear; doubling the width of a bar does not double the convection loss of a bar. The proportion of heat
loss by convection and radiation depends on the conductor size, with the portion attributable to convection normally being greater for a small
conductor and less for larger conductors.
2.2.1.1 Convection – Natural Air Cooling
The heat dissipated per unit area by convection depends on the shape and size of the conductor and its temperature rise above ambient
temperature. This value is usually calculated for still air conditions but can be increased greatly if there is forced air-cooling. Where outdoor
busbar systems are concerned, calculations should always be treated as in still (i.e. without wind effect) air unless specific information is given to
the contrary.
The following formulae estimate the convection heat loss from a surface in W/m² in free air:
For vertical surfaces:
For horizontal surfaces:

For round tubes:

where:
θ is temperature rise, °C
L is height or width of surface, mm
d is diameter of tube, mm.
Figure 4 shows the heat loss from a vertical surface (Wv) for various temperature rises plotted against surface height.
800
Heat dissipation per square metre from a vertical surface (W/m2)
700
600
500
30 degree C rise
40 degree C rise
400 50 degree C rise
60 degree C rise
300
200
100
0 50 100 150 200 250
Surface height (mm)
Figure 4 – Heat dissipation by convection from a vertical surface for various temperature rises above ambient

Figure 5 indicates which formula should be used for various conductor geometries.
Wh
Wh Wh Wh
Wv Wv Wv Wv Wv Wv Wv Wv
Wh Wh Wh Wh
(a) (b)
Wc
Wh
Wv Wv
Wh
(c) (d)
Figure 5 - Convection loss from typical bar sections
Comparing diagrams (a) and (b) and assuming a similar cross-sectional area, it can be seen that the heat loss from arrangement (b) is much larger,
provided the gap between the bars is not less than the thickness of each bar.
2.2.1.2 Convection Heat Loss - Forced Air Cooling
If the air velocity over the busbar surface is less than 0.5 m/s, the above formulae for Wv, Wh and Wc apply. For higher air velocities the following
may be used:
where:
Wa is heat lost per unit length from bar, W/m
v is air velocity, m/s
A is surface area per unit length of bar, m²/m
θ is temperature rise, °C.

2.2.1.3 Radiation
The rate at which heat is radiated from a body to its surroundings is proportional to the difference between the fourth power of their absolute
temperatures and the relative emissivity between the body and its surroundings. Emissivity describes how well a material radiates heat with a
perfect radiator (a black body) having a value of unity and a perfectly reflecting surface a value of zero.
Since the amount of radiation depends on the temperature of the bar and its surroundings, bars in enclosed spaces may lose very little heat by
radiation. Note that radiation in a particular direction will be influenced by the temperature and condition of any target surface.
The relative emissivity is calculated as follows:
where:
e is relative emissivity
ε1 is absolute emissivity of body 1
ε2 is absolute emissivity of body 2
Typical emissivity values for copper busbars in various surface conditions are:
Bright metal 0.10

Partially oxidised 0.30
Heavily oxidised 0.70
Dull non-metallic paint 0.90
The rate of heat loss by radiation from a bar (W/m²) is given by:

where:
e is relative emissivity
T 1 is absolute temperature of body 1, K
T 2 is absolute temperature of body 2, K (i.e. ambient temperature of the surroundings)
350
300
250
Heat dissipation by radiation (W/m2)
200
150
100
50
0
20 30 40 50 60 70 80 90 100
Surface temperature above 30 degree ambient (C)
Figure 6 - Heat dissipation by radiation from a surface assuming relative emissivity of 0.5 and surroundings at 30ºC

Figure 6 shows the heat loss from a surface against surface temperature. The diagrams in Figure 7 define the effective surface areas for radiation
from conductors of common shapes and arrangements. Note that there is no heat loss by radiation from opposing busbar faces since the
temperatures are approximately equal.
Wr Wr Wr Wr
Wr Wr Wr Wr
Wr Wr Wr Wr
(a) (b)
Wr
Wr
Wr Wr
Wr
(c) (d)
Figure 7 - Radiation loss from typical bar sections
The ratio of heat dissipated by convection and radiation varies considerably according to the height of the surface and the temperature rise, with
radiation becoming less important for smaller bars and lower temperature rises. Figure 8 combines the data from Figure 4 and Figure 6 with the
radiation dissipation levels, which are independent of surface height, shown as horizontal bars on the right.
In some countries it is common practice to attempt to increase emissivity to improve the dissipation by radiation by treating the surface of
busbars, for example by painting them black. Since the natural emissivity of a copper bar that has been in use for even a short time will be above
0.5 and probably approaching 0.7, the benefit of increasing it to 0.9, although positive, will be small. On the negative side, the paint layer acts as
an insulator, reducing the efficiency of the convection process. In general, painting will give little increase and possibly a reduction in the current-
carrying capacity of a busbar for a given working temperature. Painting may be worthwhile for very wide bars (where convection is less effective)
operating at large temperature rises (where radiation is more effective).

800
700
Heat dissipation per square metre from a vertical surface (W/m2)
600
500 Convection 60 degree C rise

Radiation 60 degree C rise
Convection 50 degree C rise
400
300 Radiation 40 degree C rise
200
100
0
0 50 100 150 200 250 300 350
Surface height (mm)
Figure 8 - Convection and radiation (right) losses at various temperatures
The plots given so far have been in terms of power dissipated per unit area of surface; for engineering purposes it is the heat dissipation per unit
length which is of interest. Figure 9 and Figure 10 show the heat dissipation from the major surfaces of single and parallel vertically mounted bars.
The relatively small contribution of the horizontal surfaces is ignored in these plots. These plots may be useful in determining a starting point for
detailed calculation using the formulae previously given.
400
350
Total heat dissipation per metre length (Watt)
300
250 225 mm
200 mm
175 mm
200
150 mm
125 mm
150 100 mm
75 mm
50 mm
100
50
0
20 30 40 50 60 70 80
Temperature rise above 30 degree ambient (C)
Figure 9 - Total heat losses for a single bar of various heights against temperature rise

400
350
300
Heat dissipation by radiation (W/m2)
250 225 mm
200 mm
175 mm
200
150 mm
125 mm
150 100 mm
75 mm
50 mm
100
50
0
20 30 40 50 60 70 80
Surface temperature above 30 degree ambient (C)
Figure 10 - Total heat losses for each bar of a parallel pair of various heights against temperature rise
2.2.2 Heat Generated by a Conductor
The rate at which heat is generated per unit length of a conductor carrying a direct current is the product I²R watts, where I is the current flowing
in the conductor and R is the resistance per unit length. In the case of dc busbar systems, the value for the resistance can be calculated directly
from the resistivity of the copper or copper alloy at the expected working temperature. Where an ac busbar system is concerned, the resistance
is increased because current density is increased near the outer surface of the conductor and reduced in the middle; this is called skin effect.
Eddy currents induced by magnetic fields arising from currents in nearby conductors increase losses further; this is called proximity effect. The
calculation of these effects is discussed later; for the present, a correction factor, S, is used.
The power dissipated in the conductor is

where:
P is the power dissipated per unit length
I is the current in conductor
R0 is the dc resistance per unit length at the working temperature
S is the correction factor for shape and proximity.
The process of sizing a busbar is one of iteration. Starting from an arbitrary size and the desired working temperature, the heat power loss from
the surface of a one-metre section can be calculated. The electrical power loss for the one-metre section can also be calculated. If the electrical
power dissipated is higher than the heat dissipation, the bar is too small; the size should be increased and the calculations repeated until a close
match is obtained. Note that the value of resistivity used in the calculation must be corrected for the working temperature and the value of S (the
correction for shape and proximity factors) must be recalculated for each size. A very approximate starting point is to assume an average current
density of 2 A/mm² in still air and iterate either up or down.

2.2.2.1 Alternating Current Effects – the Factor S
The factor S, introduced above, is the product of the factors due to skin effect, Sk , and proximity factor, Sp . Accurate determination of skin and
proximity effects is complex and generally requires finite element analysis. In this publication results are presented as curves and, where possible,
‘portmanteau formulae’ in the form of polynomials or ratios of polynomials valid over a given range of an independent variable such as the scaling
factor, p, described below. Most of these formulae were found by curve fitting to data computed by finite element analysis. The portmanteau
formulae do not always reflect satisfactorily a physical explanation and are usually very inaccurate outside the given range. They are normally
accurate to within ±1% within the stated range unless otherwise noted.
The graphs are plotted from the computed data. To make them readable, scaling factors and quasi-logarithmic scales are used.
The resistance/frequency parameter scaling factor combines frequency, resistivity and size. For a busbar configuration of given shape or relative
proportions, the ac resistance and inductance may be expressed as a function of a ratio of frequency to dc resistance. Two forms of the scaling
factor are used in terms of either
(a) The parameter, p, defined as
where:
f is frequency in Hz
Rdc is dc resistance in µΩ/m
also
where:
A is the cross-sectional area
δ is the skin depth
The area A and the skin depth must be in comparable units. Thus if δ is in mm then A must be in mm2.
Sometimes is used as the scaling parameter.
The parameter p is chosen since it appears widely in the literature on the subject. The plots for different cross-sectional shapes give clearer
curves when using p rather than, say, p 2 .
For p ≤ 0.5 the resistance is little different from the dc resistance. Rac is proportional to p 4 for low values of p and so proportional to the
square of the frequency but, at higher values beyond p = 5, the resistance becomes linear with p, i.e. proportional to the square root of
frequency. If efficiency is a concern, then p will normally be substantially less than 5 and more likely to be in the range 0 to 2. Figure 11
shows how p varies with cross-sectional area at a selection of frequencies for copper at 80˚C.

10.0
Parameter, p
1.0
50Hz
60Hz
400Hz
0.1
10 100 1000 10000
Cross-sectional area (mm2)
Figure 11 - Parameter p versus cross-sectional area in mm2 for typical copper at 80˚C
In some of the plots the scaling factor may be expressed simply as or in terms of γ.
(b) The ratio γ, defined as
This can often give a physical idea of the effect of size of a conductor as, for example, for round bars where and d is the bar diameter. For

low losses γ≤ 1 is desirable.
For a given busbar geometry, the values of γ and p are proportional to each other.
Quasi logarithmic scales: The resistance ratio or shape factor, S = Rac / Rdc , approaches unity at low frequencies. In some graphs the logarithm
of (S–1) is plotted against the independent variable such as scaling parameter, p . This method shows the approach to unity in more detail.
However, the values on the S scale are labelled as the actual values of S and not of (S–1). Such plots typically go down to S = 1.01 as covering a
useful range.
Busbar configuration: The choice of busbar configuration will depend on many factors. Round bars, tubes, strips or channels are common.
Factors to be taken into consideration include frequency and the presence of harmonics, cost of materials, degree of compactness sought, cooling
arrangements, magnetic mechanical forces and cost of support structures, inductance and, for longer bars, capacitance.
From Figure 11 it will be seen that, for p < 2, the cross-sectional area of bars is limited to 400 mm2 . At a typical current of 2 A mm-2 this is a
current of 800 A. The effective ac resistance may be reduced by:
1. Using two bars per phase separated by a suitable distance along their length. This possibility is discussed for some shapes in later
sections.
2. Using multiple bars for a phase and transposing them along the length of the bar
3. Interleaving phases but then the design becomes more detailed.

2.2.2.1.1 Skin Effect and Skin Depth
Current flowing in the inner parts of a conductor produces a magnetic field inside the conductor that circulates around the axis of the conductor.
This alternating magnetic field induces an electric field that drives currents in such a way that the current in the centre of the conductor is reduced
while the current in the outer parts is increased. These induced recirculating currents are referred to as eddy currents. The overall result is that
the current density is highest near the surface and falls off towards the interior.
A measure of the extent of the fall off (and so of the effect of the ac nature) is the skin depth, sometimes called the depth of penetration, and
usually given the symbol d.
where:
ρ is resistivity of conductor material, Ω m
µ0 is permeability of free space, 4π 10-7 H/m
f is frequency, Hz
δ is skin depth, m
Or, in alternative units,
where:
ρ is resistivity of conductor material, nΩ m
f is frequency, Hz
δ is skin depth, mm
Hence, for copper at 20ºC (ρ = 17 nΩ m),
≈ 9.27 mm at 50Hz
≈ 8.46 mm at 60 Hz
≈ 3.28 mm at 400 Hz
About 85% of the current is carried in a layer of thickness equal to the skin depth in conductors that are much thicker than the skin depth. In such
thick conductors the resistance is as if the current flowed with uniform current density in a layer of thickness equal to the skin depth.
Figure 12 shows how the resistivity of a typical grade of copper varies with temperature and Figure 13 shows how the skin depth varies with
temperature at typical power frequencies.

23
22
21
20
Resistivity (nΩ m)
19
18
17
16
15
0 20 40 60 80 100
Temperature (˚C)
Figure 12 - Resistivity of typical HC copper (101.5% IACS) as a function of temperature. Note 10nΩ m = 1μΩ cm
12
10
8
50 Hz
60 Hz
Skin depth (mm)
400 Hz
6
0
0 20 40 60 80 100
Temperature °C
Figure 13 - Skin depth of typical HC copper (101.5% IACS) at 50 Hz, 60 Hz and 400 Hz as a function of temperature

If the dimensions of the bar are larger than about half the skin depth, then ac effects will be important. In such a case, the extra thickness may
increase the losses rather than reduce them, but only slightly.
A useful measure of the ac effect in determining the ac resistance of a bar is the Shape Factor, S, defined as
where:
Rdc is the dc resistance
Rac is the ac resistance
thus,
where:
Rdc is resistance, Ω/m
A is cross-section, m2
ρ is resistivity, Ω m
or,
where:
Rdc is resistance, µΩ/m or mΩ/km
A is cross-section, mm2
ρ is resistivity, nΩ m
and
Although computing the bar resistance is straightforward, Figure 14 is given either as a check or for approximate estimation.
1000.0
100.0
20°C
Resistance (nΩ/m)
80°C
10.0
1.0
0.1
10 100 1000 10000 1000000
Cross-sectional area (mm2)
Figure 14 - dc resistance of typical HC Copper (101.5% IACS) versus area at 20ºC and 80ºC
One would hope in any design that the shape factor, S, would be not much more than unity.
The determination of the shape factor is a principal feature of this section.

The power, P, dissipated as heat in a bar carrying current, I, is then
where:
P is the power dissipated per metre length, W/m
Rdc is the resistance per metre length, W/m
S is shape factor
ρ is resistivity, W m
A is cross-sectional area, m2
or
where:
P is the power dissipated per metre length, mW/m
Rdc is the resistance per metre length, µΩ/m; mΩ/km
S is shape factor
ρ is resistivity, nW m
A is cross-sectional area, mm2.
The precise calculation of eddy current in conductors of general shape requires finite difference or other numerical computational procedures.
These procedures require extensive programme writing or access to one of the commercially available programmes and practice in its use.
However, for some common configurations there are exact (if complicated) formulae, while for others there are simpler approximations and
graphical methods. These are given in '2A Shape and Proximity Factors for Typical Configurations'.
2.2.2.1.2 Proximity Factor, Sp
With conductors in parallel there are also eddy current mechanisms and further current redistribution in a conductor caused by magnetic fields
arising from currents in neighbouring conductors. This is the proximity effect.
In the case of a single bar, the eddy current mechanism tends to drive current to flow in outer regions of the bar.
In the case of adjacent conductors carrying anti-parallel currents, the effect is to draw current flow towards the facing surfaces of the bars. For
round bars or tubes which are carrying anti-parallel currents there is an increase in the current density at those parts of the bar that face each
other, so losses are higher. For closely spaced parallel strips with their long sides facing each other, there is still a concentration of current near the
ends of the strips, but overall the current is more uniformly distributed across the width and overall losses can be reduced.
The proximity factor is sometimes expressed as Sp so that the effective resistance, Rac , as given by:
In section 2A, which gives charts and equations for common busbar shapes and arrangements, the shape and proximity factors are given either
separately as S and Sp or combined together as Rac / Rdc .
2.3 Conclusion
The design of a busbar system for a particular duty is quite a complex issue. It requires an iterative approach to arrive at a size and disposition that
delivers the energy efficiency levels appropriate to the duty factor of the application while ensuring good reliability and safety.

2A Shape and Proximity Factors for Typical Configurations
2A.1 Skin and Proximity Factors for Common Busbar Shapes
2A.1.1 Single Solid Rods
2A.1.1.1 Shape Factor for Single Solid Rods
For a rod of radius a calculate g:
where d is skin depth.
Substituting for d,
where:
f is frequency, Hz
Rdc is resistance in µΩ/m
then,
Sp can be calculated from the following approximate formulae:
for g < 1.98:
Error is <0.02% if g < 1.68; <0.1% if g < 1.98
for g > 3.9:
Error is <0.8% if g > 3.9, <0.2% if g > 5
In the intermediate range 1.6 < g < 3.9 the following formula is within 0.2%:
Errors lie between -0.2 % and + 0.15 %.

Computations using the formulae may be checked by comparison with the graphs in Figure 15. These graphs might also be used directly for
approximate working.
Low values of ϒ High values of ϒ
5.0
4.5
dc
dc
1.6
Shape Factor, S = Rac / R

4.0
3.5
1.4 3.0
2.5
1.2 2.0
1.5
1.0
0 0.5 1 1.5 2 2.5 3 4 6 8 10
Radius, a Radius, a
ϒ= ϒ=
Skin depth, δ Skin depth, δ
Low values of ϒ - log scale Low values of ϒ
1.50 1.10
1.08
dc
1.20
dc
1.10 1.06
1.05 1.04
1.02
1.02
1.01 1.00
0.5 1.0 1.5 2.0 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Radius, a Radius, a
ϒ= ϒ=
Skin depth, δ Skin depth, δ
Figure 15 - Plots of shape factor versus γ. In the upper left-hand figure, for low values of γ, the dotted line shows
the exact formula and the solid line the given approximation. The difference is scarcely discernable.
2A.1.1.2 Proximity Factor for Single Solid Rods
For the round bar calculate h,
where:
s is the spacing of the centres
a is the bar radius.

Then the proximity factor, for h > 2
h > 2 means that the minimum space between the opposing curved faces is greater than the bar diameter.
The factor A in the above equation is a function of and approximately for g < 2
This relationship is plotted in Figure 16.
0.6
0.5
Factor A for round bars
0.4
0.3
0.2
0.1
0
0 1 2
Radius, a
ϒ=
Skin depth, δ
Figure 16 - Factor A for round bars as a function of g

Each curve is for the
1.50 given value of
bar radius, a
ϒ=
1.30 skin depth, δ
2.0
1.8
1.6
1.10 1.4 1.3
Proximity factor, Sp
1.2
1.05
1.1
1.0
1.03
0.7 0.8 0.9
1.01
1.005
1.0 1.5 2.0 2.5 3.0 3.5 4.0
centre spacing, s
η=
bar diameter, 2a
Figure 17 - Proximity factor, SP, for single phase systems with parallel round bars
For three phase systems calculate µ:
where:
Then for a triangular arrangement of three phase busbars,
And for a flat arrangement the average proximity factor is
The factors are plotted in Figure 18 and Figure 19.

3 phase flat layout
s
1.50 a
2.0
1.30
Each curve is for the
given value of
1.8
bar radius, a
ϒ=
skin depth, δ
1.6
1.4 1.2 1.0
1.3 1.1
1.10
1.05
1.03
0.8 0.9
0.7
1.01
1.005
1.0 1.5 2.0 2.5 3.0 3.5 4.0
centre spacing, s
η=
bar diameter, 2a
Figure 18 – Mean proximity factors for flat arrangement of round bars carrying balanced three phase currents

3 phase delta layout
s
a
1.50
2.0
1.30 Each curve is for the

given value of
1.8
bar radius, a
ϒ=
skin depth, δ
1.6
1.4 1.1
1.3 1.2 1.0
1.10
1.05
1.03
0.7
0.8
0.9
1.01
1.005
1.0 1.5 2.0 2.5 3.0 3.5 4.0
centre spacing, s
η=
bar diameter, 2a
Figure 19 - Mean proximity factors for delta arrangement of round bars carrying balanced three phase currents

2A.1.2 Single Tubes
2A.1.2.1 Shape Factors for Single Tubes
Calculate parameters g and b as:

and
where:
a is radius
δ is skin depth
t is wall thickness
The units for a, δ and t must be the same.
The approximate formula for S is
where A and B are given by:
for g <1.6:
for 1.59 < g < 3.9:
for g > 3.9:
Figures 20 to 24 show plots of shape factor for various tube sizes.

β= Wall thickness, t
Tube outer radius, a
0.4 0.6 0.7
0.2 0.8
0 0.9
1.30
1.25
Shape factor, S = Rac/ Rdc
1.20
1.15
1.10
1.05
0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Wall thickness, t
g=
Skin depth, δ
Figure 20 - Shape factor for tubes
0.4 0.6 0.7
0.2 0.8
0 0.9
1.10
1.08
Shape factor, S = Rac / Rdc
1.06
1.04
1.02
0.2 0.4 0.6 0.8 1.0 1.2 1.4

Wall thickness, t
g=
Skin depth, δ
Figure 21 - Shape factor for tubes with low values of shape factor

Wall thickness, t
= 0.99 0.8 0.7 0.6 0.5 0.4 0.30
0.25
0.20
0.15
2.0
0.125
1.8
0.10
1.6
1.4 0.075
1.2 0.05
1.0
0 1 2 3 4 5 6 7
√[ f, Hz
Rdc , µΩ/m ]
Figure 22 - Shape factor for tubes
Wall thickness, t
= 0.99 0.8 0.7 0.6 0.5 0.4 0.30
0.20
0.15
2.0
0.10
1.5
0.075
1.2 0.05
1.1
1.05
1.02
1.01
0 1 2 3 4 5 6 7
√[ f, Hz
Rdc, µΩ/m ]
Figure 23 - The shape factor computed from the Bessel function formula using √(f/rdc) as the frequency parameter

Wall thickness, t
= 0.05
0.2
0.3
2.0
0.4
1.5 0.5
0.6
0.7
1.2
0.8
1.1 0.9
0.99
1.05
1.02
1.01
0.5 1.0 1.5 2.0
Wall thickness, t
g=
Skin depth, δ
Figure 24 - Shape factor as a function of the ratio,
2A.1.2.2 Proximity Factors for Tubes
As for the round bar calculate h,

where:
s is the spacing of the centres
a is the bar radius.
For moderate to large spacings, i.e. for h > 2,
with the factor A now given by

where:

NOTE: The factor a(x) is not the same as the radius a. ‘a(x)’ is a term normally adopted in the literature and a and a(x) are unlikely to be confused
once the difference is established.
For the single phase case we have Figure 25 which plots the factor A against β = t/a. This factor A is then used in Figure 26 to find Sp, the
proximity factor for values of η=s/2a=s/d.
1.0
Wall thickness, t
g=
Skin depth, δ
0.8
0.6
Factor A
0.4
0.2 g=1.0
0.9
0.8
0 g = 0.1 0.2 0.3 0.4 0.5 0.6 0.7
0 0.2 0.4 0.6 0.8 1

Figure 25 - Factor A versus β = t/a for values of g = t / δ
1.20
η=2
1.10
Proximity factor Sp
η=3
1.05
η=4
1.02 η=5
1.01 η=6
0 0.2 0.4 0.6 0.8 1.0

Factor A
Figure 26 - Proximity factor for single phase tubes as function of the factor A for values of η=s/2a=s/d
These graphs may be combined into a joint chart as in Figure 27.

1.2
η=2
1.1
η=3
1.05
Proximity factor Sp
η=4
1.02 η=5
1.01 η=6
1.0
g=0.7 g=0.8
0.6 0.9
Wall thickness, t
0.8 g=
Skin depth, δ
1.0
0.5
0.6
0.4
β
0.4
0.3
0.2
0.2
g=0.1
0 0.2 0.4 0.6 0.8 1.0

Factor A
Figure 27 - Proximity factor for single phase tubes. β = t/a , g = t/δ and η = s/2a = s/d with t = wall thickness,
a = radius, d = diameter, δ = skin depth
The chart is used by starting with the value of b on the vertical axis of the lower plot, drawing a horizontal line to the curve for the value of g to
obtain the value of the factor A. Next, draw a line vertically to the upper plot to intersect the line for the value of η. Finally move horizontally to
the left to read of the proximity coefficient, Sp , from the vertical axis of the upper plot. In the illustrated case, b = 0.45, g ≈ 0.8, A ≈ 0.42, h = 3
and Sp = 1.025.

For low values of the parameter A, and when g is small, the charts just described are not very accurate. Thus Figure 28 and Figure 29 may be used
in these circumstances.
β = 0.01
0.05
1.0
0.1
0.2
0.8
0.3
0.6
A
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
g
Figure 28 - Factor A for proximity loss factor as a function of g = t/δ for various values of β = t/a
The formula just derived applies to a single phase system with two parallel conductors. For three phase systems calculate µ:
where:
Then for a triangular arrangement of three phase busbars
And for a flat arrangement the average proximity factor is

For small spacings, i.e. for h < 2, the approximate formulae are rather bulky but the chart given in Figure 29 can be used.
η=1
1.111
2.00
1.125
1.428
1.66
1.10
2
Sp
1.01
1.001
0.5 0.7 1 1.5 2 3 5 7 10

x
Figure 29 - Proximity factor, Sp , for round bars at various spacings designated by h as defined in the figure
The parameter x is given by:
where:
d is skin depth
a is tube radius
and
where:
t is the tube wall thickness.
Also, in SI units,

with
Rdc in µΩ/m and
f in Hz

2A.1.3 Single Rectangular Sections
We consider here first single bars of rectangular cross-section which might be called strips or even straps. In subsections we consider single
phases divided into two strips, separated from each other and carrying currents in the same direction. This leads to reduced losses. Then we
consider a pair of bars carrying anti-parallel currents to form a single phase circuit. Finally, we consider three phase arrangements of bars carrying
balanced positive phase sequence currents.
A portmanteau formula is given for a single bar, but not for the other arrangements, since none are available. The reason for the dearth is partly
because it requires two or more parameters to describe the shape. Thus, resort must be made to graphs or detailed computation using a fairly
complicated programme.
2A.1.3.1 Shape Factor for Rectangular Bars
There are no analytic formulae for this case. Resort must be made to numerical methods such as finite elements, finite differences or current
simulation, i.e. particle elements.
Recollect that the parameter p is defined as:
where:
f is frequency in Hz
Rdc is dc resistivity in µΩ/m
and
where:
A is the cross-sectional area
d is the skin depth
The area A and the skin depth must be in comparable units. Thus if d is in mm then A must be in mm2.
The following plots are of shape factor for strips of different cross-sectional proportions.

2.2
1
2
2.0
4
1.8
5
b/a
Rac / Rdc
8
1.6
16
32
1.4
Infinity
2a
1.2
2b
1.0
0.5 1.0 1.5 2.0 2.5 3.0
√[ f, Hz
Rdc , µΩ/m ]
Figure 30 - Rac / Rdc as a function of the parameter √f/R dc
1
2
1.40
3
1.35 4
5
1.30
7
1.25 10 b/a
Rac/ Rdc
16
1.20
32
64
1.15
Infinity
1.10
2b
1.05
2a
1.00
0.5 1.0 1.5 2.0
√[ f, Hz
Rdc , µΩ/m ]
Figure 31 - Rac / Rdc as a function of the parameter √f/R dc . The range of the parameter is narrower than in the previous figure

At low values of p the curves for the different aspects ratios cross, as illustrated in Figure 32.
b/a = 1 3 5 16 32 Inf
1.20
1.15
1.10
Rac / Rdc
1.07
1.05
1.03
1.02
1.0 1.2 1.4 1.6 1.8
√[ f, Hz
Rdc , µΩ/m ]
Figure 32 - Shape factor, Rac / Rdc , as a function of the parameter √f/R dc .
Note the logarithmic scale for the shape factor and that the lines for constant b/a cross.
b/a = 1 2 3 4 5 7 10 16
2.00
24
1.50
32
1.20 64
Rac / Rdc
1.10
1.05
1.02
1.01
0 5 10 15 20 25 30
Long side, 2b
Skin depth, δ
Figure 33 - Shape factor versus 2b/d for various values of b/a
For values of √(f/Rdc) < 1.25 the thin sectioned bar has a slightly higher shape factor than the square conductor, whereas for higher values a thin
shape gives a lower shape factor. For copper at 20˚C with r = 17 nΩ m the transitional cross-sectional area is about 500 mm2; i.e. in old units,
about 1 sq in. One may, however, prefer a thin strip of say 6 mm by 80 mm, which is more readily cooled because it has a larger vertical flat face.
The increase in losses is of the order of one percent.

2A.1.3.2 Approximate Portmanteau Formula for Single Bars
An approximate formula which was made by curve fitting is:
with
f in Hz and Rdc in µΩ/m
with A = b/a
2A.1.4 Parallel Bars
2A.1.4.1 Proximity Factor - Anti-parallel Currents
The currents in adjacent bars induce eddy currents in neighbours. For thin wide strips placed with their long sides facing each other, there is
usually a reduction in losses as the current becomes distributed more uniformly across the width of the strip. For lower values of b/a the losses
increase. This is illustrated in the next graphs.
If the strips are too close together, the flow of coolant between them may be impaired, the temperature may rise, the resistance increase and so the
ohmic losses become larger.
Figure 34 (a-e) shows the shape factor, Rac / Rdc , as a function of the spacing expressed as s/a for various values of √f/R dc . As before, f is in Hz
and Rdc in μΩ/m.
2a
3.5
b 2b
3.0 =1
a
Shape Factor, S = Rac / Rdc
2.5 s
2.0 3.0
2.5 √(f / Rdc )
1.5 2.0
1.5
1.0 1.0
0 1 2 3 4 5
s/a
(a) b/a=1

2a
3.0
b 2b
=2
2.5 a
s
3.0
2.0
2.5
√(f / Rdc )
1.5 2.0
1.5
1.0 1.0
0 1 2 3 4 5
s/a
(b) b/a=2
2a
2.2 b
= 4
a 2b
2.0
1.8 s
3.0
1.6
2.5
1.4 √(f / Rdc )
2.0
1.2 1.5
1.0
1.0
0 1 2 3 4 5
s/a
(c) b/a=4
2a
1.7
b
=8
1.6 a 2b
3.0
1.5
s
1.4 2.5
1.3
2.0 √(f / Rdc )
1.2
1.5
1.1
1.0
1.0
0 1 2 3 4 5
s/a
(d) b/a=8

3.0 2a
b
1.4 = 16
a 2b
2.5
1.3
s
2.0
1.2 √(f / Rdc )
1.5
1.1
1.0
1.0
0 1 2 3 4 5
s/a
(e) b/a=16
Figure 34 - Plots of the shape ratio versus the ratio for various values of for anti-parallel currents
2A.1.4.2 Proximity Factor - Parallel Currents
If a single busbar is replaced by two busbars of the same total cross-section, the losses may be reduced for one or more of three reasons:
1. The exposed surface area will be larger, the cooling more effective, the temperature lower, the resistivity lower and so losses lower.
2. The shape factor Rac / Rdc will be reduced since the ratio for each bar will be lower. The reason for this is that the factor √f/R dc will be
less for each bar than for the single bar, since Rdc of each bar will be twice that of a single bar. A lower value of √f/R dc means a lower
shape factor.
3. The two new bars may be of improved shape so as to give rise to an even lower shape factor than the original design of single bar.
In considering how to design a divided bar, one may perhaps think of an original single bar and split it into two equal halves – for this reason, the
width of the bars in this section is a rather than 2a, as used elsewhere. This is the idea behind the next plots.
2.2 a
2.0 b/a = 1 2b
1.8
s
3.25
1.6 3.0
2.75
2.5
1.4
2.25 √(f / Rdc )
2.0
1.2 1.75
1.5
1.0 1.25
1.0
0 1 2 3 4 5
s/a
(a) b/a=1

a
2.0 2b
b/a = 2
1.8 s
3.25
1.6 3.0
2.75
2.5
1.4 2.25 √(f / Rdc )
2.0
1.75
1.2
1.5
1.25
1.0 1.0
0 2 4 6 8 10
s/a
(b) b/a=2
1.8
2b
b/a = 4
1.6 s
3.25
3.0
1.4
2.75
2.5
2.25 √(f / Rdc )
1.2 2.0
1.75
1.5
1.25
1.0
1.0
0 2 4 6 8 10
s/a
(c) b/a=4
a
1.6
b/a = 8 2b
1.5
s
1.4 3.25
3.0
1.3 2.75
2.5
√(f / Rdc )
1.2 2.25
2.0
1.1 1.75
1.5
1.25
1.0
1.0
0 2 4 6 8 10
s/a
(d) b/a=8

a
1.6
1.5 b/a = 16 2b
1.4 s
3.25
3.0
1.3
2.75
2.5 √(f / Rdc )
1.2 2.25
2.0
1.75
1.1 1.5
1.25
1.0 1.0
0 5 10 15 20 25
s/a
(e) b/a=16
Figure 35 - Shape factor S = as a function of the separation s for various values of with Rdc in μΩ/m and f in Hz
2A.1.5 Three-phase Configurations
In the following plots, the resistance to balanced three-phase positive sequence currents is shown. There will also be zero sequence and negative
sequence voltages but these produce no losses and are anyway smaller than the positive phase sequence voltages.
2A.1.5.1 Linear Plots
The shape factor Rac /Rdc is shown in Figure 36 as a function of the parameter p for various aspect ratios of the bar layout.
s 0.05
√(4ab) 5.0 1.6 0.8
1.5
0.1
s
0.2
0.4 √(4ab)
1.4
1.6
0.8
1.3
Rac / Rdc
2a
s
1.2
2b
1.1
1.0
0 1 2 3 4 5
Parameter, p
(a) b/a=16

2.2 s
0.05
√(4ab)
0.1
2.0
0.2
0.4
1.8
0.8
Rac / Rdc
1.6 1.6
5.0 2a
s
1.4
1.2 2b
1.0
0 1 2 3 4 5
Parameter, p
(b) b/a = 8
3.0 0.05
0.1
0.2
2.5 s
0.4 √(4ab)
Rac / Rdc
0.8
2.0 2a
1.6 s
5.0
1.5
2b
1.0
0 1 2 3 4 5
Parameter, p
(c) b/a = 4
0.05
3.5 0.1
s
0.2
3.0 √(4ab)
0.4
2.5 0.8
Rac / Rdc
1.6
2a
2.0 5.0 s
1.5
2b
1.0
0 1 2 3 4 5
(d) b/a = 2 Parameter, p

s 0.1 0.2
√(4ab) 0.05 0.4
3.0
0.8
2.5 1.6
5.0
Rac / Rdc
2.0
2a
s
1.5
2b
1.0
0 1 2 3 4 5
Parameter, p
(e) b/a = 1
2a s
2b 0.05
0.1
4.0 0.2
s
0.4 √(4ab)
3.0
0.8
Rac / Rdc
1.6
2.0 5.0
1.0
0 1 2 3 4 5
Parameter, p
(f) a/b = 2
2a s
2b
4.5 0.05
0.1
4.0
0.2
3.5 s
0.4
√(4ab)
Rac / Rdc
3.0
0.8
2.5
1.6
2.0 5.0
1.5
1.0
0 1 2 3 4 5
(g) a/b = 4 Parameter, p

2a s
2b
0.05
0.1
4.0
0.2
s
3.0 0.4 √(4ab)
Rac / Rdc
0.8
2.0 1.6
5.0
1.0
0 1 2 3 4 5
Parameter, p
(h) a/b = 8
2a s
2b
0.05
4.0 0.1
0.2
s
3.0 0.4 √(4ab)
Rac / Rdc
0.8
2.0 1.6
5.0
1.0
0 1 2 3 4 5
Parameter, p
(i) a/b = 16
Figure 36 – Linear plots of Rac / Rdc versus p where p ≈ 1·5853 with Rdc in μΩ/m and f in Hz

2A.1.5.2 Logarithmic Plots
For lower losses, using a pseudo-logarithmic scale for Rac /Rdc gives clearer plots, as shown in Figure 37.
5.0 1.6
1.10 s
0.8
√(4ab)
1.07 0.4
0.2
1.05 0.1
0.05
Rac / Rdc
0.05
1.03 0.2 2a
s
1.02
1.015 2b
1.01
0 0.5 1.0 1.5 2.0 2.5
Parameter, p
(a) a/b = 1/16
0.05 5.0
1.15 1.6
0.8 s
1.10
0.4 √(4ab)
1.07 0.2
1.05 0.1
Rac / Rdc
1.04
2a
1.03 s
1.02
1.015 2b
1.01
0 0.5 1.0 1.5 2.0 2.5
Parameter, p
(b) a/b = 1/8
0.05
1.50
0.1
0.2 s
1.20 0.4 √(4ab)
0.8
1.10
Rac / Rdc
1.6
5.0 2a
1.05 s
1.02 2b
1.01
0 0.5 1.0 1.5 2.0 2.5
Parameter, p
(c) a/b = 1/4

0.05
0.1
1.50 0.2 s
0.4 √(4ab)
0.8
1.20 1.6
5.0
1.10
Rac / Rdc
2a
s
1.05
1.02 2b
1.01
0 0.5 1.0 1.5 2.0 2.5
Parameter, p
(d) a/b = 1/2
2a
s
s
2b 0.05 0.1 0.2 √(4ab)
0.4
1.5
0.8
1.6
1.2 5.0
1.1
Rac / Rdc
1.05
1.02
1.01
0 0.5 1.0 1.5 2.0 2.5
Parameter, p
(e) a/b = 1
2a s
2b
0.2 s
0.05 0.1 0.4
√(4ab)
1.50 0.8
1.6
1.20 5.0
1.10
Rac / Rdc
1.05
1.02
1.01
0 0.5 1.0 1.5 2.0 2.5
Parameter, p
(f) a/b = 2

2a s 0.05
0.1
2b 0.2
s
1.50 0.4
√(4ab)
0.8
1.6
1.20 5.0
1.10
Rac / Rdc
1.05
1.02
1.01
0 0.5 1.0 1.5 2.0 2.5
Parameter, p
(g) a/b = 4
2a s
2b
0.05 0.1 0.2
s
0.4
1.50 √(4ab)
0.8
1.6
1.20
5.0
1.10
Rac / Rdc
1.05
1.02
1.01
0 0.5 1.0 1.5 2.0 2.5
Parameter, p
(h) a/b = 8
2a s
2b s
√(4ab)
0.05 0.1 0.2 0.4
0.8
1.5
1.6
5.0
1.2
1.1
Rac / Rdc
1.05
1.03
1.02
1.01
0 0.5 1.0 1.5 2.0 2.5
Parameter, p
(i) a/b = 16
Figure 37 – Quasi logarithmic plots of Rac / Rdc versus p where p ≈ 1·5853 with Rdc in μΩ/m and f in Hz

3.0 Life Cycle Costing
David Chapman
3.1 Introduction
Life Cycle Costing (LCC) is a technique used to estimate the total costs of a project, installation or facility during the whole of its economic life,
taking into account all costs and benefits. The use of LCC enables organisations to maximise the output value they generate for the available input
resources. In the case of private enterprise, this means increasing profit – maximising the bottom line. In the case of public services, increased
value means improving the quality and quantity of the services provided commensurate with the level of funding available.
In the context of electrical installation design, LCC is used to inform design choices. Usually, this means that all the costs, including investment
cost, running costs and end-of-life costs, are compared for design options. Since the objective is to minimise the lifetime cost rather than to
determine it, everything that remains the same for each design option can be ignored, simplifying the analysis and allowing greater focus on the
differences.
This is a departure from the normal approach, in which the installation is designed to minimise initial cost with little regard to the future; an
example might be the selection of a damper controlled ventilation system (which would be cheap to buy but very expensive to run) rather than,
say, a variable speed fan, which would be more expensive to buy, but much cheaper to run.
In established facilities, the existing structures, operating procedures and technology already in use will determine the life cycle costs of many
processes but, when new production lines or services are introduced or updated, or when additional facilities are constructed, there is an
opportunity to specifically design the infrastructure using life cycle costing to achieve the lowest lifetime cost.
Often the focus on lifetime cost reveals that, because of the high cost of energy, the running costs are by far the greatest part of the total cost,
meaning that the lowest life cycle cost design is one in which energy consumption has been carefully considered.
The advantage of LCC is obvious, so why is it so often ignored?
The first reason is that the benefit of LCC – primarily lower annual running costs – are an advantage to the operator of the facility, while the quick
profit gained by installing the lowest price equipment is a benefit to the M&E contractor. Solving this issue requires the client – the building
operator – to specify requirements to be met by the M&E contractor concerning operating costs.
Another reason is the perceived difficulty of assessing financial parameters, such as the cost of energy, discount rates, etc. for some time into the
future. However, when LCC is used as a comparative tool to assess the relative costs of different design approaches, different equipment choices
or different standards for installation practice, precise information is not needed – the objective is to determine, with a good degree of certainty,
the lowest cost, rather than the exact cost. Performing a sensitivity analysis, by estimating the LCC for each option using a few values representing
the predicted maxima and minima, improves confidence in the result.
A third perceived problem is that of dealing with risk in the project. Business decisions are always subject to change, perhaps due to changes in
market conditions or technological advances leading to the installation becoming obsolete, so that it is sometimes difficult to define ‘lifetime’. This
concern, coupled with occasionally volatile economic conditions, led to the adoption of a rather conservative view of investment which required
short payback times. This encouraged a ‘lowest investment cost’ approach, arguing that the smaller the sum invested, the easier it should be to
recover and the lower the risk. Expressing all the costs in terms of their net present value (NPV) discounts future financial sums and reflects the
possibility that the project may not be viable for as long as expected.
3.1.1 The Future Value of Money – Net Present Value
A sum of money which is to be paid or received some time in the future does not have the same value as the same sum of money today. This
is because money available today could be invested to generate a profit or used to pay an outstanding debt but, since money is not available,
it would need to be borrowed. A sum promised in the future cannot be committed and so has less value today. This is very relevant to the
assessment of life cycle cost since costs and benefits that accrue in the later years would be given undue weight if they were included at full value.
The problem in assessing the lifetime cost is that the setting up costs occur at the beginning of the project while the costs of energy losses occur
over a long period of time, so it is necessary to add together costs which arise, and have to be paid for, at different times. This is where the time
value of money must be included in the calculation.
The simple addition of all future energy costs to the initial costs would be to imagine that sufficient money is to be put aside at the start of the
project to pay all future bills or, put another way, that all future bills for energy costs would be paid in full at the start of the project.

This would not be an economical way to make provision for future payments and would unnecessarily inflate the total cost of the project. A more
practical way would be to consider what the energy costs would be worth if the money were to be set aside to earn interest until the time came to
pay each bill. It is evident that this procedure would require smaller amounts to be set aside.
It is common practice to express all costs and benefits in terms of ‘Net Present Value’ (NPV) at the start of the project using an appropriate
‘discount rate’. The discount rate depends to a large extent on the cost of financing the business which will be determined by, for example,
commercial interest rates and the cost of raising money from the markets, but it will also include an allowance for risk.
Allowing for risk can be complex. In a commercial building the risk may be low because the organisation is likely to continue to require the facility
even though the business focus may change. In other areas, such as advanced technology, the risk may be very high; the development of practical
electric vehicles requires many technological advances coupled with extensive infrastructure and behavioural changes which are not within the
control of the vehicle developer and may not happen without timely policy initiatives. Higher perceived risk implies a higher discount rate.
The discount factor, r, to be applied to sums for a particular year, is given by:

where:
n is the number of years since the start
i is the discount rate (%).
Table 8 below shows the annual discount factors for various discount rates over a period of 20 years.
Table 8 – Discount Factors for Various Discount Rates

Discount Rate (%)
Year 1 2 3 4 5 6 7 8 9 10
0 1 1 1 1 1 1 1 1 1 1
1 0.990 0.980 0.971 0.962 0.952 0.943 0.935 0.926 0.917 0.909
2 0.980 0.961 0.943 0.925 0.907 0.890 0.873 0.857 0.842 0.826
3 0.971 0.942 0.915 0.889 0.864 0.840 0.816 0.794 0.772 0.751
4 0.961 0.924 0.888 0.855 0.823 0.792 0.763 0.735 0.708 0.683
5 0.951 0.906 0.863 0.822 0.784 0.747 0.713 0.681 0.650 0.621
6 0.942 0.888 0.837 0.790 0.746 0.705 0.666 0.630 0.596 0.564
7 0.933 0.871 0.813 0.760 0.711 0.665 0.623 0.583 0.547 0.513
8 0.923 0.853 0.789 0.731 0.677 0.627 0.582 0.540 0.502 0.467
9 0.914 0.837 0.766 0.703 0.645 0.592 0.544 0.500 0.460 0.424
10 0.905 0.820 0.744 0.676 0.614 0.558 0.508 0.463 0.422 0.386
11 0.896 0.804 0.722 0.650 0.585 0.527 0.475 0.429 0.388 0.350
12 0.887 0.788 0.701 0.625 0.557 0.497 0.444 0.397 0.356 0.319
13 0.879 0.773 0.681 0.601 0.530 0.469 0.415 0.368 0.326 0.290
14 0.870 0.758 0.661 0.577 0.505 0.442 0.388 0.340 0.299 0.263
15 0.861 0.743 0.642 0.555 0.481 0.417 0.362 0.315 0.275 0.239
16 0.853 0.728 0.623 0.534 0.458 0.394 0.339 0.292 0.252 0.218
17 0.844 0.714 0.605 0.513 0.436 0.371 0.317 0.270 0.231 0.198
18 0.836 0.700 0.587 0.494 0.416 0.350 0.296 0.250 0.212 0.180
19 0.828 0.686 0.570 0.475 0.396 0.331 0.277 0.232 0.194 0.164

Energy prices are likely to increase over the lifetime of the project so the cost will increase each year by the factor:
where b is the expected average percentage rate of increase of energy prices or, in the absence of information, the rate of general inflation, over
the project lifetime.
It is also likely that the load carried by the busbar system will grow over the life of the installation and this must to be taken into account in two
ways.
Firstly, the cost of energy in later years must be modified accordingly. Although load changes will occur at discrete (and unknown) points in the
system lifetime, it is easier to assume that they occur at a uniform rate. The correction factor for each year is:
where a is the average percentage increase in load current over the project lifetime.
Secondly, the system must be designed, from a safety point of view, to cope with the final load current which is given by
where:
If is the load current at end of life
Ii is the initial load current
N is the circuit lifetime in years.
The overall correction factor, including discount rate, becomes:
The NPV of the lifetime energy cost is therefore
where:
N is the project lifetime in years
C is the first year cost
or
The term in parentheses in the equation above is a standard series so that

and

where Q, the discount ratio, is given by
Note that if r = 1, the equation for Q cannot be evaluated, but from the original series it can be seen that in that case Q = N.
3.1.1.1 Monthly Payments
Electricity accounts are usually settled each month, even though agreements with regard to prices and financial matters may be made annually.
In this case the following modifications to the derivation of the discounting factor Q may be appropriate. The difference in the value of NPV from
that based on annual payments is small.
Define the monthly value of r as rm where
Now, if the life of the project is N years, there will be 12N payments so the NPV is given by
But since
then
and
Examples
For an installation with a projected life of 10 years and a discount rate of 5%, assuming 0% energy cost inflation and no load growth, paid
annually,
If paid monthly,

For the same installation with a discount rate of 5%, energy price inflation of 3% and an average load current growth of 2%,
Note that r is > 1.
3.1.2 Loading
Few installations run at full load continuously. The maximum current must be used to select the initial conductor size and protective device rating
but, for financial purposes, the rms average load must be used. This requires some knowledge about the load profile. To find the rms load, the
profile is divided into a number of equal timeslots and the square of each current value calculated. The squares for all the time slots are summed,
averaged and the root value calculated. The result is that current which, if it flowed continuously, would be equivalent to the actual load.
However, there is a problem with this method because the temperature of the busbar, and therefore the resistance, changes with current, meaning
that energy losses for periods of low currents are overestimated, while those for periods of high current are underestimated. In practical terms the
effect is small; firstly, the error is unlikely to shift the choice of bar size from one standard size to another and, secondly, for an optimised bar the
current density is lower so the temperature rise, and therefore the variation, is much smaller.
Where the load profile is more complex, or there is a need for great accuracy, the suggested approach is to construct a straight line approximation
of the load profile, calculate the energy loss at each current level and then determine the time average of the energy loss.
3.1.3 Lifetime
Lifetime can have several meanings. At the design stage, lifetime is the period for which the installation is expected to be needed to fulfil its
original purpose. For a product line this may be quite short, while for a commercial building it will be rather longer. In practice, installations are
often used for far longer because, for example, new products can be manufactured using the same installation, perhaps with small modifications.
The useful lifetime is therefore longer than the initial ‘economic’ lifetime. The result is that the initial assessment – for the economic lifetime -
gives a lower weight to running costs than would have been the case for the actual lifetime. It also delays the end-of-life so any benefit from
recovering scrap conductor material is very heavily discounted so that it does not influence the financial decision.
3.1.4 Sensitivity Analysis
In most cases it will not be possible to accurately know all the data, for example, the design of the plant may not be finalised and future electricity
costs cannot be determined accurately, so it is necessary to test the result to check how dependent the appraisal is on each variable and make an
appropriate judgement.
3.2 Application to an Electrical Installation
Engineering design is always concerned with compromise within the boundaries of the design requirement. The functional requirements must be
met, of course, but there are many ways of doing so; using more final circuits will provide more flexibility and better isolation for sensitive loads,
while using larger conductor cross-sections will reduce energy losses and voltage drop resulting in lower energy losses and better dip resilience.
Specifying energy efficient load equipment will reduce energy consumption. Each of these decisions has an impact on the future running costs of
the installation in both energy and maintenance costs and an impact on the cost of the installation and its commissioning.

3.2.1 Installation Design
Many design decisions will be made based on the designer’s experience of having done similar designs in the past and, for many parts of an
installation, the ‘business-as-usual’ installation may also be the one with the lowest LCC because the base level is set by installation standards. For
other areas there is an opportunity to reduce the LCC by optimising the design. For example:
1. Selecting energy efficient equipment will reduce energy costs. This includes, for example, transformers, motors, pumps and fans.
2. Considering energy efficiency when determining the conductor size for cables and busbars will reduce energy losses considerably
in circuits feeding heavy loads. The 5% voltage drop at the load terminals, permitted by standards, represents an efficiency of only
95% - for a length of conductor! Unless the duty factor of the load is low, designing for a much lower voltage drop makes sense and
reduces the LCC.
3. Designing for reliability by using separate dedicated circuits for sensitive equipment and heavy fluctuating loads and designing for
reduced voltage drop in the installation to improve resilience against voltage disturbances will reduce downtime and maintenance
costs, reducing the LCC.
The decisions made at the design stage determine the scale of the LCC, so this stage is extremely important – it is worth investing time in this
process to reduce future costs.
3.2.2 Installation Costs
The installation costs include all the electrical design and installation costs for the project and, since they all occur at the start of the project, are
therefore taken into account at full cost. When comparing the LCCs of different design options, only those elements that vary would need to be
taken into account.
3.2.3 Recurring Costs
Recurring annual costs generally fall into two categories:
3.2.3.1 Maintenance Costs
All facilities require a certain amount of expenditure on maintenance. Reducing this requirement by, for example, designing for a lower working
temperature to give a longer, more reliable life, or choosing lighting equipment with a long service life to reduce replacement frequencies, can
significantly reduce costs.
3.2.3.2 Energy
Energy is expensive and the present upward trend is likely to continue making energy costs extremely important. In many industrial processes,
energy cost accounts for the majority of the total lifetime cost, so measures taken to increase energy efficiency at the design stage are often
extremely cost-effective. LCC provides a methodology to determine the economic optimum conductor size for a given loading and duty cycle.
However, designers must also consider voltage drop. Standard installation practice allows a maximum 5% voltage drop between the origin of the
installation and the load terminals. In many cases, this condition will be met or exceeded by the conductor size corresponding to the economic
optimum, but this cannot be assumed. For example, the economic optimum conductor size for a load with a low duty factor may be relatively
small, but the conductor length may introduce an excessive voltage drop. The installation designer must take care to allocate the voltage drop
‘budget’ to the various parts of the installation and, because busbars tend to be used in the higher current parts of the system, they deserve greater
attention. The voltage drop should always be checked once the economic optimum has been determined.
The questions that have to be answered are: what is the conductor size with the lowest LCC for a given circuit load, duty factor and lifetime, and is
the voltage drop sufficiently low?
The following simple example calculates the total cost – conductor cost plus energy loss - for a simple single phase circuit using a range of cross-
sections and a range of operating times. The circuit feeds a load of 500 A.

Table 9 lists the maximum working current for a range of vertically mounted busbars, assuming an ambient temperature of 35ºC and a working
temperature of 90ºC, operating in free air. The material costs listed are based on a copper price of $7500 per tonne.
Note that the current-carrying capacity of a bar is not proportional to the cross-sectional area (assuming the same ambient and working
temperatures etc). This is because convection becomes less efficient for larger bars since the air cooling flowing up the face at the top of the bar
has already been warmed as it flows up the bottom of the face. Consequently, the maximum busbar rating cannot be directly scaled although,
since the difference is not very large, scaling can be used to estimate the starting point for calculation.
Table 9 – Maximum Working Current for a Range of Busbar Sizes

Width (mm) Thickness (mm) Area (mm2) Weight (kg) Max Current (A) Material Cost (€)
25 6.3 158 1.403 530 7.518
31.5 6.3 198 1.768 639 9.472
40 6.3 252 2.245 776 12.029
50 6.3 315 2.807 932 15.036
63 6.3 397 3.536 1127 18.945
80 6.3 504 4.491 1390 24.057
100 6.3 630 5.613 1662 30.071
125 6.3 788 7.017 1987 37.589
160 6.3 1008 8.981 2433 48.114
In Table 10 the power loss for each metre length of busbar at a current of 500 A is shown, calculated at the estimated working temperature (which
is different for each size - see ‘2.0 Current-Carrying Capacity’).
Table 10 – Resistance and Power Loss at 500 Amps

Width (mm) Thickness (mm) Area (mm2) Max Current (A) Resistance at Resistance Loss at 500 Amp
20ºC (µohm/m) at working (W/m)
temperature
(µohm/m)
25 6.3 158 530 109.460 138.100 34.53
31.5 6.3 198 639 86.873 104.647 26.16
40 6.3 252 776 68.412 79.637 19.91
50 6.3 315 932 54.730 62.164 15.54
63 6.3 397 1127 43.436 48.462 12.12
80 6.3 504 1390 34.206 37.889 9.47
100 6.3 630 1662 27.365 29.982 7.50
125 6.3 788 1987 21.892 23.635 5.91
160 6.3 1008 2433 17.103 18.465 4.62
This data is plotted in Figure 38.

40.00
35.00
Power loss per metre length at 500 Amp (W)
30.00
25.00
20.00
15.00
10.00
5.00
0.00
0 20 40 60 80 100 120 140 160
Width of copper bar (mm) [Thickness 6.3 mm]
Figure 38 - Power loss versus width for 6.3 mm thick copper bars
The power loss needs to be expressed in terms of money by converting to energy loss (by multiplying by the operation time) and multiplying by the
cost of energy. For the purposes of this exercise, the cost of energy is assumed to be 0.15 Euro per kWh. Table 11 and Figure 39 show the energy
cost per metre for a range of operation times.
Table 11 – Energy Cost per Metre for Various Widths of Copper Bars at 500 A Load
Cost of Energy (€ per metre)
Width (mm) 2000 hrs 5000 hrs 10 000 hrs 20 000 hrs 50 000 hrs
25 10.36 25.89 51.79 103.58 258.94
31.5 7.85 19.62 39.24 78.49 196.21
40 5.97 14.93 29.86 59.73 149.32
50 4.66 11.66 23.31 46.62 116.56
63 3.63 9.09 18.17 36.35 90.87
80 2.84 7.10 14.21 28.42 71.04
100 2.25 5.62 11.24 22.49 56.22
125 1.77 4.43 8.86 17.73 44.32
160 1.38 3.46 6.92 13.85 34.62

300
250
50 000 hrs
200
Cost of energy (€)
150
100
20 000 hrs
50
10 000 hrs
5000 hrs
2000 hrs
0
0 20 40 60 80 100 120 140 160
Figure 39 – Cost of energy loss per metre versus width for 6.3 mm thick copper bars for a range of operating times
By comparison with the material costs in Table 9 it is immediately apparent that, for any significant operational life, the cost of energy is by far the
dominant element in the total cost of the system. Consider, for example, the 5000 hour operation case.
60
50
Total cost over 5000 hours (€)
40
Optimum cross-section Cost of material (€)

Cost (€)
30
20
10
Cost of energy losses over 5000 hours (€)
0
0 20 40 60 80 100 120 140 160
Figure 40 - Total cost per metre versus bar width for 5000 hours operation at 500 A load
Figure 40 shows the total cost, material cost (from Table 9) and energy cost (from Table 11) per metre of bar. As the width of the bar increases, the
cost of material increases linearly, while the cost of energy decreases super inversely as the resistance and temperature decrease. The total cost
curve shows a clear minimum where the total cost is optimised between investment cost (material) and operating costs (energy).
Because the cost of energy is proportional to operation time, the optimum bar width also increases with operation time, as shown in Figure 41,
which plots the total cost per metre of bar for a range of operation times. As operation time increases, the optimum bar width increases.

200.00
50 000 hours
180.00
160.00
140.00
120.00
Total cost (€)
100.00
20 000 hours
80.00
60.00
10 000 hours
40.00
5000 hours
20.00
2000 hours
0.00
0 20 40 60 80 100 120 140 160
Figure 41 - Total cost per metre of bar versus bar width for a range of operating times
Figures 40 and 41 illustrate two important points:
At the minimum bar size – i.e. the size that would be selected by conventional methods – the cost of lifetime energy is many times the cost of the
conductor; a factor of approximately 3.4 at 5000 hours. In fact, such a bar wastes its own cost in energy in less than 1500 operational hours, at
the material and energy costs used in this example.
The minimum of the total cost curve is shallow, meaning that there is some latitude for judgement in the selection of the final size with a high
degree of confidence in the outcome. For example, if the operational hours were not well known, but could be expected to be between 20 000 and
50 000 hours, a 100 mm bar could be confidently selected, being a little over optimum size for 20 000 hours and a little under optimum size for
50 000 hours.
Figure 42 is an alternative plot of the data in Figure 41 showing the optimum current density for a range of operating hours. As would be
expected, the optimum current density reduces as operating hours increase.

200
50 000 hrs
180
160
140
Total cost (€/metre)
120
20 000 hrs
100
80
60 10 000 hrs
40
5000 hrs
20 2000 hrs
0
3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4
Current density (A/mm2)
Figure 42 - Total cost per metre against current density for a range of operation times
Figure 43 shows the estimated working temperature plotted against bar width at 500 A load assuming an ambient of 35ºC. As would be expected,
lower current density results in reduced temperature rise, bringing the benefit of increased reliability due to reduced thermal stresses on joints and
mountings, as well as optimised lifetime costs.
90
80
Estimated working temperature (oC)
70
60
50
40
30
0 20 40 60 80 100 120 140 160
Width of Bar (mm)
Figure 43 - Estimated working temperature versus width of bar (mm) for 500 A load
So far, this simple analysis has ignored the time value of money; to make the argument credible to financial managers, that must be corrected.
In Table 12 the discount factors from Table 8, the material cost figures from Table 9 and the cost of energy loss figures from Table 11 are used to
calculate the net present value – i.e. the discounted cost – of operating each metre of bar for 5 years at 2000 hours load per year. The assumed
discount rate is 5% but for the sake of clarity, no account is taken of inflation or load growth.

In this case, because the energy is paid for throughout the operating time, no discount is applied to Year 0. Local accounting practice may require
discounting to be applied to Year 0 in some cases.
Table 12 – Present Value (€) per Metre of Bar

Cost of Energy (2000 hours per year)
Discount Factor Year 0 Year 1 Year 2 Year 3 Year 4 Total

1 0.952 0.907 0.864 0.823
Bar Width Power Loss Cost of Bar
(mm) per Metre (euro)
(watt) Total
25 34.53 7.518 10.36 9.8627 9.3965 8.9510 8.5263 47.10 54.61
31.5 26.16 9.472 7.85 7.4732 7.1200 6.7824 6.4606 35.69 45.16
40 19.91 12.029 5.97 5.6834 5.4148 5.1581 4.9133 27.14 39.17
50 15.54 15.036 4.66 4.4363 4.2266 4.0262 3.8352 21.18 36.22
63 12.12 18.945 3.63 3.4558 3.2924 3.1363 2.9875 16.50 35.45
80 9.41 24.057 2.82 2.6846 2.5577 2.4365 2.3209 12.82 36.88
100 7.45 30.071 2.24 2.1325 2.0317 1.9354 1.8435 10.18 40.25
125 5.93 37.589 1.78 1.6946 1.6145 1.5379 1.4649 8.09 45.68
160 4.62 48.114 1.38 1.3138 1.2517 1.1923 1.1357 6.27 54.39
Figure 44 shows the comparison between the discounted and non-discounted cost, this time for a 10 year life.
120
100
20 000 hours (No discount)
80
Cost per metre (€)
60
2000 hours per year for 10 years at 5% discount

40
20
0
0 20 40 60 80 100 120 140 160
Width of bar (mm) [Thickness 6.3 mm]
Figure 44 - Cost per metre (€) against bar width for a) 20 000 hours operation without discount and
b) 2000 hours per year for 10 years at 5% discount

The effect of discounting is, of course, to progressively reduce the value of the later years and therefore the total cost. But, as Figure 44 clearly
shows, it also shifts the minimum cost bar width towards a lower value. This is an obvious conclusion if discounting is viewed as reducing the
operating time in later years instead of the cost of each operating hour in later years.
A warning is appropriate at this point; using exaggerated discount rates will reduce the viability of any investment in the same way as imposing
too-short a payback period would, so it is important that discount rates are chosen with care and without prejudice.
3.2.4 End of Life Costs
At the end of life the installation must be removed and the materials recycled. In electrical systems the copper conductor has significant value,
making recovery worthwhile; however, it is so far into the future, it is often assumed that the discounted value will be low. However, commodity
prices are very volatile with an underlying upward trend so, if material price inflation is included in the calculation, the NPV of the residual material
may be significant.
An additional problem is that the task of removing the installation is usually carried out by a third party under contract, so the operator sees a
contract cost, while the third party receives the revenue from recycled materials.
3.3 Conclusion
Life cycle costing provides a more realistic view of the cost of a project and is particularly useful in comparing two or more design options. The
curve of LCC versus conductor size is quite flat near the minimum; selecting the largest conductor size in range provides both a short term
optimum and reduced exposure to future energy price increases.

4.0 Short-Circuit Effects
4.1 Introduction
Like all electrical circuits, busbars need to be protected against the effects of short-circuit currents. The open construction of busbars increases
the risk of faults, e.g. by the ingress of foreign bodies into air gaps, and the risk of consequent damage is high due to their high normal operating
currents and the amount of energy available.
Very high currents lead to rapid and extreme overheating of the bars with consequent softening of the material and damage to the support
structure. At the same time, the electromagnetic forces generated will distort the softened conductors which may break free from their supports.
Resonance effects may make the situation worse.
4.2 Short-Circuit Heating of Bars
The maximum potential short-circuit current depends on the source impedance of the supply and reduces along the length of the bar as the
impedance increases. For the purpose of ensuring the safety and integrity of the bar, the short circuit current should be initially calculated close to
the feed end of the bar, making no allowance for bar impedance, to establish the worst case.
When a fault occurs, the short-circuit current will be many times the normal operating current and will flow until the protective device operates.
Because the time duration is small – a few seconds at most – it is usual to assume adiabatic heating, in other words, that, over the time scale of
interest, there is no significant cooling effect and that all the heat generated by the current flow is retained in the bar. Therefore it is assumed that
the temperature rise of the bar is simply linear - this simplifies the calculation significantly and yields a conservative result.1
The amount of heat energy required to raise the temperature of a unit mass of material by one degree Centigrade is called the specific heat. For
copper, at room temperature, the value is 385 Joule/kg/K. Knowing the mass of the bar and the energy produced by the short circuit current, the
temperature rise can be calculated:
or
where:
Q is the amount of heat added to the bar (Joule)
S is the specific heat of the bar material (J/kg/K)
m is the mass of the bar (kg)
tr is the temperature rise.
The amount of energy dissipated in the bar is:

where:
P is the power dissipated in the bar (W)
T is the time for which the power is dissipated (seconds).
Therefore, or
where Dtr is the rate of temperature rise in degrees C per second.
The power dissipated in the bar is given by
or

1
A similar simplification is made in determining cable behaviour under fault conditions.

where:
R is the resistance of the bar (Ω)
r is the resistivity of the bar material (Ωm)
l is the length of the bar (m)
A is the cross-sectional area of the bar (m2).
Substituting for W and M gives:
where:
D is density (kg/m3).
Two of the physical constants, specific heat and resistivity, vary considerably with temperature so it is not obvious which values should be used.
Resistivity increases with temperature (by a factor of 1.6 from 20ºC to 300ºC) so increasing the energy dissipated as the temperature increases,
while specific heat falls by around 8% over the same range.
Using room temperature values, and adjusting to use convenient units, gives the initial rate of temperature rise as:

where:
I is current in kA
A is the cross-sectional area in mm2.
At higher temperatures – in other words, as the fault current continues to flow until the protective device operates – the rate of temperature rise
will increase as the resistance of the bar increases. The worst case rate of rise at the feed end of the bar is approximately:

However, if the short circuit is at some distance from the feed end, the fault current magnitude will be lower due to the resistance of the bar and
will reduce further as the bar temperature rises.
In practice, what is important is that the final temperature of the bar remains lower than the limiting design temperature throughout the short
circuit event. The limiting temperature for copper busbars is determined by the temperature resistance of the support materials but, in any case,
should not exceed ~ 200ºC. The maximum circuit breaker tripping time is 200/Dtr seconds.
Busbars that have been subject to short circuit should be allowed to cool and inspected before being returned to service to ensure that all joints
remain tight and that the mountings are secure. Note that, although the heating time – the duration of the fault – is quite short, the bar will
remain at high temperature for a considerable length of time. Also, because of the very high thermal conductivity of copper, parts of the bar
beyond the fault will also have become hot.
4.3 Electromagnetic Stresses
Busbars are subject to mechanical forces since each is carrying a current though the magnetic fields caused by currents in other bars. When
alternating currents are flowing, the forces have a steady component, but also a vibrational component at twice the frequency of the alternating
current. Under normal working conditions these forces are of little consequence.
However, if the bars are mounted on supports, each section will have a resonant frequency. If this frequency is close to twice the supply current
(or any significant harmonic current), then resonant vibration of these beams may occur. This rather special and uncommon circumstance can lead
to high vibrational displacements and possibly to metal fatigue or loosening of joints and connections. The problem may be avoided by choosing
an appropriate spacing of the supports or cured by introducing additional intermediate supports. Methods of estimating mechanical resonant
frequencies are given later.

If large currents flow, such as when a short circuit occurs, the forces can be more important. The unidirectional component of the forces,
exacerbated by the vibrational component, can lead to permanent bending and distortion of the bars or damage to, and even breakage of supports.
The peak, or fully asymmetrical, short-circuit current is dependent on the power factor (cos f) of the busbar system and its associated connected
electrical plant. The value is obtained by multiplying the rms symmetrical current by the appropriate factor given in balanced three-phase short-
circuit stresses.
The peak current, I, attained during the short circuit, varies with the power factor of the circuit (see Table 13):
Table 13 - Power Factor and Peak Current

Power Factor Peak Current as Multiple
of Steady State rms
0 2.828
0.07 2.55
0.20 2.2
0.25 2.1
0.30 2
0.50 1.7
0.70 1.5
1.0 1.414
The theoretical maximum for this factor is 2√2 or 2.828 where cos f = 0. If the power factor of the system is not known, then a factor of 2.55
will normally be close to the actual system value, especially where generation is concerned. These peak values reduce exponentially and, after
approximately 10 cycles, the factor falls to 1.0, i.e. the symmetrical rms short-circuit current. As shown in Figure 45, the peak forces therefore
normally occur in the first two cycles (0.04 s).
Theoretical peak current
Peak current (fully asymmetrical)
Symmetrical current
Figure 45 – Short-circuit current waveform

Bars carrying like parallel currents are attracted towards each other. Thus, if the currents in one phase are carried in separate conductors in parallel
in the same phase, then the gap between them may be reduced by the mechanical forces. More usually we deal with bars carrying anti-parallel
currents when the bars repel each other.
Eddy current effects in bars carrying anti-parallel currents bring the current flow in a bar closer to the sides facing the other conductor. This will
lead to an increase in the repulsive force. Precise formulae for this are not available but an over-estimate of the magnitude of the force under
these circumstances may be made by replacing the centre line distance in the formula for the force per unit length by the distance between the
faces.
4.3.1 Estimating the Forces Between Parallel Sets of Bars
Assuming that the bars are round will give values for the forces which, used with a factor of safety in estimating strength, are normally reliable.
If there are many bars in the configuration then the forces between individual pairs of bars may be simply added together, taking into account
their direction. In the case of ac, one may need to take into account the phases of the currents.
Bar shape usually has a minor effect unless the bars are very close together, as will be illustrated in the example of rectangular bars.
4.3.1.1 Round Bars
For two bars with spacing d and carrying parallel currents I1 and I2 , the repulsive force per unit length is

where:
d is spacing in m
µo is absolute permeability = 4π x10-7 Henrys per metre
F is force per unit length in N/m.
For two bars, at spacing d, carrying anti-parallel direct currents (i.e. go and return currents) of magnitude I, the repulsive force per unit length is
-6

For two bars at spacing, d, carrying anti-parallel alternating currents (i.e. go and return currents) of magnitude I, the repulsive force per unit length is
-6

where:
d is spacing in m
� = 2πf
f is frequency in Hz.
Thus we have a steady repulsive force, together with a force alternating at twice the frequency of the current of amplitude equal in magnitude to
the steady component of force. The result can also be seen as a periodic force varying from zero to twice the average force.

For balanced three phase currents of rms value I at frequency � = 2π f (� in rad/s, f in Hz), the currents are:
The standard triangular and inline arrays are shown in Figure 46.
y FBA
C
FCB FAB FBC

d 30°
A B C
d
A d d FAC
d x
FCA x
B
(a) (b)
Figure 46 - Three phase system with spacings d (mm). (a) triangular array, (b) inline array
4.3.1.1.1 Triangular Array
The net force per unit length in the x-direction on bar A in Figure 46(a)
The net force in the y-direction on bar A in Figure 46(a)
The net result is a steady force outwards along the perpendicular to the line BC of magnitude
The alternating forces in the x and y direction are in phase quadrature and of equal magnitude and so constitute a force tending to move the
conductor in a circle. The magnitude is the same as that of the steady force.

4.3.1.1.2 In-Line Array
The force per unit length in the x-direction (i.e. an outward repulsive force) on bar C in Figure 46(b) is
This is a steady outward force with a superposed 2nd harmonic vibration. The amplitude of the harmonic force is slightly greater than the steady
force.
The force per unit length on the centre bar, B, is purely vibrational. In the x-direction it has magnitude
The force per unit length in the x-direction on bar A is
This too is on average a repulsive force outwards in the negative x-direction.
4.3.1.2 Bars of Rectangular Section
For bars of rectangular section set out orderly, the force between the two may be expressed as the force computed as if they were round bars
multiplied by a correction factor. Thus the force, in N/m, between two rectangular bars with centre line spacing d, is
The factor K depends on the length of the sides, a and b, of the rectangular cross-section and the spacing, d, of the centre lines.
K is shown in Figure 47 as a function of
Note that the curves labelled a/b = 0 and b/a = 0 represent limiting values.

d
b
=0
Factor, K a
1.4 b
1.3 a a
a
=5
1.2 b
a
= 2. 5
1.1 b
1.0
a
=1
0.9 b
a
0.8 = 0.5
b
0.7
a
0.6 = 0.25
b
0.5
a
= 0.1
0.4 b
0.3
a d
=0
0.2 b
0.1 b
0
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 a a
d-a
a+b
Figure 47 - Factor K for calculating the force between two bars of rectangular section
For a pair of rectangles whose long sides face each other, it will prove helpful to use the equations below in which d is replaced by b and factor K
replaced by Kb which is plotted in Figure 48.
Form Factor, kb
d
3.2 Form Factor, kb
3.0 1.0 b
2.8
d
2.6
0.5 a a
2.4
2.2 b
2.0
0.2 1
1.8
a a
1.6
0.1 2.5
1.4 a
1.2 b
5
1.0 0.05
0
0.8
0.1 10
0.6
0.25 a
0.4 b 0.02 16
0.5
0.2
1
0 2.5
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
d-a d-a
(a) a+b (b) a+b
Figure 48 - Form factor Kb. (a) low values of a/b (long sides facing each other) (b) high values of a/b.

An approximate formula for Kb accurate to within 4% for and is:

where:
4.4 Mounting Arrangements
The mountings are required to restrain the busbars under all conditions to ensure that:
• the material is not over stressed

• a safe distance is maintained between bars in normal service and under short-circuit conditions by limiting maximum deflection
• the bars do not vibrate excessively (which would impair the long term efficacy of joints).
4.4.1 Maximum Permissible Stress
The maximum permissible stress in a conductor is the resultant of:
• its own natural weight (Fm) and

• in the case of busbars installed out of doors, the additional forces of wind (Fw) and ice (Fi) loadings and
• the magnetic forces resulting from a short circuit, noting that the direction of a short-circuit force (Fs) depends on the position of
adjacent phases and the direction of the currents in them.
In a general case, the resultant can be calculated by the following method:
FS
Fw
Fm+F1

and
The maximum skin stress in the conductor can then be calculated using the following formula:

where:
f is maximum skin stress, N/mm2
L is span, mm
I is moment of inertia, mm
D is diameter or depth of rectangular section, mm.
This equation is valid for a uniformly loaded single beam which is freely supported at both ends or freely supported at one end and fixed at the
other. For a beam which is horizontally fixed at both ends, maximum skin stress is reduced to one third at the centre and to two-thirds at its ends.
The maximum permissible stress is dependent on the conductor material, temper, etc. but must not exceed the material proof stress otherwise
permanent deformation will occur. For a conductor manufactured from hard drawn copper, the value is approximately 245 N/mm2.
4.4.1.1 Moment of Inertia
In the above formula the moment of inertia, I, for the section of the beam has to be calculated about the neutral axis which runs parallel to the
beam where the beam has zero tensile forces. In most cases this is the same axis of the centre of cross-section.
Shaped profiles have much higher values of moment of inertia than simple bars and rods (see Section 5.0 Busbar Profiles).
For a rectangular section of depth D mm and breadth B mm
For a circular section of diameter D mm

or
For a tubular section of internal diameter d mm and external diameter D mm

or
It should be noted that the value of I for a given cross-section is dependent on the direction in which each individual force is applied.

4.4.2 Deflection
It is important to know how far a bar will deflect so that the necessary clearances between bars and other structures can be maintained.
The maximum deflection of a beam carrying a uniformly distributed load, and freely supported at each end, is given by the formula:
where:
D is maximum deflection, mm
Fm is weight per unit length of loaded beam, N/mm
L is beam length between supports, mm
E is modulus of elasticity (124 x 103N/mm2)
I is moment of inertia of beam section, mm4.
If one end of a beam is rigidly fixed in a horizontal position, the deflection is 0.415 times that given by the above formula. If both ends of a beam
are rigidly fixed in a horizontal position, the deflection is 0.2 times that given by the above formula.
Thus, with a continuous beam freely supported at several points, the maximum deflection in the centre spans may be assumed to be 0.2 times that
given by the formula, while the deflection in the end spans is 0.415 times. Therefore, the deflection in the end spans may be assumed to be twice
that in the centre spans, assuming equal span distances.
4.4.3 Natural Frequency
The natural frequency of a beam simply supported at its end is:
and for a beam with both ends fixed horizontally it is:

where:
fn is natural frequency, Hz
D is deflection, mm.
As the deflection with fixed ends is 0.2 times the value with freely supported ends, it follows that the natural frequency is increased by 2.275 times
by end-fixing; fixing one end only increases the natural frequency by about 50%.
Busbar systems should be designed to have a natural frequency which is not within 30% of the vibrations induced by the magnetic fields resulting
from the currents, including any significant harmonic currents, flowing in adjacent conductors.
Where equipment is to be mounted outside, natural frequencies of less than 2.75 Hz should be avoided to prevent vibration due to wind eddies.

5.0 Busbar Profiles
David Chapman
5.1 Introduction
Busbar profiles are very often used in distribution panels and switchboards where they have some distinct advantages. The design considerations
for this application are significantly different from the relatively long vertical and horizontal busbars discussed previously. For example, the bars
are semi-enclosed in cabinets, so it is the air flow through the cabinet which determines the cooling available. Moreover, these cabinets contain
a substantial number of devices such as circuit breakers, which also generate heat and rely on the busbar system to remove it. Space constraints,
mounting considerations and ease of assembly also favour the use of profiles.
Profiles are more difficult and costly to manufacture than flat bars due to technical contraints, including:
• Complex design
• More elaborate machine set-up
• More delicate production process
• More complicated packaging.
These constraints are now mostly under control and currently several fabricators produce hundreds, or even thousands, of different shapes for
electrical applications.
Many electrical parts, which are often cut from sheet, can be economically produced by slicing a profiled bar.
Figure 49 - Typical copper profiles

5.2 Reasons for Using Profiles
Profiles can be used to reduce the weight of copper conductors in an existing product. For an existing oversized conductor, the design flexibility
available to define the shape of drawn copper profiles allows the engineer to design a profile smaller without affecting those surfaces which are
constrained by the presence of existing support points, or contact interfaces.
5.2.1 Skin Effect Reduction
Skin effect losses are perhaps less important in switchgear, where the lengths are relatively short, than it is in long distribution busbars. However,
taking proper account of skin effect often allows the amount of material to be significantly reduced.
5.2.2 Weight and Cost Saving
As discussed elsewhere, the apparent impedance of thick section conductors is increased due to skin effect, meaning that the current density in
the centre is much lower than that at the surface. In a simple rectangular bar, there are areas of the cross-section in which the current density is
comparatively low; using a correctly designed profile allows for a near homogeneous current density across the section that allows material to be
removed from these areas, saving material, cost and weight.

5.2.3 Integrated Fixings and Mountings
Profiles can be designed to allow fast, hole-free assembly and jointing, which can significantly reduce the time required to assemble bars in a
cabinet, improving consistency and quality.
5.2.4 Retention of Intellectual Integrity
Using bespoke profiles allows OEMs to ensure that cabinets assembled by sub-contractors conform to their specifications, avoiding the risk of
substitution of inferior materials and assembly processes.
5.3 Economics of Profiles
The material savings generated by replacing a standard conductor by a profile of even a slightly smaller cross-section can easily offset the higher
processing costs. At current refined copper prices, by far the largest contributor to the cost of a drawn copper profile is the cost of the copper
material itself.
Using profiles also gives the opportunity to capture other savings:
• Hole-free joints and mountings yielding to lower assembly time and reduced complexity
• Lower scrap generated at the point of use (drilling and cutting not required)
• Flexible sourcing (optimum profiled conductor or standard product with identical supports and mounting points)
• Optimum material usage for the specific application
• Easier re-engineering and conductor optimisation on an existing product.
5.4 Practical Profiles
5.4.1 Manufacturing Process
Profiles are manufactured by extrusion, followed by one or more stages of drawing. The copper is extruded, either continuously or discontinuously
through a die along a horizontal bench which supports the material. It is either cut into lengths or coiled for further processing.
The profile is then drawn through one or more dies, the last of which defines the final size and shape. Drawing is a cold working process so the
drawn material is significantly hardened. The choice of extruded size and the number of drawing stages and their reductions determines the final
hardness of the material; usually, the aim is to achieve the final size and shape without excessive hardening.
For profiles it is important that the size reduction during drawing is uniform – in other words that, while the size is reduced, the shape remains
similar – to prevent differential hardening of the bar. For example, inappropriate tool and process design can, among other things, lead to
problems with twisting, more so for asymmetric profiles.
5.4.1.1 EN 13605
EN 13605 is the standard that specifies the composition, properties and dimensional tolerances for copper profiles and profiled wire for electrical
purposes which would fit within a circumscribing circle with a maximum diameter of 180 mm. The standard also specifies how ordering
information should be structured between supplier and customer.

180 mm diameter envelope
hmax s smin
bmax
Figure 50 – A non-typical profile indicating the dimensions used in the standard
It is not intended here to describe the requirements of the standard in great detail but some aspects which are particularly relevant to profiles
deserve mention.
Figure 50 shows a non-typical profile and the dimensions used to specify tolerances. Note that the standard is limited to those profiles that have
a cross-section that can be contained within a circle of 180 mm diameter. A large majority of profiles are smaller than this, typically limited to 165
mm x 30 mm or 65 mm square with a weight of 30 kg/m. Some manufacturers are capable of manufacturing much larger profiles with maximum
dimensions and the appropriate tolerances must be agreed between customer and supplier. Maximum dimensions for two processes are given in
Table 17.
The tolerances on the major dimensions, b and h in Figure 50, are categorised by value and defined in six groups according to:
• The diameter of the circumscribing circle diameter in three groups, up to 50 mm, between 50 mm and 120 mm and between 120 mm
and 180 mm
• The ratio of the maximum dimension (b or h) to the minimum thickness (s) into two sets:
o bmax < 20 smin or hmax < 20 smin
o bmax ≥ 20 smin or hmax ≥ 20 smin.
Tables 14, 15 and 16 list the tolerances.

Table 14 – Tolerances for Dimensions b and h for bmax or hmax <20:1
Nominal Dimensions b and h (mm) Tolerances for Dimensions b and h within a Circumscribing Circle
Over Up to and including <= 50 >50, <=120 >120, <=180
- 10 ±0.11 ±0.18 ±0.29
10 18 ±0.14 ±0.22 ±0.35
18 30 ±0.17 ±0.26 ±0.42
30 50 ±0.20 ±0.31 ±0.50
50 80 ±0.37 ±0.60
80 120 ±0.44 ±0.70
120 180 ±0.80
Table 15 - Tolerances for Dimensions b and h for bmax or hmax ≥ 20:1

Nominal Dimensions b and h (mm) Tolerances for Dimensions b and h within a Circumscribing Circle
Over Up to and including <= 50 >50, <=120 >120, <=180
- 10 ±0.18 ±0.29 ±0.45
10 18 ±0.22 ±0.35 ±0.55
18 30 ±0.26 ±0.42 ±0.65
30 50 ±0.31 ±0.50 ±0.80
50 80 ±0.60 ±0.95
80 120 ±0.70 ±1.10
120 180 ±1.25
The tolerances on thickness s in Figure 50 are specified for profiles with a circumscribing circle of 50 mm and below and one of greater than
50 mm.
Table 16 - Thickness Tolerances

Nominal Thickness (mm) Thickness Tolerance within a Circumscribing Circle
Over Up to and including <=50 >50, <=180
- 3 ±0.13 ±0.20
3 6 ±0.15 ±0.24
6 10 ±0.18 ±0.29
10 18 ±0.22 ±0.35
18 30 ±0.26 ±0.42
30 50 ±0.50
5.4.1.1.1 Straightness, Flatness and Twist
Profiled sections are often used as pre-manufactured components for switchgear assemblies where space is limited and clearance distances are
important, so it is essential that the sections fit correctly. Since profiles are much stiffer than flat bars, they cannot easily be pulled into alignment
during assembly, so it is essential that they are straight, flat and free from twisting. Measurement of straightness, flatness and twist is made with
the test piece lying on a flat reference surface.
In Figure 51, h represents the deviation from straight over length l ; h1 must be less than 3 mm per metre length (l1 ) up to 3 m while h2 must be
less than 1.2 mm for any 400 mm length. It should be noted that, for special requirements, a straightness of less than 1 mm in a 1000 mm length
can be achieved.

Test piece
h1
l2 h2
Reference surface
l1
Figure 51 – Measurement of straightness
Figure 52 illustrates the measurement of flatness.
0.03h
0.01b b
Figure 52 – Measurement of flatness
Figure 53 illustrates the method for measuring twist. The twist tolerance is defined by
where:
v is the displacement from the reference plane
b is the width of the test piece
f is a coefficient, the tolerance limit.
Again, the limiting value of f is grouped according to the diameter of the circumscribing circle (≥ 15 and ≤ 50; >50 and ≤120; >120 and ≤180) as a
maximum value per metre of length and as a maximum for any length over 2 m.
l
Reference plane
Test piece
v
Figure 53 - Measurement of twist

Table 17 – Coefficient for Twist Tolerance
Diameter of Circumscribing Circle (mm) Coefficient f for Twist Tolerance v
Over Up to and including Per 1m length On total length l over 2m
15a 50 0.08 0.15
50 120 0.05 0.10
120 180 0.04 0.08
a including 15
5.4.2 Design for Manufacturing
As might be expected, some profile designs are easier to manufacture than others; taking account of the capabilities of the manufacturing process
during the design phase may help to reduce the cost and result in more consistent quality of the final product.
The overall size of the profile must lie within the capacity of the available plant. Manufacturers quote different limits depending on the process
used, so advice should always be sought from potential manufacturers.
Table 18 – Maximum Sizes of Profile According to Two Manufacturers

Process 1 Process 2
Maximum width 250 mm 250 mm
Maximum depth 115 mm 60 mm
Maximum cross-sectional area 8600 mm 2
3000 mm2
Maximum weight per metre 80 30
Some specific design awareness allows the achievement of the right characteristics:
• Have a uniform wall thickness

• Have simple, soft lines and radiused corners
• Be symmetrical
• Have a small circumscribing circle
• Not have deep, narrow channels.
5.4.2.1 Wall Thickness
Small wall thicknesses require higher extrusion force (or, alternatively, a lower extrusion speed) and result in faster die wear. For simple, cost-
effective, single pass drawing, wall thicknesses of less than 3 mm should be avoided.
5.4.2.2 Avoid Sharp Corners
Very sharp corners are difficult to produce consistently because the die wears very easily. Specifying a radius of 1 mm is often sufficient. In
general, soft contours are to be preferred wherever the functionality of the profile allows.
5.4.2.3 Symmetry
Wherever possible, profiles should be designed to be symmetrical so that the shear centre is coincident with the centre of gravity. In an asymmetric
profile, loads on the profile (including those introduced in manufacture) will result in torsional stresses and result in twisting. Unbalanced
(asymmetric) profiles tend to twist during the drawing operations: tight geometric tolerances will require longer setup time and more complex
tooling development.

5.4.2.4 Be Compact
Profile cross-sections with a low aspect ratio are easier to produce. The profile fits into a smaller circumscribing circle so the die will be smaller
and stronger and the tolerances smaller.
5.4.2.5 Avoid Deep Narrow Channels
For profiles with channels, the depth of the channel should not exceed three times the width, otherwise the strength of the die will be
compromised. A slightly higher ratio (1:4) can be achieved if the bottom is semi-circular and the corners at the opening have a large radius or are
tapered.
5.4.2.6 Avoid Hollow Chambers
While hollow chambers can be formed in extruded profiles, the die is more complex and less robust. As a result, the product uniformity and die
life, as well as productivity, are reduced.
5.4.3 Functional Design
Some advantages of using profiles rather than flat bars stem from the ability to integrate features into the design that will aid subsequent
assembly processes such as mounting and jointing, as illustrated in Figure 54.
The lug on the end of the profile (1) can be made to match existing moulded supports. The channel (2) can be dimensioned to hold captive a
hexagonal-headed bolt for attaching jointing plates. Fins (3) reduce skin effect and increase the surface area and, therefore, heat dissipation
allowing the bar to run at lower temperature. The thin central section (4) also reduces skin effect.
3
2
4 1
Figure 54 - Profile cross-section showing mounting lugs and slots for bolt-head
5.5 Electrical Design Considerations
5.5.1 Skin Effect
Eddy-current effects in profiles can be reduced with respect to rectangular bars because a significantly higher proportion of the copper conductor
material can be located further from the centre of the magnetic field. Because skin effect is reduced, the amount of material required is reduced.
This is particularly important where two or more conductors of the same phase are placed side by side. In this configuration the current density in
some areas of the bars is comparatively very low, so the bars can be shaped to remove this ineffective material.
For slim profiles – where the material thickness does not exceed 25 mm - an approximate way to assess skin effect is to calculate the average
profile thickness and obtain the relevant resistance ratio from Figure 55.

2.2
1
2
2.0
4
1.8
5
S b/a
8
1.6
16
32
1.4
Infinity
2a
1.2
2b
1.0
0.5 1.0 1.5 2.0 2.5 3.0
√( f, Hz
Rdc, µΩ/m
)
Figure 55 - Rac /Rdc as a function of the parameter √f/Rdc with f in Hz and Rdc in µΩ/m
5.5.2 Thermal Dissipation
Many of the components used in switchgear generate a significant amount of heat, which is dissipated by conduction via their connections to
the bars. This heat, as well as that generated in the bars, must be lost by convection and, to a lesser extent, by radiation. Ultimately, the heat is
removed from the cabinet by the flow of air through the cabinet by natural convection or by forced ventilation.
Since a profile has a larger surface area than an equivalent rectangular bar, the larger air-to-bar contact surface means that more heat can be
lost by convection. This is particularly true for vertically orientated profiles where the fins form channels to guide the rising air. In long vertical
channels, the air moves at increasing velocity as it heats up, but it also becomes less efficient at absorbing heat from the profile and it is necessary
to provide baffles to spill the air so that it can be replaced by cooler air.
On the other hand, the presence of fins and channels may significantly alter the natural air flow and cause air traps where air flow is stalled. Also,
the air wetting one wall of a vertical channel may interfere with the air wetting the other walls. These effects can be reduced by ensuring that:
• Horizontal fins are adequately spaced from each other

• Vertical channels are sufficiently wide
• Vertical channels are divided so that the air flow is interrupted - at fire stops, for example – and fresh air introduced.
Figure 56 shows a thermal image of two common profiles under load.

Figure 56 – Thermal image of two profiles under load
5.5.3 Jointing and Mounting
Since profiles cannot normally be overlapped, joints are usually achieved by the use of fishplates bolted to adjacent lengths of profile.
Alternatively, the profiles can be designed to include channels to retain bolt heads (as shown in Figure 54) for mounting and jointing. Connecting
plates are fastened to each of the outside faces of two adjacent sections of profile, as shown in Figure 57.
In the example shown, each plate is 5 mm thick and 120 mm long and is attached by four M8-bolts to each profile. The bolt torque is 19 Nm,
resulting in a high contact pressure. Assembly requires the use of a torque spanner and can be accomplished by relatively unskilled operators.
Figure 57 – Hole-free joints

Although the contact area is rather smaller than a normal flat overlap joint, the contact pressure is rather high and efficiencies (i.e. the ratio
between joint conductance and conductance of an equal length of continuous profile) of higher than 100% are possible, with freedom from hot-
spots and the ability to withstand the stresses due to short-circuit currents.
5.5.4 Short Circuit Performance - Moment of Inertia

x1 x2
Under short-circuit conditions busbars are subject to very large
electromagnetic forces which deform the bars, reducing clearance
distances. For the same current, the amount of deformation is determined
by the moment of inertia. A higher moment of inertia results in smaller
deflections, so that busbars can be placed closer together making the 2 y2
installation more compact.
It must be remembered that higher momentum does not increase the

short-circuit capacity of the system, which is usually limited by the short-
circuit temperature rise and the effect this might have on the mechanical
Composite axis
properties of the bars and other associated components.
Because profiles are designed to have a high proportion of their mass

as far as possible from the centre (to reduce skin effect), they have a x3
relatively high moment of inertia and deform less under short-circuit
conditions.
5.6 Calculation of Moment of Inertia of Complex Sections

y1 1 3 y3
The following section gives an outline of the procedure for calculating
the moment of inertia of a complex section about the neutral axes
perpendicular to the resolved components of the applied force. It is valid
only for shapes that can be broken down into elements that have neutral
axes which are parallel to each other, implying that the elements are
either parallel or perpendicular to each other.
For a profile, the moment of inertia about the neutral axis can be
calculated by the following procedure. The cross-section of the profile
x4
is divided into regular shapes, as shown in Figure 58, in which four
rectangular elements are shown. First, the moment of inertia of each
element about its own neutral axis is calculated (using the formula
for a simple strip). Next, the composite neutral axis is determined, the
individual moments of inertia referred to it and then the overall moment 4 y4
determined.
dc
d1
d4
Neutral axis Neutral axis

element 1 element 4
Reference axis
Figure 58 – Calculation of moment of inertia of a complex shape

The moment of inertia of each segment is calculated by:
where the symbols are defined in Figure 59.
b b
Neutral axis
Neutral axis
Figure 59 – Parameters used in the calculation of moment of inertia of each element
The position of the composite neutral axis is calculated as follows:
The individual moments of inertia are referred to the composite neutral axis by:
and, in total
where, for di > dc
else

6.0 Jointing of Copper Busbars
David Chapman
6.1 Introduction
Busbar joints are of two types; linear joints required to assemble manageable lengths into the installation and T-joints required to make tap-off
connections. Joints need to be mechanically strong, resistant to environmental effects and have a low resistance that can be maintained over the
load cycle and throughout the life of the joint.
6.2 Busbar Jointing Methods
Efficient joints in copper busbar conductors can be made very simply by bolting, clamping, riveting, soldering or welding. Bolting and clamping are
used extensively on-site. Shaped busbars may be prefabricated by using friction stir welding.
Bolted joints are formed by overlapping the bars and bolting through the overlap area. They are compact, reliable and versatile but have the
disadvantage that holes must be drilled or punched through the conductors, causing some distortion of the current flow in the bar. Bolted joints
also tend to have a less uniform contact pressure than those made by clamping but, despite these issues, bolted joints are very commonly used and
have proven to be reliable. They can be assembled on-site without difficulty.
Nut
Flat washer
Flat washer
Bolt
Figure 60 - A typical bolted joint
Clamped joints are formed by overlapping the bars and applying an external clamp around the overlap. Since there are no bolt holes, the current
flow is not disturbed resulting in lower joint resistance. The extra mass at the joint helps to reduce temperature excursions under cyclic loads.
Well-designed clamps give an even contact pressure and are easy to assemble, but take up more space than a bolted joint and are more expensive
to manufacture.
Nut Clamp plate
Flat washer
Flat washer
Bolt
Figure 61 - A simple clamped joint
Riveted joints are similar to bolted joints. They can be efficient if well made; it is difficult to control the contact pressure. They cannot easily be
dismantled or tightened in service and they are difficult to install.

Rivets
Washers
Figure 62 - A riveted joint
Soldered or brazed joints are rarely used for busbars unless they are reinforced with bolts or clamps since heating under short-circuit conditions
can make them both mechanically and electrically unsound.
Solder layer
Figure 63 - A soldered joint
Welded joints are made by butting the ends of the bars and welding. They are compact and have the advantage that the current-carrying
capacity is unimpaired, as the joint is effectively a continuous copper conductor. However, it may not be safe or desirable to make welded joints in
situ.
Welding of copper is discussed in Copper Development Association Publication 98, Cost-Effective Manufacturing: Joining of Copper and Copper
Alloys.
Figure 64 - A welded joint
The following sections apply to bolted and clamped joints.
6.3 Joint Resistance
In principle, a clamped or bolted joint is made by bringing together two flat surfaces under controlled (and maintained) pressure, as shown in
Figure 65.
Figure 65 – An overlapped joint
The resistance of a joint is mainly dependent on two factors:
1. The streamline effect or spreading resistance, Rs , due to the diversion of the current flow through the joint
2. The contact resistance or interface resistance of the joint, Ri .
The total joint resistance, Rj , is given by:
Rj = Rs + Ri
This applies specifically to direct current applications. Where alternating currents are flowing, the changes in resistance due to skin and proximity
effects in the joint zone must also be taken into account.

6.3.1 Streamline Effect
When current flows through a joint formed by two overlapping conductors, the lines of current flow are distorted and the effective resistance of
the joint is increased since current flows only through a portion of the material.
Provided that the width of both bars is the same, the streamline effect is dependent only on the ratio of the length of the overlap to the thickness
of the bars and not on the width. This is shown in Figure 66.
2.0
l
1.8 b
1.6
1.4
Resistance ratio, e
1.2
1.0
0.8
0.6
0.4
1 2 3 4 5 6 7 8 9 10
Overlap ratio, l/b
Figure 66 - Streamline effect in overlapped joints
The current density in the direction perpendicular to the bar, i.e. as current transfers from one bar to the other, is highly non-uniform and is
concentrated around the edges.
The resistance ratio e in Figure 66 is the ratio of the resistance of a joint due to streamline effect Rs , to the resistance of an equal length of single
conductor Rb , i.e.
where:
a is the width of bar, mm
b is the thickness of bar, mm
l is the length of overlap, mm
r is the resistivity of the conductor, mW mm
Rs is the resistance of overlap section in mW (to which contact resistance must be added)

hence:
From the graph it can be seen that the streamline effect falls very rapidly for l/b ratios up to two, and then very much more slowly for values up to
ten. This means that, in most cases, the streamline effect has a limited effect as the overlap is often much greater than five times the thickness in
order to allow space for bolting or clamping. There is no advantage in allowing very long overlaps; it is only necessary to allow enough space to
accommodate sufficient bolts to achieve the required contact pressure.
In the case of bolted joints, the bolt holes also reduce efficiency due to the streamline effect. The resistance ratio of a bolted overlap section can be
estimated by:
where:
d is the diameter of the holes
n is the number of holes across the width of the bars.
It follows that holes should be placed in-line along the length of the joint, as shown in Figure 67; offsetting the holes increases the resistance by
increasing the disturbance of the current flow. In Figure 67a, the value of n is 2 while in Figure 67b the value of n is 4.
a) Correct bolt placement
b) Incorrect bolt placement
Figure 67 - Bolt placement in overlapped joints
It has been found that the distortion effect in the tap-off of a T-joint is about the same as that in a straight joint. Note that the current flow in the
straight bar is disturbed by the presence of bolt holes.
It has been shown that the current distortion is reduced if the ends of the bars are angled at less than 45 degrees, as shown in Figure 68. The
initial joint resistance is reduced by 15%. Because the current flow is more uniform, the development of localised hot spots is reduced, leading to a
factor of 1.3 to 1.5 reduction in the rate of increase in resistance under current cycling.
Figure 68 – Overlap joint between bars with angled ends

6.3.2 Contact Resistance
There are two main factors that affect the actual interface resistance of the surfaces:
1. The condition of the surfaces

2. The total applied pressure.
6.3.2.1 Condition of Contact Surfaces
In practice, an electrical contact between the solids is formed only at discrete areas within the contact interface and these areas (known as
‘a-spots’) are the only current conducting paths. The a-spots typically occupy an area of the order of 1% of the overlap area.
Obviously, the larger the number of a-spots, the more uniform the current distribution across the joint area will be. This can be encouraged by
ensuring that the surfaces of the conductors are flat and roughened (which removes the oxide layer and produces a large number of asperities)
immediately before assembly. As the contact pressure is increased, the higher peaks make contact, disrupt any remaining surface oxide and form
metal to metal contact.
In some areas an oxide film may remain. Copper oxide films on copper form relatively slowly and are semiconducting because copper ions
diffuse into the oxide layer. When copper oxide films are compressed between two copper surfaces, diffusion can take place in both directions
so conduction takes place in both directions. This is very different from aluminium, where the oxide is a very good insulator and forms within
microseconds of exposure to air.
Since the area of each a-spot contact is small, the current density is high, leading to higher voltage drop and local heating. In a well-made
joint this heat is quickly dissipated into the mass of the conductor and the temperature of the interface will be only slightly above that of the
bulk material. However, if the contact pressure is too low and the joint has deteriorated, local over-heating may be enough to induce basic
metallurgical changes including softening and melting of the material at the a-spot. At first sight this may appear to be advantageous, however,
as the joint cools the material contracts and fractures and is subsequently liable to oxidise.
Since elevated temperature is the first symptom of joint failure, maintenance procedures should be established to monitor the temperature of
joints with respect to that of nearby bar using thermal imaging. If, under similar load conditions, the differential temperature increases, it may
be a sign of early joint degradation. As a first step, more intensive monitoring should be undertaken and, if the trend continues, remedial action
taken.
It is not normally recommended that the surfaces of copper-to-copper joints are plated unless required by environmental considerations. In fact,
plating may reduce the stability of the joint because, as soft materials, the plating may flow at elevated temperatures leading to reduced contact
pressure.
However, to ensure a long service life, a contact aid compound is recommended to fill the voids in the contact area and prevent oxidation or
corrosion. Many proprietary compounds are available or, if none are available, petroleum jelly or, for higher temperatures, silicone vacuum grease
may be used.
6.3.2.2 Effect of Pressure on Contact Resistance
Joint resistance normally decreases with an increase in the size and number of bolts used. Bolt sizes usually vary from M6 to M20 with either four
or six bolts being used. The appropriate torque for each bolt size depends on the bolt material and the maximum operating temperature expected.
Contact resistance falls rapidly with increasing pressure, as shown in Figure 69, but above a pressure of about 30 N/mm² there is little further
improvement. In most cases it is not advisable to use contact pressures of less than 7 N/mm², with pressures above 10 N/mm² being preferred.
The contact resistance for a joint of a particular overlap area is obtained from Figure 69 by dividing the contact resistance for 1 mm2 by the overlap
area in mm2.

5000
4000
Resistance (µΩ) for 1mm2
3000
2000
1000
0 10 20 30 40 50 60
Pressure N/mm2
Figure 69 – The effect of pressure on the contact resistance of a joint
Contact pressure for both bolted and clamped joints is normally applied by tensioning one or more bolts. For bolted joints, the pressure is applied
around the bolt holes, so using more bolts will result in a more even pressure distribution. Large, thick washers can be used to spread the load. For
clamped joints, the load transferred from the bolts, which are outside the width of the conductors, depends on the rigidity of the clamps. Where
the clamps are narrow, the pressure distribution provided by clamps can be quite uniform but, for wider conductors, the very rigid clamps required
may be impractically large.
In everyday practice, contact pressure is impossible to measure and has to be inferred from the torque applied to the bolts from the following
equation:
T = KFD
where:
T is the tightening torque (Nm)
K is a constant, often referred to as the ‘nut factor’ – see Table 19
F is the force (kN)
D is the nominal bolt diameter (mm) - see Table 21.
The ‘nut factor’ depends on a number of factors including the coefficient of friction, the surface finish and state of lubrication of the threads and
other bearing surfaces. Table 19 gives typical nut factors for different states of lubrication.

Table 19 - Nut Factors for Different States of Lubrication
Bolt Lubrication Nut Factor
Dry 0.20 – 0.22
Contact aid compound 0.19 – 0.21
Boundary lubricant (Mo2S) 0.15 – 0.16
It is important to control the rate of applying the torque as well as the final value; bolts should be gradually tightened in rotation.
The correct tightening torque must be carefully determined to provide sufficient initial contact pressure at ambient temperature while not
exceeding the proof or yield stress of the bolt material over the working temperature range of the joint. Differential expansion between the bolt
and bar materials results in an increase or decrease in bolt tension (and therefore contact pressure) as temperature changes. Galvanised steel
bolts are often used with copper busbars but copper alloy bolts, e.g. aluminium bronze (CW307G), are preferred because their coefficients of
expansion closely match that of copper, resulting in a more stable contact pressure. Copper alloy bolts also have the advantage that the possibility
of dissimilar metal corrosion is avoided and are also to be preferred where high magnetic fields are expected. Because these alloys do not have
an easily discernible yield stress, however, care has to be taken not to exceed the correct tightening torque and the bolt stress over the working
temperature range should not exceed 95% of proof strength.
Because of the strength of copper, deformation of the conductor under the pressure of the joint is not normally a consideration.
Table 20 shows the proof strength and coefficient of thermal expansion of some typical bolt materials compared to copper. It is clear from this
table that the choice of bolt material will determine the thermal stability of the joint.
Table 20 – Proof Strength and Coefficient of Thermal Expansion for Copper and Typical Bolt Materials
Material Proof Strength Coefficient of Expansion
MPa Per Degree C
Copper (reference) Fully annealed - 50 16.5 x 10-6
Full temper - 340
High tensile steel 700 11.1 x 10-6
Stainless steel 316 414 15.9 x 10-6
Aluminium bronze CW307G 400 16.2 x 10-6
Stainless steel 304 207 17.2·x 10-6
Silicon bronze C651000 365 17.8·x 10-6
The increase in force (Fsupp) due to an increase in temperature (DT)is given by:
where:
aa is the coefficient of expansion of the busbar conductor
ab is the coefficient of expansion of the bolt
Ab is the bolt cross-sectional area
Eb is the elastic modulus of the bolt
Ea is the elastic modulus of the busbar
t is the thickness of the washer
a is the thickness of the busbar
Aw is the apparent area under the washer
Aa is the apparent area of the joint overlap.

The change in bolt stress is proportional to (αa -αb ) so, for a joint made with high tensile steel bolt, the tension will increase considerably
(Da = 5.5 x 10-6) while, if the joint were made with CW307G bolts, the tension will reduce only slightly (Da = -0.3 x 10-6).
In any case, the joint must be designed so that the maximum tension in the bolts, at any temperature within the working range, must be less than
95% of yield stress to avoid the risk of plastic deformation which would ultimately lead to loosening of the joint and failure.
The stress in the bolt is calculated using the Tensile Stress Area (see Table 21), not the nominal area.
Table 21 – Typical Thread Characteristics

Size Designation Nominal (Major) Nominal Pitch Pitch Minor Tensile
Diameter Shank Area (mm per thread) Diameter Diameter Area Stress Area
Dn An p dp As Ats
M6 6.00 28.274 1.000 5.3505 17.894 20.123

M8 8.00 50.265 1.250 7.1881 32.841 36.609
M10 10.00 78.540 1.500 9.0257 52.292 57.990
M12 12.00 113.10 1.750 10.863 76.247 84.267
M14 14.00 153.94 2.000 12.701 104.71 115.44
M16 16.00 201.06 2.000 14.701 144.12 156.67
M20 20.00 314.16 2.500 18.376 225.19 244.79
M22 22.00 380.13 2.500 20.376 281.53 303.40
M24 24.00 452.39 3.000 22.051 324.27 352.50
M27 27.00 572.56 3.000 25.051 427.09 459.41
M30 30.00 706.86 3.500 27.727 518.99 560.59
M33 33.00 855.30 3.500 30.727 647.19 693.55
M36 36.00 1017.9 4.000 33.402 759.28 816.72
If the design requirement is such that high tensile steel bolts must be used, the incremental force, Fsupp , may be so high as to exceed 95% of the
proof stress of the bolts. In these cases, disc-spring, or Belleville, washers must be used, as shown in Figure 70. The height and the spring rate of
the washer are selected to reduce the value of Fsupp according to the following equation:
where:
h is the overall height of the disc-spring washer
K is the spring rate of the disc-spring washer.
In practice, the joint would be assembled normally with the required torque for the required contact pressure. In service, as the joint temperature
rises, the spring is compressed, limiting the increase in bolt tension to a safe value.

Belleville washer
Flat washers
Cu Cu
Cu Cu
Steel bolt Bronze bolt

Figure 70 - Possible bolting techniques for copper busbars
Changing the design of a bolted joint, for example by introducing a longitudinal slot (see Figure 71), can reduce the contact resistance by 30
to 40%. The reduction in resistance is attributed to an improvement in the uniformity of contact pressure in each ‘leg’ of the joint leading to
increased contact area.
Upper bar Lower bar
a)
Slot cut in
both bars
Upper bar Lower bar
b)
Figure 71 – Joint with a longitudinal slot
6.4 Bolting Arrangements

Although the required bolting arrangements should always be calculated for the circumstances of the installation, many sources give
recommendations. Those given in Table 22 have been used for many years and are given here as a rough guide.
The recommended torque settings may be used for high-tensile steel (8.8) or aluminium bronze (CW307G, formerly C104) fasteners with
unlubricated threads of normal surface finish.

Table 22 - Typical Busbar Bolting Arrangements (Single Face Overlap)
Bar Width Joint Joint Area Number of Metric Bolt Bolt Hole Size Washer Washer
mm Overlap mm² Bolts (Coarse Torque mm Diameter Thickness
mm Thread) Nm mm mm
16 32 512 2 M6 7.2 7 14 1.8
20 40 800 2 M6 7.2 7 14 1.8
25 60 1500 2 M8 17 10 21 2.0
30 60 1800 2 M8 17 10 21 2.0
40 70 2800 2 M10 28 11.5 24 2.2
50 70 3500 2 M12 45 14 28 2.7
60 60 3600 4 M10 28 11.5 24 2.2
80 80 6400 4 M12 45 14 28 2.7
100 100 10000 5 M12 45 15 28 2.7
120 120 14400 5 M12 45 15 28 2.7
160 160 25600 6 M16 91 20 28 2.7
200 200 40000 8 M16 91 20 28 2.7
6.4.1 Joint Efficiency
The efficiency of a joint may be measured in terms of the ratio of the resistance of the portion of the conductor comprising the joint to that of an
equal length of straight conductor. It is possible to make joints with an efficiency of greater than 100% - i.e. the resistance of the joint is lower
than that of an equivalent section of bar.
In terms of a complete busbar system, the proportion affected by joints is relatively small so that any inefficiency of the joints has only a small
impact on the overall performance. However, joint inefficiency is important in two respects:
1. A joint with an efficiency of less than 100%, having a higher resistance, will run at a higher working temperature and experience
greater temperature excursions than the normal bar. This could have an effect on the longevity of the joint and require more
frequent maintenance.
2. In switchgear cabinets there will be many joints close together; less efficient joints will lead to excess heating and higher voltage
drops.
The resistance of a joint, as already mentioned, is made up of two parts, one due to the distortion of lines of current flow and the other to contact
resistance. The resistance due to the streamline effect at an overlapped joint is given by:
where:
e is the resistance ratio obtained from Figure 66
a is the width of bar, mm
b is the thickness of bar, mm
l is the length of overlap, mm
r is the resistivity of the conductor, mW mm (17.24 for 100% IACS copper)
d is the diameter of the bolt holes, mm
n is the number of holes across the width of the bars. For clamped joints, the value of n is zero.

The contact resistance, Ri , of the joint is:
where Y is the contact resistance of a unit area, obtained from Figure 69.
The total joint resistance, Rj , is:
Since the resistance, Rb , of an equal length of straight conductor is given by:
the efficiency of the joint is:
From this equation it is apparent that the most important factor is the reduction in cross-section due to the bolt holes, i.e. the term nd.
Taking the parameters for a 50 mm wide busbar from Table 22, the contact force, F, is given by (noting that there are two 12 mm bolts):
The area of the joint is 3500 mm2, so the pressure is:
From Figure 69, a Y value of 3000 mW is obtained.
For 10mm thick bar, the overlap to thickness ratio is 7 so that, from Figure 66, e = 0.55.
Substituting,

This joint has a resistance of 1.12 times that of a 70 mm length of 50 mm x 10 mm copper bar, i.e. equivalent to 78.4 mm of bar. The joint
temperature will be slightly higher than that of the surrounding bar.
If the joint were redesigned with an overlap of 90 mm, using three in-line bolts at the same torque, the joint efficiency becomes:
The area of the joint is 4500 mm2, so the pressure is:
From Figure 69, a Y value of 2600 mW is obtained.
For 10 mm thick bar, the overlap to thickness ratio is 9 so that, from Figure 66, e = 0.55.
Substituting,
In this case, the joint has a resistance of 0.91 times that of a 90 mm length of 50 mm x 10 mm copper bar, i.e. equivalent to 82 mm of the bar. This
joint will run at a slightly lower temperature than the surrounding bar.
6.5 Clamped Joints
The design criteria for bolted joints apply in principle also to clamped joints. However, some aspects require particular attention:
• The clamping plates must be designed to be rigid enough to transfer the pressure without flexing. Often, ribbed castings are used for
this purpose.
• The bolts which provide the joint pressure are at the periphery of the joint and will run at a temperature somewhat below that of the
bar. In some ‘wrap around’ lamp designs, the bolts will also be physically shorter than the thickness of the stacked bars. The bolts
will therefore expand less, and joint pressure may rise excessively.
6.6 Degradation Mechanisms
The deterioration of a connector proceeds slowly at a rate determined by the nature of different processes operating in the contact zone and in
the environment. This initial stage persists for a long time without causing any noticeable changes because it is an intrinsic property of clusters
of a-spots that their overall constriction resistance is not sensitive to small changes in their size. However, when the contact resistance increases
sufficiently to raise the local temperature, a self-accelerating deterioration resulting from the interaction of thermal, chemical, mechanical and
electrical processes will be triggered, and the contact resistance will rise abruptly. Hence, no deterioration will be noticeable until the final stages of
the connector life.
6.6.1 Oxidation
Oxidation of the metal–metal contacts within the contact interface is widely accepted as the most serious degradation mechanism occurring in
mechanical connectors. Copper is not very active chemically and oxidises very slowly in air at ordinary temperatures. As mentioned earlier (see
‘6.3.2.1 Condition of Contact Surfaces’), cleaning and roughening the joint surfaces prior to assembly and the use of a contact aid will prevent
oxidation.

6.6.2 Corrosion
Corrosion is a chemical or electrochemical reaction between a metallic component and the surrounding environment. It begins at an exposed metal
surface with the formation of a corrosion product layer and continues as long as reactants can diffuse through the layer and sustain the reaction.
The composition and characteristics of the corrosion product layer can significantly influence the corrosion rate.
Busbars are potentially affected by atmospheric, localised, crevice, pitting and galvanic corrosion. The most important factor in all these corrosion
mechanisms is the presence of water. In the presence of a sulphur-bearing atmosphere, tarnishing of the copper surface occurs because of
sulphide formation from hydrogen sulphide in the atmosphere. The growth of tarnished film is strongly dependent on the humidity, which can
reduce it if a low sulphide concentration prevails or increase it if sulphide concentration is high.
6.6.3 Fretting
Fretting is the accelerated surface damage occurring at the interface of contacting materials subjected to small oscillatory movements.
Experimental evidence shows that amplitudes of <100 nm are sufficient to produce fretting.
There is still no complete unanimity on the mechanisms of fretting, specifically with regard to the relative importance of the processes involved.
Nevertheless, based on the existing knowledge of the phenomenon, it can be safely assumed that the following processes are present:
1. Disruption of oxide film on the surface by the mechanical action exposes clean and strained metal which will react with the
environment and rapidly oxidise
2. The removal of material from the surfaces by adhesion wear, delamination or by shearing the microwelds formed between the
asperities of the contacting surfaces when the contact was made
3. Oxidation of the wear debris and formation of hard abrasive particles that will continue to damage the surfaces by plowing
4. Formation of a thick insulating layer of oxides and wear debris (a third body) between the contacting surfaces.
The oscillatory movement of the contacting members can be produced by mechanical vibrations, differential thermal expansion, load relaxation,
and by junction heating as the load is cycled. Because fretting is concerned with slip amplitudes not greater than 125 µm movement, it is
ineffective in clearing away the wear debris and accumulated oxides, and a highly localised, thick insulating layer is formed in the contact zone,
leading to a dramatic increase in contact resistance and, subsequently, to virtual open circuits.
6.6.4 Creep and Stress Relaxation
Creep, or cold flow, occurs when metal is subjected to a constant external force over a period of time. The rate of creep depends on stress and
temperature and is higher for aluminum than for copper. Stress relaxation also depends on time, temperature, and stress but, unlike creep, is not
accompanied by dimensional changes. It occurs at high stress levels and is evidenced by a reduction in the contact pressure due to changes in
metallurgical structure. The change from elastic to plastic strain has the effect of significantly reducing the residual contact pressure in the joints,
resulting in increased contact resistance, possibly to the point of failure.
6.6.5 Thermal Expansion
The effect of temperature variation on contact pressure has already been discussed. Longitudinal expansion is also important since it can lead
to slip in the joint followed by loosening. It is important that long bars are provided with a flexible element so that movement can take place
elsewhere.
6.7 Conclusion
The quality of busbar joints is crucial to the long term reliability of a busbar system. It is important to take care over the choice of joint design, the
tightening torques, bolt types and the effect of temperature to ensure reliability. In-service maintenance should ideally include thermal imaging of
joints so that any problems can be found before failure occurs.

Annex: Coatings
David Chapman
A1.0 Introduction
Although coatings have been applied to busbars for several perceived reasons, there are in reality few genuine engineering needs to do so. Since
coatings are expensive to apply and may require additional maintenance, they should be applied only when truly necessary.
A2.0 Reasons for Coating
A2.1 Coating to Provide Electrical Insulation
The primary reason to insulate busbars is to provide protection against electric shock in the event of direct contact.
Two significant problems arise. Firstly, the electrical insulating material is also a thermal insulator, so the bars will run at a higher temperature or
will be limited to a lower working current. As a result, bars with a larger cross-sectional area will be needed. Although coating can reduce the
current-carrying capacity, the short circuit capacity is not affected because the SCC is always calculated assuming that no heat is lost.
Secondly, it is difficult to apply insulation to the required degree of integrity during installation. Even when using pre-insulated bars, additional
insulation has to be applied at joints and connections, requiring clean working conditions and skilled operatives.
It is preferable to avoid the need for direct insulation, ensuring safety by mounting non-insulated bars out of reach or behind appropriate barriers,
together with the provision of the appropriate fixed safety warning labels.
A2.2 Coating to Inhibit Corrosion
In most circumstances, copper is highly resistant to corrosion and does not normally need additional protection. However, in environments
containing ammonia, sulphur and chlorine compounds, especially where the humidity may be high, protection is required.
A2.2.1 Metal Coatings
Historically, tin, tin-lead alloys, nickel or silver coatings have been used.
• Tin and tin-lead alloys: Pure tin coatings should be avoided because they tend to form whiskers, which can cause transient short
circuits. Tin-lead alloys are rarely used now due to environmental concerns. Tin-containing coatings are relatively expensive.
• Nickel is the preferred material since it is relatively cheap, is durable (except at high humidity) and provides a harder surface than
the alternatives. Nickel oxide is tough, so high pressure is required at joints to ensure reliability. Nickel may corrode significantly in
contact with more noble metals such as gold.
• Silver is expensive but is very effective, except where sulphur compounds are present.
Protection will only be effective if it is continuous over the whole surface, including joints and connections. This means that metal coatings, which
are always relatively thin (2 to 5 µm), are free of pinholes and other defects which would allow corrosion of the conductor material. Plated bars
must be carefully protected before and during installation to avoid damage.
Metal coatings will not significantly affect the current rating or working temperature of the busbars.
A2.2.2 Non-Metallic Coatings
In very corrosive environments, busbars may need additional protection, using similar materials and techniques as would be used for electrical
insulation. Obviously, the current rating of the busbars will be reduced due to the additional thermal insulation.
A2.3 Coating to Increase Current Rating
Matt black surfaces have higher emissivity than light-coloured polished surfaces, so it has often been suggested that matt black paint should be
applied to busbars to improve cooling by radiation.
In practice, the potential improvement is small and sometimes negative. The natural emissivity of a copper bar that has been in use for even a
short time will be above 0.5 and will rise over time to 0.7 or above, while a well applied matt black paint will have an emissivity of 0.9 to 0.95. As
a result, an increase of about 30-35% in cooling by radiation may be expected. Even at a temperature rise of 60°C, this corresponds to an overall

increase in cooling of barely 12% and an increase in current-carrying capacity of less than 6%. At lower temperature rises, or where bars are
facing each other, the radiation loss is lower and the ‘gain’ due to painting is much lower.
On the negative side, the paint layer acts as an insulator, reducing the efficiency of the convection process, perhaps by as much or more than the
increase in radiation efficiency. In general, painting will give little increase, and quite possibly a reduction, in the current-carrying capacity of a
busbar for a given working temperature.
Painting may be worthwhile for very wide bars (where convection is less effective) operating with high temperature rises (where radiation is more
effective).
A2.4 Coating for Cosmetic Purposes
It is inevitable that fingermarks will occur on busbars during assembly and, in the first few weeks of operation, the natural grease in the marks
darkens much more quickly than the surrounding copper. In some countries this is thought to be unsightly and the busbars are painted in
order to hide the marks. There is no technical merit in painting busbars – it is an additional cost, increases the working temperature and makes
maintenance more difficult. On the other hand, painting may be required to meet the demands of the local market.
A2.5 Coating to Improve Joint Performance
Plating of the contact area of joints is not recommended unless it is really necessary to combat corrosion.
As described in Section 6.0 ‘Jointing of Copper Busbars’, a contact is formed by a large number of cold welds between the surfaces of the two
conductors. If plated conductors are used, these cold welds cannot form and the softer plating material tends to flow, reducing contact pressure
and reducing joint reliability.
A3.0 Methods of Coating
Coating of high conductivity copper is normally limited to the application of insulation to conductors in very long lengths, a method not well
suited to the relatively short lengths used for busbars. In other, non-electrical applications, protection is usually achieved by the choice of a copper
alloy with intrinsic resistance to the application environment, so there is a limited requirement for other types of coating. As a result, most direct
experience with local coatings is with steel, primarily in the oil, gas and chemical industries. For use with HC copper, it is important to establish
that the chosen coating will be compatible in the long-term. Many coatings methods will require some form of pre-treatment to ensure good
adhesion and this should also be included in the compatibility assessment.
Whatever type of coating is used, it is important that the coating is continuous and free of pinholes and voids. To avoid thinning of coatings at
edges, conductors with rounded, rather than sharp, edges should be used to reduce mechanical and electrical stresses.
Coatings can be applied either during manufacture of the conductors or during on-site assembly. In practice, there will probably always be a need
to apply at least some coating on-site at joints and connection points, requiring an assessment of the mutual compatibility with the factory-
applied coating. The following sections give some examples, but many other techniques are available.
A3.1 Factory Application Methods
Factory application of coatings during manufacture of the conductor allows a much greater choice of materials and processes and should ensure
a more consistent quality of coating at lower cost than on-site coating. However, it is necessary to locally coat the busbar joints in a compatible
manner during on-site installation.
A3.1.1 Extrusion
Extrusion is viable only for bars supplied in very long lengths, typically relatively small diameter round rods or tubes. It is cost-effective and is
compatible with traditional electrical insulation materials such as PVC and XLPE.

A3.1.2 Powder Coating
As the name suggests, these coatings are applied in powder form, by electrostatic spraying or in a fluidised bed, and then cured either by heat, UV
light or some other means. Nylon and epoxy powders are often used. For many materials, the heat curing temperature and time requirements will
not result in the annealing of the copper conductor and will tolerate temperatures of 150°C.
A3.2 On-Site Application Methods
A3.2.1 Heat-Shrinkable Sleeve
Heat-shrinkable sleeve is normally made from cross-linked polyolefin. It will shrink to about 50% of the supplied perimeter temperatures
above about 70°C, with a consequent increase in thickness. It is easy to apply on-site using an adequately powered hot air gun but can be time
consuming. It can typically tolerate working temperatures up to 90°C.
A3.2.2 Painting
Painting, by brush or spray usually requires a primer coat. Note that paint cannot be expected to provide adequate shock protection because the
integrity of the coating cannot be guaranteed.
A3.2.2.1 Alkyd Paints
Alkyd paints are traditional oil-based paints and are suitable for use at temperatures up to about 120°C. They have a limited lifetime of a few years.
A3.2.2.2 Two-Part Epoxy
Two-part epoxy paints are applied immediately after mixing and cure at normal ambient temperatures. The cured paint has good abrasion
resistance with a lifetime of over ten years but it is not resistant to UV. It is typically suitable for use at temperatures up to about 150°C.
A4.0 Inspection and Maintenance
As with uncoated bars, routine inspection and maintenance is required. The biggest potential problem is the loosening of joints, which is normally
detected as a local rise in temperature, detected during regular inspection with a thermal imaging camera.
In the case of coated bars, the presence of a coating on the nuts and bolts may initially prevent loosening but, once loosening has occurred, will
certainly make it more difficult for the joint to be correctly re-tensioned. Because the coating thicknesses may vary, especially in the areas of the
joints, it is important to establish the temperature profile in the early life of the installation for future comparison.
It is also necessary to visually inspect the coating for integrity, looking for cracking, especially at joints where different coatings meet or the size
changes, and for disturbance of the surface that might indicate underlying corrosion.

Notes

Notes

Copyright © 2014. Copper Development Association and European Copper Institute
Copper Development Association

5 Grovelands Business Centre
Boundary Way
Hemel Hempstead, HP2 7TE
UK
[email protected]
www.copperalliance.org.uk

Copper For Busbars All Sections

Uploaded by

Copyright:

Available Formats

Copper For Busbars All Sections

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

Copper For Busbars All Sections

Uploaded by

Copyright:

Available Formats

Copper for Busbars

Guidance for Design and Installation

Copper Development Association Publication No 22

European Copper Institute Publication No Cu0201

Revised May 2014

Copper Development Association is a non-trading organisation that promotes

Founded in 1996, ECI is a joint venture between the International Copper

Professor Toby Norris

Cover page picture acknowledgment:

2 | COPPER FOR BUSBARS

1.0 Introduction.................................................6 3.0 Life Cycle Costing..................................... 56

COPPER FOR BUSBARS | 3

5.5 Electrical Design Considerations................................................85

4 | COPPER FOR BUSBARS

COPPER FOR BUSBARS | 5

1.1. About this Guide

• Temperature rise due to energy losses

1.2 Materials for Busbars

1.2.1 Material Requirements

• Low electrical and thermal resistance

6 | COPPER FOR BUSBARS

Table 1 – Properties of Typical Grades of Copper and Aluminium

1.2.2.1 High Conductivity

Table 2 – Properties of 100% IACS Copper

It follows from the first two of these values that:

COPPER FOR BUSBARS | 7

0 0.05 0.10 0.15 0.20 0.25 0.30

Figure 1 – Effect of small concentrations of impurities on the resistivity of copper

8 | COPPER FOR BUSBARS

1.2.2.1.2 Effect of Cold Work on Conductivity

1.2.2.2 Mechanical Strength

COPPER FOR BUSBARS | 9

1.2.2.2.2 Proof Strength

0.2% Proof Strength

10 | COPPER FOR BUSBARS

Annealed (O) 60 HV max

1.2.2.2.4 Resistance to Softening

1.2.2.2.5 Creep Resistance

COPPER FOR BUSBARS | 11

Figure 3 - Typical creep properties of commercially pure copper and aluminium

Table 4 - Comparison of Creep Properties of High Conductivity Copper and Aluminium

1.2.2.2.6 Fatigue Resistance

Table 5 - Comparison of Fatigue Properties of High Conductivity Copper and Aluminium

12 | COPPER FOR BUSBARS

1.2.2.2.7 Bending and Forming

Table 6 – Minimum Bend Radius of High Conductivity Copper

Table 7 - Self-extinguishing Arcs in Copper and Aluminium Busbars

COPPER FOR BUSBARS | 13

1.2.3 Types of High Conductivity Copper Available

1.2.3.2 Oxygen-free High Conductivity Copper, CW008A (C103)

1.2.4 Available Forms

14 | COPPER FOR BUSBARS

2.1 Design Philosophy

2.2 Calculation of Maximum Current-Carrying Capacity

2.2.1 Methods of Heat Loss

COPPER FOR BUSBARS | 15

2.2.1.1 Convection – Natural Air Cooling

For vertical surfaces:

For horizontal surfaces: