OPTIMISING LINE DESIGNS

TUTORIAL FOR TB 638

R. STEPHEN
GUIDE TO OVERHEAD OHL DESIGN TB 638
AORC CONFERENCE BALI
Table of contents

• AC LINE DESIGN
• ELECTRICAL CHARACTERISTICS
• MECHANICAL CHARACTERISTICS
• THERMAL RATING
• PLANNERS REQUIREMENTS
• OBJECTIVE INDICATOR
• DC LINE DESIGN
• ELECTRICAL CHARACTERISTICS
• OBJECTIVE INDICATOR
• COST OF LINES
• EXAMPLES OF OPTIMISATION
 OPTIMISING LINE
DESIGNS
TUTORIAL FOR TB 638
AC LINE DESIGN

GUIDE TO OVERHEAD OHL DESIGN TB 638

What is a power line?
• A device to transmit power over distances.
• Design of the line can be tailor made to meet planner’s requirements.
• Load flow depends on R, X and B values.

4
LINE MODEL

• Reduce R and L (resistance and inductance)

• Increase C (capacitance)

5
Maximise Power Transfer
• Zs is surge impedance
• SIL is the surge impedance loading
• Reduce L and increase C to maximise transfer
L is series
inductance
C is shunt
capacitance

VLL is phase to phase

voltage
6
Determination of R, X and B
• Resistance is a function of
− Conductor construction material and line length
o Lay ratio, ACSR, AAAC, number of layers, diameter of strands.
− Temperature
o The higher the temperature the higher the resistance
− Current and frequency
o Transformer effect
o Eddy currents.

7
Variation with current
1.14

1.12

1.10

1.08

Rac/Rdc
1.06 Penguin(1+6)
107 mm2
Tcond=80°C
1.04

1.02

1.00
0 100 200 300 400 500 600 700 800
I(A)

1.10

1.08

1.06
Rac/Rdc

800 mm2
Falcon(19+54)
1.04
Tcond=80°C

1.02

1.00
0 1000 2000 3000 4000 5000 6000 7000 8000 I(A)

8
Determination of L
• L is a function of Geometric mean radius (GMR) and Geometric mean Distance
(GMD)
• Larger bundle radius and closer phase spacing gives lower L

9
Determination of C
• To increase capacitance
keep phases closer together.

10
Summary
• SIL (L and C) can be varied by
− varying phase spacing closer is better
− Increasing bundle size larger is better
• Resistance can be improved by
− Varying lay ratios per layer (not practical)
− Different materials
− Homogeneous conductors

11
Corona limitations
• Corona can produce audible noise under certain weather conditions. This is very
difficult to mitigate. It is desirable to avoid corona inception.
− Smaller bundle radius will reduce corona up to a point.
− Wider phase spacing better
− More sub conductor bundles better.

12
 AC LINE DESIGN
MECHANICAL CHARACTERISTICS

GUIDE TO OVERHEAD OHL DESIGN TB 638

Mechanical considerations
• Wind load is major consideration in tower design
− Less conductors in the bundle the better
− Less UTS the better (Lighter strain towers). Higher tension to increase height.
• Vibration is a function of tension
− Need to design to the recommended T/m ratio
− Small bundle sizes (twin triple), need more care in vibration damping design
• Galloping mitigation needs to be taken into account
− Includes phase configuration
− Pendulum dampers, interphase spacers

14
Tower top Geometry
• Tower top geometry design applies to conventional towers with metal surrounded
center phase (tower window) as well as three phases in the same window as is
the case with the cross rope suspension.
• The interaction between the phases as well as the shielding angle for the
conductors needs to be carefully designed to ensure optimal insulation co-
ordination providing the required level of reliability.
 AC LINE DESIGN
THERMAL RATING

GUIDE TO OVERHEAD OHL DESIGN TB 638

Thermal loading
• Load at which the safety or annealing criteria of the line is met
− Current at which the height above the ground is in line with OHS act
− Height determined by voltage and flashover distance
• Heat Balance equation used.

17
Joule and magnetic heating
• Joule dependent on AC resistance and temperature
• Magnetic heating dependent on current and conductor layers.

18
Solar heating
• Darkness of conductor
• Diameter of conductor
• Solar radiation

19
Convective cooling
• Dependent on
− the conductor diameter (bigger is better)
− Wind speed
− Temperature difference (bigger is better)
− Roughness

20
Templating temperature
• Templating temperature is the conductor temperature at which the height above
ground is in accordance with the OHS act

Templating Normal Emergency

Conductor temperature deg C Amps Amps
TERN 50 611 814
TERN 60 784 991
TERN 70 911 1138
TERN 80 1023 1257
ZEBRA 50 642 859
ZEBRA 60 818 1049
ZEBRA 70 963 1203
ZEBRA 80 1080 1325

21
SUMMARY

SIL Corona Mechanical Thermal

loading rating
Phase spacing Good Bad Good Neutral
decrease
Large al Bad Bad Good Bad
area/cond (less
conductors)
Diameter Good Bad Bad Neutral
Bundle
increase
High steel Neutral Neutral Bad Good
content

22
 AC LINE DESIGN
PLANNING REQUIREMENTS

GUIDE TO OVERHEAD OHL DESIGN TB 638

Planning requirements
• Planners need to specify the following
− Load transfer requirements
− Load profile daily, annual
− RXB parameters, high and low
− Line Voltage for AC
− Length of line
− Location, start and end points
− Reliability requirements

24
Conductor size and temp

25
Conductor type

TACSR (Thermal resistant)

GTACSR (Thermal resistant
aluminium with gapped core)

XTACSIR (Thermal resistant with

invar core) ACSS (Steel supported) 26
Selection of conductor

27
INSULATOR SELECTION
• Location of conductor bundle determined to meet insulation co-ordination
requirements
• Insulator creepage, dry arching distance, basic insulation level (BIL) determined.
− Depends on pollution levels in the line route
• Insulator configuration depends on tower selection, I or V or other.
• Material depends on pollution, vandalism, maintenance.
• Hardware depends on material (corona rings for composite), live line
requirements
• For cross rope towers I string permits less pollution accretion especially from
birds.
LIGHTNING CONSIDERATIONS
• Shield angle determination depends on tower type.
− Negative shield angles generally give better performance.
• Tower footing resistance needs to be determined and reduced on the line
− “crows foot”, buried earthwire, bentonite mix
• Note towers with large footprints generally give lower tower footing resistances.
− Cross rope suspensions provide excellent results
• If tower footing resistance still high may consider line surge arresters installed on
certain towers.
• Shield or earth wires are now often OPGW.
− Care to be taken for fault current in the earth wire.
 AC LINE DESIGN
OBJECTIVE INDICATOR

GUIDE TO OVERHEAD OHL DESIGN TB 638

Indicator to determine best design
• Need to combine
− SIL
− Thermal rating
− Cost initial and life cycle
o (Taking into account corona, magnetic fields, mech loading etc)

31
FACTOR 1 Life Cycle Cost (k1)
• Covers determination of optimum aluminium area required. (Kelvin’s law)
• Cost of maintenance (estimate)
• Cost of losses – use system losses not line losses. (Due to power flow in
interconnected system)

32
FACTOR 2 THERMAL (k2)
• Cost is directly proportional to Thermal rating
− Higher rating higher initial cost
• A ratio is therefore needed
− Initial cost/MVA thermal (emergency or normal)
• The lower the ratio the better.

33
FACTOR 3 SIL (k3)
• The higher the SIL the higher the initial cost (normally)
• Ratio is therefore also required
− Initial cost/MVA sil

34
COMBINATION OF THE FACTORS
• Objective Matrix method
− Present practice is given 3/10
− 0 or 10 level is determined (normally trial and error) and a linear interpolation
used.
• ATI = w1k1+w2k2+w3k3
− wn are weighting factors

35
CONDUCTOR SELECTION
CONDUCTOR OPTIMISATION

900.00

800.00

700.00

600.00
2 X IEC800
4 X KINGBIRD
LCC(Rmil)

500.00 3 X BERSFORT
3 X TERN
400.00 3 X YEW
3 x Bersfort 4 X TERN

300.00
4 x Tern

200.00
4 x Kingbird

100.00
3 x Tern

0.00
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
Pmean(MVA)

36
TOWER SELECTION

37
EXAMPLE LINE

− Quad “Zebra” guyed Vee tower

− Triple “Bunting” conductor guyed Vee tower
− Quad “Bunting” cross rope suspension (CRS) tower phase spacing of 6,5m.
− Quad “Rail” conductor with a CRS tower with a 6,5m phase spacing.
− Triple “Bittern” conductor with a CRS tower with a 6,5m phase spacing.
− Quad “Boblink” conductor with a CRS tower with a 6,5m phase spacing.
− Triple “Bersfort” conductor with a CRS tower with a 8,2m phase spacing.

38
ATI SCORES
CASE AL AREA DESCRIPTION K1 (LCC) K2 K3
mm2 (CI/MVAth) (CI/MVAsil)

1 1715 4XZEB V 103,53 28,13 7,43

[3,30] [3,07] [3,19]
2 1817 3XBUNT V 84,4 19.48 6,31
[6,25] [5,20] [5,38]
3 2423 4XBUNT 88,36 13.27 7,02
CRS 6.5m [5,64] [6,73] [3,96]
4 1935 4XRAIL 87,76 14.32 5,94
CRS 6.5m [5,73] [6,47] [6,12]
5 1933 3xBIT 82,91 17.86 6,31
CRS 6.5m [6,48] [5,60] [5,38]
6 2901 4Xbob CRS 93,33 17.04 8,06
6.5m [4,87] [5,80 [1,88]
7 2059 3xBers CRS 80,41 16.23 6,30
8.2m [6,81] [6,00] [5,40]

39
 ATI WEIGHTING

CASE W1;W2;W3 W1;W2;W3 W1;W2;W3 W1;W2;W3

0,8;0,1;0,1 0,6;0,2;0,2 0,4;0,3;0,3 0,2;0,4;0,4
1 2,82 [7] 2,89 [7] 2,96 [7] 3,03 [7]
2 5,80 [3] 5,67 [4] 5,55 [4] 5,42 [4]
3 5,23 [5] 5,18 [5] 5,14 [5] 5,09 [5]
4 5,56 [4] 5,74 [3] 5,93 [2] 6,11 [1]
5 6,04 [2] 5,90 [2] 5,77 [3] 5,63 [3]
6 4,33 [6] 4,21 [6] 4,08 [6] 3,96 [6]
7 6,42 [1] 6,24 [1] 6,06 [1] 5,88 [2]

40
FINDINGS/BENEFITS
• Tower, foundation, hardware, electrical designers work together with planners
(iterative process)
• Indicator very sensitive and detects errors rapidly
• Line optimisation is possible looking at overall line design.
• Reliability is assumed constant for options
• Cost system is critical
• Most aspects of the line design are taken into account

41
CONCLUSIONS
• Line design options can be objectively determined
• ATI is a guide from which options can be finalised.
• Alignment with Planners requirements

42
 OPTIMISING LINE
DESIGNS
TUTORIAL FOR TB 638
DC LINE DESIGN

GUIDE TO OVERHEAD OHL DESIGN TB 638

 DC LINE DESIGN
ELECTRICAL CHARACTERISTICS

GUIDE TO OVERHEAD OHL DESIGN TB 638

Power Transfer
V  vd %
2
Pmax =
100R x L
where vd% is the voltage drop expressed as a % of V,

and the equation applies for both 10% and 15% volt drop. {1/sqrt(44) is an approximation of 15/100} and the format
removes the generality of the basic and simple equation.

V=Sending end voltage, pole to ground in kV

Rx=DC Resistance of the conductor in ohm/km

L=Distance in kilometres.
DC resistance
• DC resistance not dependent on current
• Dependent on conductor geometry and conductivity of material.
• Dependent on temperature.
Effect of conductor radius (TB388)

The higher the conductor diameter and the more sub-conductors in the bundle the more resistant the bundle is to
corona and therefore the designer is more able to raise the voltage to ground and hence increase the power capability
of the line.
Corona power loss
 E mas   d  n  HS 
P = 11 + 40 log  + 20 log  + 15 log  − 10 log 
 25   3.05  3  225 

Where P is the corona loss is dB above 1 W/m, Emax is the positive polarity maximum bundle gradient in kV/cm, d is the
sub-conductor diameter in cm, n is the number of sub-conductors in the bundle, H is the average conductor height in m
and S is the pole spacing in m.
INSULATOR CONSIDERATIONS
• Similar requirements to AC as far as tower window design.
• For glass insulators need germanium glass as normal glass will shatter
− Zinc collar also required
• Creepage larger than for AC
• Space charge considerations as well as uneven pollution on insulator to be taken
into account.
• Composite insulators can be used for AC and DC lighter weight often suit long
insulator installation.
• Porcelain disc are also successfully used.
Summary of options

Action Parameter Voltage drop Corona Mechanical loading Thermal rating

+ and - pole spacing Neutral Bad Good Neutral

decrease

Large Al area/cond (less Good Bad Good Bad

conductors)

Diameter bundle Neutral Bad Bad Neutral

increase

High steel content Neutral Neutral Bad Good

Voltage selection (TB 388)
 DC LINE DESIGN
OBJECTIVE INDICATOR

GUIDE TO OVERHEAD OHL DESIGN TB 638

Optimisation process
• Select voltage (TB388)
• Determine range of conductor, bundle diameter and number of sub-conductors as
well as height above ground and pole spacings that will meet corona limitations
and power flow.
• Determine range of tower, foundation and conductor configurations.
• Finalise by further analysis the most suitable tower, foundation, conductor bundle
option. Recheck with power flow requirements, converter cost and technology.
 Objective indicator
IC
ATI DC = w1LCC + w2 IC * Plosscorona + w3
MVAth

ATIdc Appropriate Technology Index for DC lines

LCC is the life cycle cost expressed in terms of a score from 1 to 10 and IC is the initial cost.

Plosscorona is the power loss due to corona.

IC is the initial cost.

MVAthermal is the thermal rating of the line and depends, as in the AC case, to the templating temperature of the line.
Application of HVDC Indicator
Line requirements

• DC voltage (V) 600;700;800 kV

• Number of sub-conductors per pole (N) 4; 5; 6
• Conductor type ACSR
• Line length 1750 km
• Transmitted Power 3000 MW (bipolar)
• Cost of the losses 60 U$/MWh; loss factor =0,5
• Life= 30 years; yearly interest rate= 10%
• Interest during construction 10%; maintenance 2% per year
Alternate designs
Scores
ATIdc weighted scores
 OPTIMISING LINE
DESIGNS
TUTORIAL FOR TB 638
COST OF LINES

GUIDE TO OVERHEAD OHL DESIGN TB 638

Purpose
• The purpose of the questionnaire was to compare the work done by WG09 in
1990 to the latest figures as many component costs may have changed.
• The questionnaire was in two parts, the first to compare component costs of
existing projects, the second to compare costs of an actual line with given
parameters.
• The response was generally poor with around 13 respondents compared with
over 100 in 1990.
North America
Europe
Africa
Asia
Cost per km comparison
Competitive line costs
• From the graphs, if the conductor cost as a percentage of total line cost is greater
than 10% the cost is likely to be relatively low. The range of percentages in the
examples received indicated that the conductor cost for relatively low cost per km
lines, should vary between 10 and 15%.
• In the previous survey it was found that the conductor cost was 32% of the
material cost which was 63%. This results in 20% of the total line cost. It could
be concluded that the cost/km of lines has increased from 1990 to 2014 in real
terms mainly due to cost of labour and environmental issues.
•
• A good target for a competitive cost per km line would be that the conductor cost
should be between 15 and 20% in 2014.
COMPARISON TO 1991
Conductor, shield wire etc % as a function of material costs.
MATERIAL CONSTRUCTION
YEAR CONDUCTORS SHIELD WIRE INSULATORS TOWERS FOUNDATIONS
COSTS COSTS
For all lines and voltages 1991 63.7 36.3 32.7 3.8 8.1 36.2 19.2
2013 42.4 57.6 31.8 2.7 7.6 46.3 11.6
For all lines up to 150kV 1991 64.3 35.7 31.6 4.1 8.8 36.0 19.5
2013 46.4 53.6 28.6 2.0 7.9 49.6 11.9
For all lines over 300kV 1991 62.6 37.4 34.1 3.9 6.9 36.4 18.7
2013 46.8 53.2 35.7 3.0 7.2 42.7 11.4
All single circuit lines 1991 63.6 36.4 33.1 4.2 8.2 35.6 18.8
2013 42.8 57.2 33.4 2.8 6.9 43.7 13.3
All double circuit lines 1991 63.8 36.2 32.0 3.3 7.9 37.1 19.7
2013 31.0 69.0 24.7 2.3 10.6 58.1 4.3
Guyed structure lines 1991 59.6 40.4 32.8 3.2 8.3 36.0 19.8
2013 55.0 45.0 36.5 3.2 6.3 41.3 12.7
Lines with 1 conductor/phase 1991 64.4 35.6 32.2 4.2 8.5 36.3 18.8
2013 38.7 61.3 28.3 2.0 7.8 45.1 16.9
Lines with 2 conductors/phase 1991 64.6 35.4 32.3 4.0 8.1 36.2 19.4
2013 38.0 62.0 32.3 2.3 10.6 48.4 6.3
Lines with 3 conductors/phase 1991 60.8 39.2 35.1 3.7 7.0 40.3 13.8
2013 41.5 58.5 36.6 4.6 6.6 42.6 9.6
Lines with 4 conductors/phase 1991 61.4 38.6 33.4 2.7 7.6 33.4 22.9
2013 56.5 43.5 34.2 3.4 7.9 37.9 16.7
Percentage comparisons with 1991
MATERIAL CONSTRUCTION SHIELD
YEAR COND INSUL TOWERS FNDN
COSTS COSTS WIRE

For all lines and

voltages 1991 63.7 36.3 32.7 3.8 8.1 36.2 19.2

2013 42.4 57.6 31.8 2.7 7.6 46.3 11.6

For all lines up to

150kV 1991 64.3 35.7 31.6 4.1 8.8 36.0 19.5

2013 46.4 53.6 28.6 2.0 7.9 49.6 11.9

For all lines over

300kV 1991 62.6 37.4 34.1 3.9 6.9 36.4 18.7

2013 46.8 53.2 35.7 3.0 7.2 42.7 11.4

All single circuit

lines 1991 63.6 36.4 33.1 4.2 8.2 35.6 18.8

2013 42.8 57.2 33.4 2.8 6.9 43.7 13.3

All double circuit

lines 1991 63.8 36.2 32.0 3.3 7.9 37.1 19.7

2013 31.0 69.0 24.7 2.3 10.6 58.1 4.3

Guyed structure
lines 1991 59.6 40.4 32.8 3.2 8.3 36.0 19.8

2013 55.0 45.0 36.5 3.2 6.3 41.3 12.7

Lines with 1
conductor/phase 1991 64.4 35.6 32.2 4.2 8.5 36.3 18.8

2013 38.7 61.3 28.3 2.0 7.8 45.1 16.9

Lines with 2
conductors/phase 1991 64.6 35.4 32.3 4.0 8.1 36.2 19.4

2013 38.0 62.0 32.3 2.3 10.6 48.4 6.3

Lines with 3
conductors/phase 1991 60.8 39.2 35.1 3.7 7.0 40.3 13.8

2013 41.5 58.5 36.6 4.6 6.6 42.6 9.6

Lines with 4
conductors/phase 1991 61.4 38.6 33.4 2.7 7.6 33.4 22.9

2013 56.5 43.5 34.2 3.4 7.9 37.9 16.7

Summary of Trends
• Material and construction costs – the trend appears to be that the material cost has reduced as a function of total cost with the construction cost being
the more prevalent cost. This appears to be the case over the entire range of lines investigated.
•
• Conductors – in the 2013 cases the steel shield wire is included in the conductor cost. Even with this inclusion, it appears the conductor cost is
generally the same or lower percentage of the total cost as compared to 1991.
•
• Shield wire – this cost is related to the OPGW cost for 2013 and the steel wire cost for 1991. The sample for double circuit and single circuit lines for
2013 is very small and therefore cannot be considered to be representative. However it indicates a similar percentage to the 1991 costs even though
the shield wire is more complicated and expensive in real terms in 2013.
•
• Insulators – the percentage of total cost spent on insulators seem to be slightly lower than in 1991. This could be due to the advent of composite
insulators which have dropped in price considerably over the years as well as glass being more competitive with merger of manufacturers.
•
• Towers – The percentage of the total cost spent on towers seem to be higher than in 1991. This cost includes the erection cost which could indicate
the higher cost of labour which is reflected in the construction cost compared to material cost. As mentioned previously the environmental constraints
on current lines could have resulted in more angle or strain towers as well as more aesthetically pleasing towers such as the Wintrack towers.
•
• Foundations – The percentage of the total cost spent on foundations seems to be lower than is 1991. This may be due to the higher level of
mechanisation and perhaps use of more pile foundations but this is not confirmed.
Tower development

Proposed CRS 6% Existing guyed V

saving on line cost
71
Cost Savings
R450,000

R375,000

46%
R300,000 52% Saving
Saving
R225,000 Misc Costs
Insulation
R150,000 Hardware
Tower Erection
Tower Supply
R75,000
Foundations

R0

72 structures
0-15 degree 15-30 degree structures
705A tower at NETFA

73
 REFERENCES/
ACKNOWLEDGEMENTS

[Stephen 2004] Stephen R. “Use of indicators to optimise design of overhead transmission lines”. Paper 330-1
Shanghai Symposium, Cigré 2003. (Held in Lubljana April 4-6 2004)
[Stephen 2011] Stephen R “Objective detetermination of Optimal power line designs” PhD thesis submitted in 2011
University of Cape Town.
[Muftic]. Muftic D, Bisnath S, Britten A, Cretchley DH, Pillay T, Vajeth R “The Planning design and construction of
overhead power lines” Published by Crown publications 2005 ISBN 9780620330428
[Southwire] Overhead Conductor Manual First edition copyright 1994.
Prof. C.T.Gaunt (UCT) acknowledged for comments and input.
J. Lindquist AC/DC conversion TB 583
Nolasco, Jardini, TB 388

74