0186 Economic Assessment of HVDC Links

186
ECONOMIC ASSESSMENT
OF HVDC LINKS
Working Group
14.20
June 2001
ECONOMIC ASSESSMENT
OF HVDC LINKS
WORKING GROUP
14.20
Members of the Working Group:
J.F. Allaire (Canada) W. Peter Lewis (UK)

M.H. Baker Convenor (UK) W.F. Long (USA)
H. Bilodeau (Canada) C.A.O. Peixoto (Brazil)
J.C. Campos Barros (Brazil) M. A. Reynolds (USA)
S. Chakravorty (India) A. Roy (India)
P.W. Christensen (Denmark) P. Smith (Ireland)
M.G. Dwek (UK) C.V. Thio (Canada)
J. F. Eggleston (Australia) D.R. Torgerson (USA)
G.J. Georgantzis (Greece) R. Witzmann (Germany)
J. Henje (Iceland) R. R. Wood (USA)
N.G. Hingorani (USA) D.A. Woodford (Canada)
Y. Inoue (Japan)
14.20newD
Draft 8. December 2000
TABLE OF CONTENTS
1. INTRODUCTION
2. GENERAL CONSIDERATIONS OF AC AND DC TRANSMISSION
2.1 Objectives
2.2 The Benefits of Interconnection
2.3 Technical Considerations
2.3.1 Long Distance Bulk Power Transmission
2.3.2 Interconnection by AC or HVDC
2.3.3 HVDC Multi-terminal System
3. ASSESSMENT METHODOLOGY
3.1 Identifying alternatives for interconnections

3.2 Technical Assessment
3.2.1 Performance and Cost Analysis
3.2.2 HVDC Electrodes
3.2.3 Losses
3.2.4 Reliability
3.2.5 Spare Parts for HVDC
3.2.6 AC System Strength
3.2.7 Commutation Failure
3.2.8 Subsynchronous Frequency Considerations
3.2.9 Harmonic Modulation
3.3 Tools
4. TRANSMISSION SYSTEM COSTS
4.1 Check List of Cost Areas

4.1.1 Engineering cost
4.1.2 Environmental costs
4.1.3 Implementation costs
4.1.4 Operational Costs
4.2 Station Costs
4.3 Overhead Line Costs
4.4 Stages in Expansion of HVDC Transmission
4.5 Environmental Aspects
5. TRANSMISSION BENEFITS
5.1 Technical Benefits

5.1.1 AC System Support
5.1.2 Limitation of Faults
5.1.3 Limitation of Short Circuit Level
5.1.4 Power Flow Control
5.1.5 Voltage Control
5.2 System Reserve
5.3 Environmental Benefits
6. ECONOMIC ANALYSIS
6.1 Factors to be Considered

6.2 Economic Criteria and Decision Rules
2
14.20newD
6.3 Economic Analysis Methodology

Figure 6.1 Flow Chart of Scheme Development
6.3.1 DCF spreadsheet examples
6.3.2 Cost Streams for Least Cost Ranking
6.3.3 Avoided Cost Concept
6.4 Coping with Uncertainty
6.5 Non-Quantified Benefits and Costs
7. OTHER FACTORS AFFECTING A DECISION
7.1 Risks
7.1.1 Technical Risks
7.1.2 Economic Risks
7.1.3 Social and Political Risks
7.2 Dependency and Interdependency
7.3 Market and Institutional Arrangements
7.4 HVDC as a Barrier
7.5 Supplies to communities en route
8. CONCLUSIONS
ACKNOWLEDGEMENT
APPENDIX I – HVDC CONFIGURATIONS
I 1. Monopolar HVDC Systems

I 2. Bipolar HVDC Systems
I 3. Point-to-Point HVDC Overland Systems
I 4. Point-to-Point HVDC Submarine Systems
I 5. Back-to-Back HVDC Links
I 6. AC and DC in parallel
I 7. Switching possibilities on DC side
I 8. Multi-terminal HVDC Systems
I 9. Intermediate Supplies
APPENDIX II – EXPERIENCE REVIEW
II 1. Cross Channel link

II 2. East and West Europe
II 3. Cahora Bassa
II 4. Quebec - New England
II 5. India
APPENDIX III – MINI-SPEC
III 1. Scheme Type

III 2. AC System for each Terminal
III 3. DC System
III 4. Generators
III 5. Auxiliary Supplies
III 6. Controls & Telecommunications
III 7. Sites
III 8. Site Conditions
REFERENCES
3
14.20newD
1. INTRODUCTION
High Voltage Direct Current (HVDC) has been a viable technique since the 1960s [1] and has
played a vital part in both long distance transmission and in the interconnection of systems.
Its relative economic position has been much discussed over 30 years [2,3,4,5,6] and this
report is both an update of those considerations and a review of the present position.
This report reviews many aspects of transmission development that are common to both AC
and DC transmission, with a concentration on the specific aspects of DC as they give
advantage or extra dimension to the long distance transmission of electrical power.
Interconnection of high voltage transmission (grid) networks can enable electricity to be

provided to communities more reliably and at a lower cost. Pooling of generation capability
and the opportunity to utilise more diverse primary energy resources can be harnessed for
the more economic production of electrical energy. The larger market achieved via
interconnection also enables economies of scale to be realised in the operation of power
plants and in accommodating demand growth. Interconnection facilitates the introduction of
competition into electricity supply.
It is self evident for successful interconnection that the benefits and costs must be
distributed so that all participants gain over the alternative of individual power development
and that users are presented with a supply that is better (e.g. more reliable) or cheaper or
both. The administrative and managerial burdens should be shared, as well as the political
risks that invariably accompany joint developments. It follows that strategies for
interconnection and electricity trade should be based on a careful review of how cooperation
between countries and utilities can develop, allowing for different degrees of integration
according to the individual needs and aspirations of participants. Some implications of this
are included herein.
Chapter 3 discusses the technical features particular to DC, Chapter 4 the costs and Chapter
5 the benefits to be expected. A method of analysing the economics of inter-connection
against continuing separate development is given in Chapter 6, with some non-quantifiable
factors relevant to the decision in Chapter 7.
Appendix I details alternative configurations of DC transmission, Appendix II some

experience from different parts of the world. Appendix III offers a trial mini-specification to
collect further data. The references are a sample from a large quantity of available
literature.
2 GENERAL CONSIDERATIONS OF AC & DC TRANSMISSION
2.1 Objectives
A transmission system needs to fit the demands expected of it with a minimal impact on the
environment. Transmission usually, but not always, has spare capacity because that can
readily be built into the original design. A system is spoken of as having (n-1) capacity, that
is it can supply rated power if any one single element is out of service. Some systems have
(n-2) capacity; even if a double-circuit line suffers a tower failure, the system can still deliver
full power. Some HVDC links consist of a single unit, where a single failure can interrupt
power flow, but reserves are held elsewhere in the system as a whole.
4
14.20newD
Transmission links between systems not before synchronised can be AC or DC. AC demands
that they become synchronised. To determine if this is feasible, it is necessary to consider
• If the submarine cable length is not above the cable’s capacity to carry its own
charging current;
• that the link is able to maintain the synchronism of the two systems under all but
extreme conditions; and
• that it is feasible to arrange generation and frequency control in the joint system
on a common basis.
The HVDC link avoids these limits and has the following advantages:
• no technical limit to the length of a submarine cable connection;
• no requirement that the linked systems run in synchronism;
• no increase to the short circuit capacity imposed on AC switchgear;
• precise delivery of power in response to the operator’s setting; HVDC avoids the
normal constraints of an AC system and delivers power independently of
impedance, phase angle, frequency or voltage;
• the linked systems can preserve their independent management of frequency and
generator control; the link “relocates” the remote generation or system at the
inverter terminal which can supply power output, frequency modulation and
voltage regulation just as a separate generator might;
• a link can be relied upon as part of a system’s generation reserve. Though not
loaded (or loaded in an exporting direction), a link to a separate system can
remain connected but on stand-by and able to import power as soon as required;
and
• it can improve either or both AC systems’ stability, and therefore internal power-
carrying capacity, by modulation of power in response to frequency, power swing
or line rating.
2.2 The Benefits of Interconnection
By way of a review, the principal benefits from interconnecting two, or more, power systems
are set out below:
(a) Economies of Scale. In general, large generation units are more efficient than
small. However, too large a unit size runs the risk of major disturbances if it
develops a fault. The larger the interconnected system, the easier it is to withstand
the loss of a large unit.
(b) Fuel Economy. The dispatch of generating plant aims to use efficient plant for
continuous load and less efficient for meeting peaks. Wider fuel choices from
interconnected systems offer opportunities to optimise dispatch on a larger system,
where more plant options are likely to be available, and thus to reduce supply costs.
(c) Reduction in Reserve Capacity. A margin of reserve capacity has to be

maintained in operating generation of any system to cater for plant maintenance
and/or breakdown. The interconnection of two or more separate systems enables
their standby capacity to be shared, and the larger the combined system, the smaller
is the statistically determined reserve requirement. Additions of generating plant to
meet rising predicted demand can be deferred by interconnecting, or some of the
reserve is freed to supply demand.
(d) Diversity in Demand. Different types of consumer mix, East/West time zone shift,
North/South seasonal shift and even different religious observance (Moslem Friday,
5
14.20newD
Christian Sunday and different festivals) can all result in non-coincident system
peaks, such that the interconnected system’s maximum demand is considerably less
than the sum of demands on the separate systems.
(e) Fuel Source Diversity. Different types of generating plant have differing operating
preferences. Large coal-fired, combined-cycle gas and nuclear plant are efficient
when operating continuously. Hydro and gas turbines can be more suitable for
peaking and reserve duties. If a system with a high proportion of the former can be
interconnected with one with a high proportion of the latter, higher load factors can
be achieved on thermal plant by interchanging energy. Surplus hydro energy in flood
periods can be put to good use, saving fuel in the thermal system. There may also
be strategic advantages in a system having access to the alternative fuel sources of a
neighbouring system.
(f) Reliability and Security of Supply. The security and reliability of existing
networks will be increased by an interconnection making available the additional
variety of generation type and standby capacity from outside.
(g) Environmental Benefits. Many of the above factors have consequential

environmental benefits. Improved efficiency of energy delivery is the most obvious.
A particular benefit associated with hydro is the realisation of integrated operation
that can permit some reduction of carbon dioxide releases into the atmosphere from
fossil sources in addition to the saving in fuel. Further the cooperative operation of
run-of-river and storage hydro plants via interconnection enables significantly more
energy to be produced from a given hydrological scenario. This can apply within
and/or between river systems, and be coupled with opportunities of conserving water
for other uses or in drought conditions. It follows that interconnection also provides
opportunities to reduce the scale of new storage hydro projects, and thereby limit the
inundation of land and destruction of habitat.
(h) Financial Participation. The owners of interconnected systems are able to share
both the cost and the benefits of large projects, e.g. hydro, which may otherwise not
be economically justified.
(i) Technology Exchange. Shared systems tend to encourage standardisation of

designs, operational practices and information exchange.
(j) Pooling opportunities. In 1997 there were some fifty pooling agreements or
directives linking utilities in the world, from the formal to the informal, to realise the
benefits of interconnection. Most used a traditional central-planning style as the
mechanism for gaining the benefits. However, there is now the alternative of the
wholesale competitive market, giving third-party access via a transmission system for
bidding energy resources to determine the system marginal price for dispatch. As
well as energy, there are markets developing in power reserve, frequency regulation,
black-start capability and voltage control. These will call for a more precise ability to
control the operation of the power system and the advantages of DC over AC in this
respect will be a consideration for planners and developers. Unconstrained access,
under the transmission operator’s control will be vital in this world.
These benefits apply across the whole system as long as they meet the objectives that load
demand be reliably served at minimum total system cost [e.g. Ref. 13]. However, they can
often only be obtained within the framework of a power pool that coordinates planning and
6
14.20newD
operations so that economies of scale, increases in reliability, and other "system" benefits
are achieved to the extent practicable.
2.3 Technical Considerations
When independent, asynchronous systems are to be interconnected, there are some

configurations for which DC is the only solution or at least an interesting alternative, e.g.
sea-cable interconnection. DC transmission technology has been well described in the
literature [1, 14]. There are comprehensive and accessible bibliographies to HVDC literature
available from The Bonneville Power Administration (BPA) in the USA [7, 8, 9, 10, 11, and
12].
Normally with bulk power transmission, whether an interconnection or not, there is a choice
between AC and DC, and the determination may in some cases be a matter of economics.
In the case of DC a large investment is required in terminal equipment for conversion, and
this is mostly independent of the length of transmission. However, DC overhead lines are
cheaper than AC lines for the same power transfer and DC line losses are less than those for
AC lines. Recent trends suggest that overhead line costs are increasing at a higher rate than
the costs of terminal equipment. This means that break-even distances might reduce, that
is the distance at which the overall cost (including losses) of AC and DC alternatives are
equal. However, there are two more significant effects that increase the breakeven
distance. One is that series compensation can reduce the effective length of a line for
stability and FACTS controllers (power electronics applied to the control of voltage and
power flow) can extend the range or capacity of AC systems [15]. The other is the compact
AC line, which draws the phases closer together to lower the reactance and increase the
shunt capacitance and therefore improve the AC system’s capacity [16].
The break-even points of transmission cost between AC and DC taking into account the
investment and the operational losses, depends strongly on the basis of the loss evaluation.
From the technical point of view neither the transmission distance nor the amount of power
to be transmitted is practically limited with DC.
2.3.1 Long Distance Bulk Power Transmission
When large amounts of power are to be delivered over long distances, DC is always an
alternative to be considered.
AC transmission becomes limited by:
• the acceptable variation of voltage over the transmission distance and at all expected
loading levels
• the need to maintain stability, i.e. synchronous operation across the transmission, after a
disturbance, both transiently and dynamically
• the economic effect of additions necessary to correct the above.
The comparison of long distance power transmission with gas pipelines is beyond the scope
of this document. Suffice it to say that the pipeline would show advantage where gas has
many uses in the receiving area. However, where capacity is over 1GW, the distance
relatively long and the usage predominantly electrical, electricity transmission is expected to
be more economic. Obviously the many economic parameters need investigation. We can
observe that there are instances of both pipeline and transmission being built more or less in
parallel.
7
14.20newD
The DC line, needing as few as two conductors (one only for submarine with earth return)
compared to the AC line’s use of three, requires a smaller right of way and a less obtrusive
tower. Fig 2.1 shows schematically the tower configurations for 1200MW (two circuits AC,
bipolar DC) and 1500 to 2000MW transmission at EHV AC single circuit or monopolar DC by
alternative tower designs. (Caution; a single circuit or a single pole above 1600MW capacity
has not been built to date (2000) because of the effect on the system of the loss of such a
high capacity circuit.)
Figure 2.1
Comparable Tower configurations for AC and DC transmission
In the case of a remote hydro station connected only to a long transmission, it is possible to
consider an integral generation and DC link design with two results: to gain efficiency from
the hydraulic turbine by varying generator speed, and to combine the transformers for the
generators and the DC. The concept dispenses with AC filters, requiring the generators to
absorb the harmonics and supply the necessary reactive power. CIGRE has prepared a full
treatment of Unit Connected Generation, but such a scheme has not yet been built [17,18].
As an AC line reaches either the limit imposed by system stability or its thermal capacity and
if adding a parallel line is impossible, it may be possible to convert it to DC. Up to three
times the AC capacity should be obtainable by altering the tower head configuration, but not
the foundations, tower size nor the right of way [19, 20]. Running AC and DC on the same
tower is also possible [21,22]. No example of these being put into effect can be reported.
8
14.20newD
2.3.2 Interconnection by AC or HVDC
If two or more independent systems are to be interconnected by a synchronous AC link,

common rules have to be decided upon concerning security, reliability, frequency control,
voltage control, primary and secondary control of reserve capacity etc. When the basis for
synchronism is established, it depends on the structure and the strength of the power
systems and the number of interconnecting lines whether or not stability problems, e.g.
inter-area oscillations, may occur. In most cases more than one AC link is necessary for
reliability, but there are examples of single-circuit interconnections for energy and reserve
exchange, where limited reliability of the link is accepted.
Interconnecting the systems with DC removes any constraints concerning stability problems
or control strategies. The common rules listed above concerning security etc. can largely be
left within the jurisdiction of the separate AC systems, remaining independent of the
agreement to link. The interconnection can be made by HVDC back-to-back stations along
the border or by interconnecting load and generation centres within the systems by long
distance transmission. Examples for back-to-back interconnection are between the eastern
and western systems in USA and Canada, where a number of links connect the two
independent systems.
For submarine interconnection AC cables generate an increasingly wide variation of voltage

with power flow as distance lengthens, until the rating of the cable is fully taken up by its
charging current. Since intermediate, reactive compensation units cannot be installed, the
maximum practical distance was 50km until recently. The advent of the XLPE cable (cross-
linked polyethylene) for submarine use, with a lower shunt capacitance than earlier types,
has increased this limit to about 100km in recent years. Beyond this, DC is the only
technically suitable solution. A DC connection requires only positive and negative (go and
return) conductors, or in some cases a single conductor with sea return and, as stated
above, there is no practical limit to length except cost.
2.3.3 HVDC Multi-terminal Systems
When power is to be transmitted from a remote generation location across different

countries or different areas within one country, it may be economically and politically
necessary to offer a connection to potential partners in the areas traversed. Multi-terminal
DC is a possibility (see also Section 7.5).
HVDC multi-terminal systems allow more participants [22]. They have proved to be feasible
e.g. the SACOI 3-terminal cable system between Italy, Corsica (France), and Sardinia (Italy)
and the Quebec-New England 3-terminal overland system in Canada/USA [23,24] and series
taps have also been suggested [25, 26]. The Pacific Intertie and the Nelson River DC links
are examples of multiterminal DC put to practical use [22].
A further example for interconnecting more systems via long distance HVDC links is the
planning of the EURO-link, connecting Russia, the Baltic States, Belarus, Poland and
Germany, where a multi-terminal HVDC system is under consideration. The advantages of
interconnection can be exploited without establishing common rules (of frequency control for
example) and the AC systems can continue to operate and develop independently. If in the
longer term the requirements for AC interconnection are fulfilled and it is agreed to
synchronise, the HVDC transmission becomes a strong backbone in the interconnected
system and brings considerable stability advantages.
9
14.20newD
A choice is available to operate with a coordinated master power controller, or that each
terminal has its individual power control and the voltage-controlling terminal supplies the
balance. New control concepts may become available to overcome the need for a master
controller and to allow expansion with more terminals, each convertor operating with locally
available information.
Care has to be taken when weak systems have to be integrated into a multi-terminal system,
so that faults within them do not cause too widespread a disturbance [26, 27].
Furthermore, if a multi-terminal system is to develop and grow independently, as AC
systems do, the integration of a new converter station needs a review and re-coordination of
the control structure and parameters of all converters. However, smaller convertors (with
current rating below or equal to the current margin, about 10%, of the existing system) can
quite easily be integrated at a later date. See also Appendix I.8.
3 ASSESSMENT METHODOLOGY
Part 3.1 addresses specifically the process of assessing a new interconnection. Parts 3.2 and
3.3 explore specific aspects of using HVDC.
3.1 Identifying Alternatives for Interconnections
The following form part of the adjudication of potential options for constructing an
interconnecting transmission within or between power system networks:
• costs and benefits of each versus individual system expansion;

• reliability and security with and without interconnection;
• autonomy considerations and safeguards;
• trading arrangements, legal and institutional matters; and
• financing multi-participant projects.
Initial studies should focus on identifying the possible system benefits arising from
interconnection, based on existing plans for power system expansion in the target networks,
which otherwise might develop independently.
This initial analysis should be followed by a pre-feasibility study, which uses least-cost
development techniques to evaluate the expected costs and benefits of interconnection.
Where a link forms part of a distant generation project, the plant and transmission are
treated as one. Where a link is to be an interconnection, in the absence of a long-term
contract, the benefits are calculated from the energy expected to be traded with the buy and
sell prices, plus the capacity benefit to each system, plus any ancillary services (frequency
control, voltage control) that the link can deliver.
Proposals can then be made for project-specific detailed technical and economic feasibility
studies within different trading, institutional, and financing scenarios. The feasibility studies
should be undertaken to develop interconnection design concepts and transfer capabilities
and to generate a number of interconnection options according to type (AC, DC or a mix),
size (MW), location, length and timing.
The interconnection options are analysed and ranked with respect to power transfer
performance over a range of normal and contingency conditions, with more advanced
studies to confirm technical viability. Detailed cost estimates are prepared for those options
10
14.20newD
that show the best overall performance. Finally, the feasibility study is concluded with a
review of the operation, control and management requirements.
In order to generate points of comparison, a rigorous process of technical and economic

equalisation is suggested for the comparison of the various alternatives. The following areas
would demand attention during this process.
a) Power rating and overloading requirement. Once the objective of the

interconnection is clear, i.e. whether it is for energy exchange or firm power transfer,
the rating and overload capacities should be chosen carefully and kept the same for
all alternatives. All the alternatives should be able to perform equally under identical
sending and receiving end system conditions: voltage and frequency variations and
also the minimum and maximum short-circuit levels.
Thus if one alternative provides better stability or security to the network(s), the
criteria should be reconsidered as to whether such improvement should be required
of all alternatives.
b) Equal demands on the performance parameters of all the alternatives should be

considered. These parameters which particularly influence the cost of the
interconnection include the following:-
- Reactive power exchange by the interconnection at both ends. Identical

exchange limits should be set for all the alternatives, as a function of line
loading and the needs of each AC system.
- Harmonic filtering requirements. It is well known that DC and power

electronic devices introduce harmonics to the network, and provisions are
needed to limit the distortion. However, care should be taken that the levels
specified are not below the actual measured values of harmonic distortion
already present or allowable by convention. An alternative may be to require
a distortion improvement by all alternatives.
In some cases, modifications may be necessary to existing communication

lines (e.g. open wire telephone) paralleled by a DC overhead line.
- Overvoltage considerations. The limits on overvoltage should not differ

among the alternatives. Pollution determination (ESDD levels etc.) and sizing
of insulators should be done using the same probability of flashover for all
alternatives.
- Reliability and Availability. Criteria should not be different from the expected
performance of any other important element in the system, say the largest
generator or transformer.
- Criteria for fault clearing and recovery should be fixed and should strive to
provide identical security for all the alternatives.
c) Right of way. The cost of right of way has to be determined in terms of license fees,
cost of de-forestation, access, etc. as applicable to each country and for each
alternative.
11
14.20newD
d) Evaluation of losses. This point is of particular interest when assessing bulk power
transmission to be embedded in an existing synchronised network. The overall
reduction in system losses is calculated for normal flow patterns for each alternative
and the savings in the losses can then be capitalised to capture the relative
advantages of the alternatives.
e) Intermediate Loads. A long-distance transmission project, whether AC or DC, may

have to supply intermediate loads. Alternatives may be to add a separate supply,
possibly on the same towers, or to apply a ripple current via the ground wire for
small loads under 1MW (see part 7.5).
The assessment process establishes the interconnection as a viable alternative to other

system expansion options, and tests the technical equivalence of various proposals for AC or
DC interconnection. These proposals are then economically compared and ranked according
to least discounted cost, with particular attention given to identifying the risks and
uncertainties in considering economic choice, as discussed in Section 6.
In some instances some of the different interconnection options may be of different

capacities, especially in the earlier planning stages. In this case the economic and technical
comparison must consider costs and benefits associated with the power transfer. In a
deregulated network, market simulation techniques can be used to forecast the value of the
additional available transfer supplied over an interconnection.
3.2 Technical Assessment
A process of technical assessment is necessary, whether a link is AC or DC or a hybrid, to

obtain a budgetary cost assessment and to determine if an acceptable rate of return can be
expected. The technical studies are best led by the staff of the purchaser of the link and
may involve the use of outside consultants and/or manufacturers. Specific studies relating
to DC are included below.
3.2.1 Performance and Cost Analysis
The process in brief leading to a feasibility assessment follows:
1. Electrical Performance, AC and DC Systems
• Many technical aspects associated with HVDC links relate to the strength of the
AC systems to be connected. They are detailed by the two CIGRÉ WG14.07
reports [28, 29].
• Determine the expected use of the facility for continuous bulk transfer, seasonal
flows, reverse flows, or variation hour-by-hour according to market opportunity.
• Steady state power flows for different transfers with a number of critical outage
contingencies determine the need for AC network reinforcement or reactive
compensation.
• Assess the dynamic and transient stability of the network under a wide range of
operating conditions and disturbances, especially at minimum AC system
strength.
• Switching surge and temporary overvoltages at either convertor station AC
busbar caused by faults, other transient events or load rejection lead to insulation
coordination and surge arrester requirements. DC station costs are significantly
affected by the adopted AC voltage control strategy. Load networks in the
vicinity of convertor stations must not be subjected to damaging overvoltages.
12
14.20newD
Studies will identify steady state and short time over voltage ratings as well as
the requirement for voltage control and compensation. Each DC scheme will be
unique in this regard.
• System protection requirements.
2. Performance of a DC System
• Harmonic distortion is examined to establish filtering requirements under steady

state conditions. Also, recovery of the DC link from AC disturbances might be
influenced by harmonics if post fault transformer saturation causes distortion of
the AC commutation bus at the inverter which in turn may delay DC link recovery.
The reactive power balance and the mitigation of harmonic distortion need to be
coordinated because both the reactive power requirements and levels of
harmonics generated can vary considerably with the level of power transfer.
• Subsynchronous resonance effects should be looked at to ensure that the DC link

does not contribute any negative damping to subsynchronous modes of oscillation
possible within any of the turbine, generator and exciter masses.
• Performance of the DC link during and after faults is important. When the short-
circuit ratio (see Section 3.2.6) at either of the convertor stations is low, switching
surge and temporary overvoltages require particular attention. Costs are affected
by the strategy of AC voltage control and studies will identify equipment ratings,
steady state and short time, and any requirement for compensation. Any need to
reinforce the AC network is relevant to these aspects.
3. Costs of transmission line and cable need to be determined if they form part of the
project. Estimates for the construction costs of an overhead line follow the same
general procedures for an AC or a DC line. Costs of submarine cable can be very
significant and should consider not only its manufacture, but also its laying with
submarine trenching (which may be necessary for its physical protection from
shipping and fishing), as well as possible costs for repairs in case of faults or
damage.
4. Alternative termination points for the connection can be considered and if so, the
studies (1), (2) and perhaps (3) above need to be repeated and results compared.
5. Consider particular aspects of HVDC, as outlined below in parts 3.2.2 to 3.2.9.
6. Prepare a preliminary outline of the plant required for the stations and transmission
lines and develop corresponding mini-specifications (Appx III for HVDC) for line and
station equipment. Obtain budget estimates from suppliers and manufacturers for
input into the feasibility assessment.
3.2.2 HVDC Electrodes
Long distance HVDC by land or sea can use the earth as a return conductor.
Continuous use in a monopole is rare, but temporary use as the return path when one pole
of a bipole is out of service is well established. Electrodes need careful siting, whether for
land or sea conduction. Metallic return is the alternative, either as a third conductor from
end to end or by the switching the available transmission circuits as either pole or the
neutral return path.
13
14.20newD
Assuming firstly a bipolar system, (Appendix I.2) a certain quantity of current will flow in the
electrode due to slight unbalances between the two poles, which never can be avoided
completely. It is therefore possible, at both terminals, to connect the station neutral point to
the ground mat of the HVDC station, to which the line side star points of the convertor
transformers are also connected, and to stipulate that the unbalance current cannot exceed
a low value [30]. Currents high enough to carry power would lead to corrosion not only of
the ground mat itself but also of other metal structures embedded in the earth e.g. cables,
pipelines. If direct current will flow more often than momentarily, it is common practice to
locate the DC electrode at some distance (10 to 50 km) by means of an electrode line. The
location of the electrode is chosen so that power currents are safely conducted to the earth
or sea. Siting must take into account environmental and technical conditions such as
accessibility, step and touch voltage, the distance to other metallic structures, geological
structure of the ground etc.
Continuous operation with ground return is rare for long-distance overhead line
transmission, but more common for submarine connections. For sea crossings, electrodes
can be installed on the sea bottom or as a shore installation, the principles being to feed the
current to the sea water at all tide conditions and to protect the installation from wave action
or storm.
Using the sea is low-cost (sea water has low resistance), but the direct current in the HV
cable will affect the deviation of a magnetic compass on board a ship nearby. Some
authorities restrict this, even though it does not affect modern navigation tools. A solution
that has been used at Baltic cable [31] is to route the cable to the sea electrode alongside
the HV cable wherever the water depth is shallow enough for the monopole cable to affect
shipping.
In bipolar systems the electrodes normally carry only a small fraction of the nominal current.
However, in the event of a failure of one pole, the operation of the remaining system can
continue either with the full current flowing through earth or by switching to metallic return.
Either method contributes significantly to the availability of the bipolar HVDC system. The
electrodes are designed to carry the full current at least for a limited time. As it is not
foreseeable which pole fails, the current may be in either direction and both electrodes are
to be designed for anodic operation.
In monopolar systems (Appendix I.1) the electrode must carry the load current continuously.
The current rating must allow all modes of operation, including overload. Even if power
direction can change, the current direction is not changed in monopolar systems and the
electrodes at both ends can be different in order to fulfil their specific tasks as the anode
and the cathode. For anode operation the selection of the material and the design of the
electrode depends on the material loss caused by current flow. The design of an earth or
sea electrode operated exclusively as a cathode is not a problem.
An alternative is the construction of an earth electrode at great depth. One such is in trial
service [31]. A full treatment is offered in the CIGRE report on Electrodes [32].
3.2.3 Losses
The technical assessment is not complete without an evaluation of the impact of losses.
For the key components such as valves, converter transformers, smoothing reactors, AC and
DC filters, DC line or cable, the respective fixed and variable losses will be determined by
14
14.20newD
bidders in order that bids may be compared, Ref. [33] gives the loss calculation procedure
now established by IEC.
But it is for the user to give appropriate values to these future losses. Each may have his
own analytical approach to assess the capitalised costs of losses. Values will vary according
to the period of time the losses occur as well as the hour or season, but these future values
must be captured, usually from the fixed and variable components, in a capitalised value per
kW at a few most probable loading levels. These loss values will later be applied in
component design and the selection of conductor size for the DC line or cable.
If various alternative configurations are to be evaluated and compared, the effect of each on
the network’s losses is needed. A base case network configuration is established for the
condition before the link is in service and total system losses are determined by power flow
studies. Next, the total system losses are determined with the link in place and operating at
load for the various configuration options under consideration, and the incremental changes
in total system losses are compared against the base case. If the added link relieves an
overstressed system for the case where no new load or generation is added, system losses
may decrease. If so, then in the capitalised cost of loss evaluation, a loss credit is
determined.
The following, developed when considering HVDC specifications, may be useful.
Losses are made up of a power element, the cost of providing capacity to generate the kW
that are lost, and an energy element, the fuel consumed. The power element can be taken
as the cost of generation reserve, either the marginal costs of another gas turbine or the
average cost of all generation installations in the network, in $/kW. The latter seems fairer
to many. The energy element is the average fuel cost over time whether a typical day or
season or year.
The above applies to a predominately thermal powered system. For a hydro system, there is
a choice for the capacity element of either the average cost of the principal generation or the
marginal cost of the next item of peaking plant, whether hydro or gas turbine. For the
energy charge, the cost of the firm average kW applies.
Today however, it is more realistic to accept that market prices vary enormously by season
and time of day and to select an appropriate average of the day by day capacity and energy
charges that apply.
These two elements (in $/kW/year) are summed and present-worth calculated over the
expected life of the project and then calculated for no-load (continuous 24-hour loss) and
load (when the link is at full power). Step by step then, first for no-load loss evaluation, it is
necessary:
• To select average generation cost per kW

• To project forward and discount at a typical cost of capital (an interest rate of funds) for
the number of years of the life of the project
• To select the average energy cost
• To evaluate the total cost of supplying this for a year, 24 hours a day
• To calculate the discounted cost of this energy for all future years of the life of the
project
• To add the capacity and energy terms to give a total discounted cost of losses in $/kW.
Second, for full load loss evaluation, follow a similar procedure, where:
15
14.20newD
• The capacity and energy charges are averages of conditions when the link runs at full
power
• The energy cost is applicable not to 24 hours a day but to the number of hours when
that power is expected to run
• The addition of these gives a total discounted cost of losses at full power.
Lastly, a formula relating the duration the link is expected to be used at full load, no-load
and other load levels should be approximated. For example, for 15% of time it might be on
standby, for 60% of time at 75% output and for 25% at full load.
The final $/kW values are reduced by the factor of duration at that output. The loss
evaluation to be used in the choice of equipment is then published in a specification as a
formula to be completed as follows.
A loss of kW at 0 MW flow, @ $/kW = $

A loss of kW at 50% flow, @ $/kW = $
Total : $
The load duration curve for the DC link may be quite different on a seasonal basis and
therefore an adjustment for different seasons may be necessary.
3.2.4 Reliability
An assessment of reliability influences the economic evaluation. Reliability calculations will

indicate the requirements for spares to achieve a required performance. For a DC system,
this applies to convertor transformers and smoothing reactors, spare filters and
compensation, all of which can be decided by analysis of reliability.
The reliability of a HVDC system is expressed in the targetted number of forced outages and
the non-availability of the system. WG 04 of CIGRÉ Study Committee 14 developed
definitions for the reliability terms and parameters of prime interest and prepared a protocol
[34] for use in collecting and compiling data. Operators of HVDC systems collect data of
availability and reliability. These data are assembled by this Working Group and published
every two years, e.g. [35]. The WG also prepared a paper in 1997 summarising the
performance of systems using 1983 to 1994 average unavailability data [36]. The CIGRÉ
data can be used to make some overall statements about the reliability of HVDC systems.
In the CIGRÉ information, the Forced Energy Unavailability (FEU), together with the number
and duration of outage events by category, is considered to be the most useful
measurement of system performance for operation evaluation and for planning of future
systems. The FEU is defined as the per unit amount of annual energy, based on full rated
capacity, that cannot be transmitted over the DC system due to forced outages of convertor
station and other equipment. Therefore the FEU values are in equivalent hours, which are
the sum of unavailability hours after the durations have been adjusted in proportion to the
percentage reduction in capacity. For example, for a two hour outage of one pole of a
bipolar system (50% loss of capacity) the equivalent outage hours would be one hour.
From the data available so far, an important conclusion from the perspective of assessing
overall system security and reliability is that a high percentage of the total FEU is due to
16
14.20newD
faults within an individual valve group. Only small FEU contributions are due to pole and
bipole outages. Another significant observation is that bipoles with only one convertor per
pole appear to date to be more reliable on average than bipoles with two convertors per
pole. This appears to support the logic that the more items of equipment, the more there is
to go wrong.
The data published to date suggests that the FEU of two-terminal line/cable systems could
be as follows:
(a) For two-terminal, one convertor/pole system:

V.G. plus pole FEU = 0.3%; bipole FEU = 0.015%;
Total FEU = 0.315%
(b) For two-terminal, two convertor/pole systems:

V.G. FEU = 2.45%; pole FEU = 0.08%, bipole FEU = 0.015%
Total FEU = 2.55%
The above values are for both convertor terminals inclusive, but they exclude outages due to
transmission lines or cables. Based on the average of all reports, the lines/cables have
resulted in an average FEU of about 0.25% per scheme.
Considering all schemes, the overall performance of transmission/cable systems is

comparable to back-to-back schemes. However, the average unavailability for schemes
commissioned after 1985 is significantly increased by some high outage values in just a few
bipoles due to special reasons. There were high Itaipu (Brazil) outage values in 1989/90,
but in the subsequent four years this scheme gave excellent performance with an average
FEU of 0.33% for the two bipoles. Overall, the performance of transmission/cable systems
continues to be very good, i.e. in the range 95 to 98%.
The document “Reliability of HVDC Convertor Stations” produced by the IEEE Substations
Committee, WG I7, gives advice on preparing the reliability sections of a HVDC specification
[37].
3.2.5 Spare Parts for HVDC
HVDC Convertor Stations, as any other plant, need a stock of spare parts for maintenance
and repair. The determination of how many spare parts are required should be based on a
calculation of the required total system reliability and availability and the expected statistical
failure rates. It is not necessarily a question of “as reliable as possible”, but a consideration
of the objective of the link, its reliability with reference to the other elements in the power
system and its likely performance.
Spare parts philosophy directly affects the cost of a HVDC system. The most important and
costly single components are the convertor transformer and the smoothing reactor. In
smaller links, if the transport weight of the convertor transformer is not exceeded, the three-
phase three-winding transformer is the most favourable solution. However, if a reserve
transformer has to be ordered, e.g. because availability analysis indicates the requirement
for a spare, a single-phase configuration may be more economical since the spare is
provided by a fourth transformer of single-phase rating, instead of a complete three-phase
spare.
As HVDC transmission systems are seldom repeated (if there are other convertor stations
within a reasonable distance of each other, their rating is not necessarily identical), the
17
14.20newD
spare part philosophy for HVDC components is unique for each link. So the distance
between the stations and transport determine whether or not spare parts are separate or
common to both stations.
Reference [38] describes a HVDC link built in Canada for energy interchange and exchange
of reserve generation. This link was not required to be more reliable than a single generator
in either of the systems it interconnects. It was therefore not necessary to have any spare
transformer, because the 6 to 9 month repair time was no worse than a similar event on a
generator unit or its transformer. The link has in practice given 10 years service without any
transformer problem.
3.2.6 AC System Strength
Most HVDC systems use line-commutated convertors. Smooth operation of these convertors
occurs when there is a strong AC system with constant voltage and frequency. The ratio
between the AC system short circuit MVA (three phase) and the DC power, the so-called
short-circuit ratio (SCR) is a measure for the strength of the AC system.
Fault level ( MVA)

SCR =
DC Power ( MW )
A further measure is the Effective Short Circuit Ratio (ESCR) where the shunt connected
filter and capacitor Mvar are subtracted from the short circuit MVA before dividing by the
scheme MW.
Fault level ( MVA) − Shunt capacitor ( MVA)

ESCR =
DC Power ( MW )
For strong system conditions e.g. SCR > 3 no major problems are to be expected. In case
of weak system conditions e.g. SCR < 2 problems may occur such as increased number of,
or repetitive, commutation failures. [Refs. 28, 29] have an extensive treatment of weak AC
systems. Low ESCR values, down to about 0.9, are possible with special design features
[38, 39]. For these cases detailed studies are performed in order to design the interface
between AC and DC. Critical to this is the reactive power balance, usually the first concern
to the DC designer. His job is easier if a flow of reactive power to and from the AC system is
possible.
If reactive power is constrained tightly, the individual filter banks must be very small (adding
cost) and offset at low DC power by shunt reactors in order to maintain the filtering.
Sufficient reactive power is needed at the receiving end to maintain a useful AC voltage, but
if supplied by shunt capacitors, it reduces the ESCR.
If the (relatively inexperienced) design incorporating a series capacitor in the convertor

station is used, the reactive power variation across the range of power being transmitted is
lessened [40].
A specific situation is the feeding of HVDC power into a system without generation. The
short-circuit capacity necessary can be installed by synchronous condensers and the start-up
of such a system has to be supported e.g. by black start diesel generators. Recent designs
using self-commutating inverters incorporating GTO thyristors or IGBT devices avoid the
18
14.20newD
need for rotating machinery [41]. These are available to approximately 200MW power
ratings.
3.2.7 Commutation failure
If the commutation of the current from one phase to the next in a line-commutated
convertor is not successful, this is called a commutation failure. It can occur only at the
inverter and the effect is a short circuit on the DC-side and a momentary loss of power
transmission.
Commutation failures are often or mostly caused by events in the AC system. A voltage
drop or a phase angle change in the AC system caused by a switching event or system fault
might cause a commutation failure in the DC system. See Reference [42] for a wider
treatment.
The risk of commutation failure is of a random and statistical nature, and the design
objective, therefore, is not to avoid commutation failure but to lower the rate and to
minimise any consequential effects. The normal solution is to use a minimum value of the
extinction angle, which allows sufficient margin to take care of most unforeseen events that
affect the commutation process. However, the greater the extinction angle, the greater the
DC system voltage ratings and costs. Increased losses are also a drawback of using a higher
margin, because the reactive power absorption of the HVDC plant is increased.
For multi-infeed configurations, where separate DC links are connected to a system with low
electrical distances between the convertor stations, the weak system condition can be
reached very quickly. Insufficient decoupling increases the chance of simultaneous
commutation failures of convertors in the area. These situations and the effect on the AC
system need study, as a disturbance to a number of links might stress the whole network
beyond its limits. Ref. [43] contains a treatment of these effects.
3.2.8 Subsynchronous Frequency Considerations
Large turbo-generators are sensitive to subsynchronous torsional interactions (SSTI) or

subsynchronous oscillations. If the mechanical resonance of any rotor shaft is the same as
the complement of the electrical resonance of the AC system (i.e. electrical frequency =
fundamental freq. – mechanical freq. between 10 and 40Hz), a small excitation may lead to
increasing subsynchronous oscillations and under worst-case conditions, to damage to the
shaft. Power electronics such as a HVDC rectifier station with constant power regulation,
may initiate such an effect [44]. However, power modulation for AC system stability
assistance is usually required at frequencies no higher than 3 to 5 Hz and the control
response can be therefore be suitably tailored, both to provide this service and to avoid SSTI
with an adequate margin.
This risk may need careful study if any of the following apply:
- Turbo-generators and HVDC rectifier station are located close together

- Weak interconnection of the turbo-generator to the AC system
- Rated power of the HVDC and the turbo-generator are of the same order of
magnitude.
19
14.20newD
3.2.9 Harmonic Modulation
There have been concerns over the exchange of harmonics between two AC systems,
running at slightly different frequencies, by a back-to-back DC link. The matter has been
thoroughly examined and is not seen as a barrier to the operation of such interconnections
[45].
3.3 Tools
The tools for simulating and analysing system conditions to arrive at a viable technical
assessment are well established and are detailed in reference [46]. The main tools are:
• AC power flow program to evaluate expected system performance, component rating at

load and under contingency operation.
• Short circuit program for evaluation of short circuit ratio and circuit breaker duty.
• Transient stability program to determine the transient performance of the system, and to
determine dynamic operating limits for various contingencies.
• Eigen-value analysis program to confirm the dynamic performance of controls and check
the elementary interaction between DC links, synchronous generator excitation control
systems and FACTS controllers.
• Dynamic stability program for analysis of system damping and to check on the results of
the eigenvalue analysis and operation of the DC link within the AC system.
• Electromagnetic transient program for insulation coordination and non-linear control
design and validation. The electromagnetic transient program is also a check on the key
results determined using other programs.
• Long term dynamics program
• DC simulators and Real Time Digital Simulators
• Tools for system reduction to a size that can be handled by simulators
• Economic cost models, such as a spreadsheet by year, discounted cash flow, production
cost model and a market simulation package.
The key requirement for the above tools is a suitable representation of HVDC link
characteristics for steady state, dynamic and transient operation.
4. TRANSMISSION SYSTEM COSTS
This section covers both typical costs, where available, of installations and the factors
affecting them. The possible staging, the reliability objectives, and aspects of power
systems or the environment that affect costs are contained here.
4.1 Check List of Cost Areas
This section lists areas where costs need to be considered. These apply equally to AC and
DC projects.
4.1.1 Engineering Costs
During the feasibility stage of the project, the following engineering activities can be
identified.
20
14.20newD
• Feasibility studies include the comparison of alternatives and, depending on the number
of parties involved and the regulatory environment which must be satisfied, can take
some time to complete
• Consulting costs
• Design and engineering costs, including preparation of specifications
• Bid evaluation costs, plus financial and contractual arrangements associated with each bid
• Contract and technical administration costs.
4.1.2 Environmental Costs
Prior to the commitment phase of the project, considerable effort may need to be expended
on environmental and related activities:
• Costs for environmental studies, which may include special consulting services.
• Costs for obtaining permits
• Costs of environmental mitigation
• Costs for public relations and public consultations
• Compensation to municipalities and landholders, which the transmission line or cable may
traverse. Prolonged delays in gaining permits for right-of-way may cause serious
financing costs if these permits are withheld because expected compensation to the local
community is not provided. Provision for value added compensation to affected
communities may help to avoid such delays.
4.1.3 Implementation Costs
With the project committed, the costs of construction and implementation will have
components for:
• Acquisition of real estate and right-of-way costs

• Site and route preparation costs
• Transmission line or cable
• Stations and reactive compensation equipment (may require strengthening in the AC
network)
• Electrodes and electrode lines
• Supply to intermediate small loads
• Communications between terminals, to be used for control, protection, etc.
• Spares provision
• Customs duties if imported materials
• Construction scheduling and staging costs including installation and commissioning
• Warehousing, both during construction and for spares after commissioning
• Insurance not covered by the contractors
• Any mitigation measures required to avoid interference in existing communication
facilities
• Mitigation measures for radio interference and audible noise at stations.
• Change orders
• Interest during construction as well as costs of financing and the allowance for escalation
of prices and labour over the duration of construction
• Training
• Contingency during construction
21
14.20newD
4.1.4 Operational Costs
Operational costs may be capitalised and added to other capital costs to evaluate their
impact on feasibility and selection of variation or supplier:
• Costs of losses.
• Operating and maintenance costs will depend on the degree of remote control,
automation and condition monitoring included in the stations. For the line or cable,
maintenance and inspection needs to be anticipated including on-going clearing of the
right-of-way. Operation and maintenance may be undertaken by utility staff or by
contractors.
It is prudent to assume that the convertor stations will need continuous manning for the
first two or three years, after which a remote control may apply. There are examples of
remotely controlled stations [38].
• Down time costs for component outage, both forced and scheduled, to some extent
influenced by the quality of the training of the maintenance staff and by the spares
available. Typical outage durations can be estimated from historical outage statistics
published by CIGRÉ (Working Group 14.04) [35, 36].
4.2 Station Costs
Many utilities have in-house, up-to-date and valid capital cost estimates for most AC
transmission equipment. These costs are used for planning new facilities and for the capital
budgeting and decision-making processes.
However, most utilities do not have readily available cost estimates for DC transmission
facilities. Consequently, in the case that DC is a competing option for a project, obtaining
costs requires a preliminary investigation and correspondence with manufacturers and
perhaps consultants.
One of the difficulties with this exercise is the vitally important requirement that capital cost
estimates for competing options are obtained on a comparable basis. When obtaining
capital cost estimates of DC for planning purposes from convertor station suppliers, it is
valuable to prepare a “mini-spec” listing of the essential parameters. Also important is the
commercial framework expected of a future enquiry: turnkey or civil work separate,
financing source, predicted time scale, warranty period, operating responsibility, etc. See
Appendix III for a questionnaire leading to a mini-spec.
The following issues and questions may become relevant:
• When quoting and using DC costs, it is important to recognise what facilities and cost
add-ons are, or are not, included, such as AC switchgear at the terminals
• How do estimates compare with usually quoted values?
• What is the cost variation in respect of scheme size, voltage and configuration? (A cost
breakdown by components may be of interest.)
22
14.20newD
• What are the advantages/disadvantages of the DC stations being designed and installed
by the Owner versus a turnkey supply and what are the cost comparisons?
• How competitive is DC compared to an alternative AC option? What is the “break-even”

transmission distance with today’s costs and technologies? This question is relevant in
the situation of conventional land transmission lines and water crossings where there are
no special requirements or asynchronous systems requiring the inherent advantages of
DC.
• What is the real potential for reduced DC costs due to new and emerging technologies
such as, for example, new filtering concepts and new valve types or designs?
• In a ‘bottom-line’ sense, what are the approximate capital costs that utility planners
should be using today for DC facilities?
In any economic assessment of either overall project viability or comparison of competing

options, the base capital costs are invariably the most significant items and hence the most
important in terms of accuracy. Throughout this exercise and indeed throughout the process
of defining a transmission project, the accuracy of a cost estimate will improve, as the
parameters (technical and commercial) become better known.
In recent years, several reviews have been done on the capital costs of HVDC equipment
supply [5,6].
TABLE 4.1 HVDC Turnkey Cost Division
Cost values given in year 2000 US$/kW (both ends inclusive) for one valve group per pole.
Back-Back Monopole Bipole Bipole Bipole

500kV ±500kV ±500kV ±600kV
200MW 500MW 500MW 1000MW 2000MW 3000MW
% % % % % %
Valve Groups 19 19 21 21 22 22
Convertor Transformers 22.5 22.5 21 22 22 22
DC Switchyard & 3 3 6 6 6 6
Filtering
AC Switchyard & 11 11 10 9.5 9 9
Filtering
Control/Prot./Comm. 8.5 8.5 8 8 8 8
Civil/Mech. Works 13 13 14 14 13.5 13.5
Aux. Power 2 2 2.5 2.5 2.5 2.5
Project Eng. & Admin. 21 21 17.5 17 17 17
100 100 100 100 100 100
$130 $90 $180 $170 $145 $150

Total per kW
This material has been reviewed again and a cost breakdown analysis for recent DC stations
23
14.20newD
is presented in Table 4.1. It obviously must be treated with caution, as quoted costs of DC
stations are subject like anything else to the vagaries of the marketplace. Though they have
recently been dropping, it cannot be predicted whether such conditions will continue or
reverse.
In table 4.1, the “total” level gives typical turnkey costs of the vendor’s HVDC supply and
installation. These costs cover both terminals (of a two-terminal scheme), and are based on
some simplifying assumptions. They assume that the DC bipole is made up of one valve
group per pole and also that no special measures are required for reactive power
compensation and/or voltage control to incorporate a DC scheme into a weak AC system.
These costs also do not include any costs by the purchaser entity itself, taxes, interest
during construction or other money borrowing costs. In certain applications, the purchaser’s
costs can be substantial.
If ±500kV is selected in place of ±600kV for a 3000MW bipole, a station cost approximately
5 to 10% lower applies. In view of market volatility, the above estimates should be treated
as having an accuracy no better than ±20%.
These costs can be used to explore development options but confirmatory figures obviously
need to be obtained from manufacturers. Each power system is different with respect to
voltage, system strength, harmonic and reactive power limits. Each HVDC scheme is
therefore unique and caution is needed when utilising the turnkey costs and facility cost
variations discussed above for comparing options.
A “check list” of possible additional costs is as follows. There may be specific project
conditions also.
(a) Multiple valve group poles - In some cases, usually for the larger power ratings,
single valve group poles may not always be acceptable. There could be restrictions
due, for example, to weight transport limitations of convertor transformers, to system
reliability implications of the loss of one large valve group, or to earth currents or the
losses in metallic-return during a valve group outage. Multiple valve group poles
require additional equipment and switching on both the AC and DC sides. It is
estimated that there would be about a 20% total cost premium to the turnkey supply
for a same-size bipole having two parallel valve groups per pole instead of one. If
the two valve groups are in series, however, this extra cost applies only to the
second (higher-voltage) valve groups.
(b) AC stations - The AC station costs included in Table 4.1 are those required for the DC
and associated equipment connections only. Switchgear for outgoing lines must be
added and, especially for the larger power bipoles, more extensive AC stations may
be required for additional transmission line connections or for other equipment, to
provide equal performance to other schemes under comparison.
(c) Compensation for reactive power/voltage control - For short circuit ratios below about
3.0, costs may rise due to special compensation measures in addition to the AC
filters. Costs may rise where
- there are large values of capacitor banks or AC filters, beyond the needs of
the HVDC link itself, which reduce the effective short-circuit ratio and may
require measures to avoid low-order resonance.
24
14.20newD
- the accepted step change in voltage when an individual filter is switched in is

small, since it dictates a large number of small filters and associated
switchgear, plus perhaps off-setting shunt reactors for use at low power
levels.
Additional shunt capacitors cost about $10/kvar. More sophisticated control devices,
such as SVCs or STATCOMs, cost from about $30 to $50/kvar (total installed cost).
In the case of very weak systems where additional system strength in the form of
synchronous compensators is applied, larger costs from about $70 to $90/kvar may
be implied.
(d) Electrodes are included, but not their connecting transmission lines or cables because
they depend on length.
(e) Communications beyond and between the convertor stations are not included.
4.3 Overhead Line Costs
Depending on the degree of system reliability required and the sensitivity tolerated for
transient and permanent line faults, various types of HVDC overhead lines can be
constructed with different remaining transmission capacity after line faults (fig. 4.1). An
increase in reliability means an increase of cost of transmission lines. The figures in
parenthesis in Fig. 4.1 assume that the two station poles can be switched to operate in
parallel (at a small increase in the station cost) and that the line conductors have the
thermal capacity for twice the current. This may be inherently available as conductor sizes
are selected to satisfy design limits for corona discharge.
Little has been written of the difference between DC and AC line construction costs, since a
very early treatment [47]. Some indication of comparative line costs is necessary to assist
the reader’s assessment of DC and AC transmission. Since terrain and local practices differ
markedly around the world, WG 14.20 has sought industry opinion, by requesting not actual
prices but cost ratios between AC and DC transmission line construction. Designers will be
familiar with the actual costs of AC lines in their territories, and it was hoped that the ratios
would apply, broadly, anywhere.
25
14.20newD
Variant Tower Remaining Transmission Capacity Relative

Configuration cost %
Loss of one pole Tower breakage
Ground return
permitted not permitted
Single
monopolar 0 0 0 85
line
single 50
bipolar 0 0 100
(100)
line
double
bipolar 100 100 0 114
line
Two 50 50
monopolar 0 126
(100) (100)
lines
two lines
(bipolar or 100 100 100 136
homopolar)
Figure 4.1
HVDC Overhead Line Configurations
The working group suggested four applications of approximate equivalent capacity as

tabulated below. Line designers were asked to calculate, for a typical situation familiar to
them, the ratio of costs per km of each DC line, using the corresponding AC line cost as 1.0
per unit per km length. Suitable design parameters, including conductor sizes, with which
they were familiar for each case, could be used, assuming a simple bipolar tower without a
metallic neutral return. The results are given in Table 4.2. The intent was to compare the
cost of towers, conductors and construction only, without taking into account other parts of
the system.
26
14.20newD
TABLE 4.2
Cost ratios for DC and AC transmission line construction
Case AC equivalent line Cost pu HVDC bipolar line Range of Costs pu

ratings
1. 230kV, double circuit 1.00 ±250kV, 500MW 0.68 to 0.95
4. 765kV double circuit 1.00 ±500kV, 3000MW 0.33 to 0.7
Note these figures are not derived from a statistically significant sample, but from only 3
approximate estimates by practitioners. The working group warns the reader that these
figures may not be of great value due to the wide range that resulted. It may be concluded
that a DC line has the potential to cost no more than two thirds of an AC line of similar
rating, but local knowledge pertinent to a project should be consulted.
The working group has felt that Cable costs have been outside its scope of work. Typical
prices depend on the power to be carried, the distance and the water conditions of depth,
temperature, etc. They can be readily obtained from the manufacturers. CIGRE Study
Committee 21 covers Cables and there are general articles available [48]. Burial at sea has
a considerable cost but has proved a most valuable asset in terms of the reliability of the
link, as is demonstrated by the difference in performance between the first (unburied) and
second (buried) Cross-Channel links between England and France, see Appendix II & [49].
4.4 Stages in Expansion of HVDC Transmission
HVDC transmission can be fitted more readily than AC to a gradual expansion plan for
transfer of power. In this way, unnecessary investments can be avoided or delay of
investments can be obtained. AC transmission often has to be built from the start with a
high capacity to maintain stability, but DC can be tailored to discrete stages.
The most common staging in DC transmission is first to build a monopole and later a bipole.
To develop further from this stage, a new bipole can be added or the convertor stations can
be upgraded in current and/or voltage by adding convertors in parallel or series.
In many applications HVDC is chosen for large power transfers on a long-term basis. The
transfer may however be low in the initial stage and higher after a certain period. Based on
the build-up timing and having the investment costs for convertor stations in mind, it is
natural to evaluate different approaches of a stepwise implementation of the total HVDC
transmission scheme.
The major alternatives are:
1. Stage 1 Pole 1
Stage 2 Extension with Pole 2.
Usually the two poles have equal power and voltage ratings. The transmission line
might be designed as a bipolar line from the beginning and additional convertors are
added at both ends.
27
14.20newD
2. Stage 1 Bipole 1 at reduced voltage

Stage 2 Upgrading by increasing the DC voltage.
This requires additional convertors in series.
3. Stage 1 Bipole 1
Stage 2 Upgrading by increasing the DC current.
This requires additional convertors in parallel. Both alternatives 2 and 3 should have
the transmission line designed for the higher voltage or current capacity from the
beginning.
4. Stage 1 Bipole 1
Stage 2 Bipole 2.
The two bipoles need not be equally rated, but additional security can be gained by
paralleling poles under line outage conditions.
If the transmission includes submarine cables, it is often economical to install the required
rating and no more at each stage.
4.5 Environmental Aspects
Increasing environmental awareness throughout the world is impacting both the

implementation and the cost of transmission projects. Environmental objections to projects
can lead to lengthy delays in construction, and thus to increased cost. Interconnection
projects may span different jurisdictions, each with its own regulations and application
procedures. More directly, measures to mitigate environmental impact lead directly to cost
increases. On the other hand these cost increases have to be compared to the cost of delay
and especially to the income not received from a scheme that had proved to be economical.
The environmental issues which can give rise to increased costs include the following:
• As with AC overhead transmission lines, there is increasing resistance to the construction

of new DC transmission lines. The objections are typically based on visual impact and
concerns about electromagnetic fields. Objections to overhead lines led to a decision to
underground the land portions of the KONTEK scheme (Denmark-Germany), and to
prolonged delays to the Italy-Greece interconnection. These objections can affect
associated AC connections and system reinforcements as much as the DC transmission
scheme itself. This has been the case with the Moyle interconnection (N. Ireland-
Scotland), where a description of the necessary process has been published [50].
• Many submarine HVDC schemes are conceived as monopolar schemes, thus minimising
cable costs. However, the use of a sea/earth return path leads to questions of corrosion
of other metallic objects (pipelines, cable sheaths etc.), production of chlorine gas and
impact on fish populations.
• Monopolar submarine HVDC schemes cause magnetic compass deflections, depending on

cable orientations, water depth and current magnitude. In some jurisdictions there are
limits to magnetic compass deflections, which may require a return cable alongside the
pole cable, or the use of coaxial cable with integral current return, where ratings are
appropriate.
28
14.20newD
• Burial of submarine cables may be necessary to minimise the risk of mechanical damage
to the cables by trawling or shipping. However, the disturbance of the seabed caused by
trenching and cable laying may have adverse impacts on marine life.
• Stringent acoustic noise limitations may be placed on convertor station installations.

Measures to limit noise from reactors, valve cooling systems, filter banks etc. can have an
impact on the cost of the convertor station.
• Direct voltage across line insulators tends to attract and polarise airborne dust particles.
Anti-fog insulators are usually needed and creepage distance requirements are greater
than on an AC line.
• An objection to any higher-voltage transmission line may be that it is difficult to supply

small loads to villages along the right-of-way. Solutions for DC links are offered in
Appendix I.8 and it may be necessary to add such costs to any alternative under
consideration.
5. TRANSMISSION BENEFITS
In this section the benefits of HVDC transmission are listed. Most reflect that DC can offer
greater flexibility of operation and therefore a more satisfactory realisation in practice.
5.1 Technical Benefits
The classical application of HVDC is the transmission of a large amount of power over long
distances because the overall cost for the transmission system is less than AC and the losses
of the DC system are lower. A significant advantage of the DC interconnection is that there
is no stability limit concerning the amount of power or the transmission distance.
5.1.1 AC System Support
An AC load flow depends on the difference in angle between voltage vectors in different
parts of the network. This angle cannot be influenced directly but depends on the power
balance. Secondly, a change in power generation or in the load demand will cause a change
in system frequency that has to be restored by altering the generation. As this task has to
be fulfilled by the generator speed controllers, the frequency restoration is a slow action.
System stability also depends on there being sufficient flexibility to allow the automatic
adjustment of the voltage vectors.
If stability problems are encountered which can be solved by fast frequency control, HVDC
can fulfill this task by drawing the energy from the remote network. Due to the ability to
change the operating point almost instantaneously, HVDC can feed (or reduce) active power
into the disturbed system to control the frequency much faster than a normally controlled
generator. If the feeding AC system is strong enough the DC link can, within its rating,
control the frequency in the receiving system. A prerequisite for this kind of system support
is only the appropriate mode of control.
Take the case of an AC system containing relatively long transmission lines, where
electromechanical oscillations can be excited by system faults and are weakly damped.
Assume the addition of a DC link (point-to-point or back-to-back) from outside into this
system. Control features for power modulation with the appropriate phase angle can
actively introduce damping torque. In general this valuable feature of an HVDC link is
29
14.20newD
inherent and requires no significant extra costs. Where the systems at each end of the DC
have different natural frequencies of oscillation, the damping torque can be applied to either
system or both simultaneously if necessary.
Two controls are available. Where a terminal’s AC network is part of a large system, the DC
controls can react to swings of power and attempt to mitigate their effect by damping power
to maintain synchronism. Where it is a separate system, applying a slope characteristic
similar to that of a generator can be used to apply frequency control.
5.1.2 Limitation of Faults
Faults causing depression of voltage on power swings do not transmit across a DC barrier.
They may emerge on the other side of a DC link as a reduction in power, but voltage is not
affected. Constraining the influence of certain critical faults can be a valuable attribute of
DC.
5.1.3 Limitation of Short Circuit Level
When new lines are built to extend AC systems, the short circuit level of the system will
unavoidably be increased. The switchgear apparatus must cope with the short circuit
requirements or a refurbishment has to take place. Since reactive power is not transmitted
over a DC link, it provides a means to extend the active power exchanged without increasing
the short circuit level.
5.1.4 Power Flow Control
A DC link operates at any condition of voltage and frequency of the two AC systems. An
independent control is therefore available to transmit power, leaving each system’s existing
load frequency control to act normally. A valuable strategy then is to hold in reserve the
system control features given above (5.1.1) for occasions when voltage or frequency stray
outside the normal bands of operation.
Where a link is contained within one AC system the same applies, but special stability
controls act when system oscillations exceed a certain band of, for example, the rate of
change of busbar voltage angle.
5.1.5 Voltage Control
A DC link can also be used for voltage control. The convertor absorbs reactive power
depending on the control angle, which normally will be compensated by filters and/or
capacitor banks. By extending the control angle operating range (to lower voltage) and
additional capacitor banks (to raise voltage) together with a fast acting transformer tap-
changer, the reactive power demand can be used for independent voltage control at both
connection points. This operation outside the optimum (minimum) control angles leads to
higher short-time operational losses and stress on components, but these are usually
marginal compared to the operational improvement. If it is to be used as a permanent
feature, this method of operation has to be taken into account in the design phase [38].
It is important to realise that the normal constant power regime of a DC link can destabilise
an AC network under distress. A normal feature of the DC link is the voltage-dependent
current limit, where DC power is limited when voltage drops below the normal band, so that
the reactive power is made available to the AC system.
30
14.20newD
In disturbed conditions, it is a good principle to look after the AC voltage first, and order the
power flow second. There are substantial AC filters at the convertor stations, which can be
used to bolster AC voltage if stability is threatened. The DC control drops DC power, so that
the convertors absorb less reactive power and the reactive capacity of the filters is available
to the network. Though the loss of power flow is unwelcome, the boost to AC voltage may
be more valuable.
Self-commutated voltage-sourced convertors are able to provide independent control of real

and reactive power. Reactive power generation or absorption is possible, within convertor
ratings, at any DC power transfer [41].
5.2 System Reserve
The maximum unit site of generation in the system is determined by the maximum loss of
power for which the system frequency can be maintained within defined limits. When a
large amount of power is fed into an AC system by a HVDC long distance transmission
system, it can also be thought of as generation. The maximum power of one pole of the
HVDC link is in the same way limited by the system parameters.
The largest possible loss of power of a HVDC link in case of a fault depends on the DC line
tower configuration and on the ability to transmit power via ground or metallic return in case
of line outages. Assuming that the current carrying capacity of a conductor is well above the
nominal current there can be a short-time capacity of overload in the convertor and line on
the remaining healthy equipment, to reduce the shock to the system as a whole in case of
pole faults. Dimensioning the maximum sudden step-change resulting from a fault can
therefore be precise when DC is used.
5.3 Environmental Benefits
As well as the comparison of life cycle costs, the environmental compatibility of a design
alternative needs examination. An AC alternative constitutes AC lines plus AC substations
whilst a DC alternative has DC lines plus convertor stations plus corresponding AC
substations. The environmental impacts of the two methods will not be equal. A qualitative
comparison between AC and DC lines with regard to impact on the environment is:
• Visual Impact constitutes an environmental advantage for a DC line, since the tower size
for the same power is reduced compared to the tower size of an AC line.
• Right-of-way width of a DC line compared to an AC line is considerably reduced. This

facilitates suitable routes in densely populated areas and in regions with difficult
landscape.
• The corona phenomenon has a substantially different nature in DC than in AC

transmission. Generally for a bipolar DC transmission line and an AC transmission line
with almost the same rms conductor voltage to earth and equal transmitting capacity,
annual mean Corona Losses (CL) are more favourable for the DC than the AC case,
particularly in poor weather conditions.
• Radio Interference (RI) results from Corona discharges, which generate high frequency
currents in the conductors producing electromagnetic radiation in the vicinity of the lines.
RI measurements have shown that radio noise from a DC line is considerably lower than
from AC lines of similar capacity.
31
14.20newD
• Audible Noise (AN) values resulting from comparable DC and AC lines during fair weather
are quite similar. However, during rain, the better performance and the lower
interference levels generated by DC compared to AC lines are considered an advantage.
• Regarding magnetic fields, conditions for DC lines are quite different. Since a DC line has
an unchanged electric field, it exerts effectively no magnetic field on the surroundings.
The DC field of a monopolar line is comparable to the strength of the earth’s magnetic
field.
• Regarding generation and emission by DC lines of positively charged ions, O3, N2 and free
electrons, research studies and investigations of possible consequences have shown, up
to now, no evidence of hazard from any operating DC line.
6. ECONOMIC ANALYSIS
The rationale in planning any power project, regardless of its type or the institutional
arrangements, is to identify technically equivalent and viable alternatives for economic
comparison, in order to ensure the optimal allocation and recovery of financial resources. In
applying this rationale definitions must be clear as to what is meant by cost, and also the
quality of service to be produced as a measure of the associated benefit.
The technical analysis described in Section 3 and the cost estimation techniques set out in
Section 4 will lead to a fully costed transmission scheme. The benefit assessment as
described in Section 5 will provide an estimate of the impact of the transmission scheme on
total system costs. Project analysis texts are available [51, 52], from which it should be
noted that an important requirement is an assurance that there will be sufficient revenues to
pay for the transmission scheme.
Economic assessment is required to ascertain whether the benefits are sufficient in relation
to the costs for the project to proceed, and the notion of discounting is common to most
methods of investment appraisal. A useful reference book is Capital Budgeting Techniques
by F.M. Wilkes [55], as it defines what is to be optimised and indicates the constraints which
must be taken into account in a mainly free enterprise economy.
6.1 Factors to be considered
Transmission links may be built for a variety of purposes, and the specific factors to be taken
into account in an economic evaluation will depend on the purpose of the link and the
perspective of the party for whom the evaluation is being conducted. A central problem in
the analysis, however, is that of identifying the risks and uncertainties. These arise from
variations in the discount rate, inflation, border prices, construction costs, etc., and are now
compounded by non-economic costs associated with the environment and institutional
changes bringing competition to electricity supply.
Many existing and projected HVDC links are intended to deliver firm energy from a remote
source to a demand centre (e.g. Intermountain Power Project (USA), Nelson River (Canada)
and projected schemes to deliver energy from Inga to South Africa or to Europe). In these
cases the economic evaluation will be concerned mainly with the total cost of generation and
transmission, and the avoided generation and transmission costs in the receiving system. In
this respect generation and transmission decisions can not be taken separately.
This concept may be taken further in future in the development of merchant plants (that is
Independent Power Producers or IPPs) and their associated transmission, where the purpose
32
14.20newD
of economic assessment, from the perspective of the merchant, will be to calculate the cost
of energy delivered to the proposed market. The trend towards electricity liberalisation
shows no sign of abating and is encouraging the development of merchant plants.
Furthermore, there is no one model emerging for electricity markets.
Other schemes are developed for a one-way flow of energy, but without being associated
with a particular generation plant development. The Moyle Interconnector, between
Scotland and Northern Ireland, is an example of this type of link; it was proposed to enable
Northern Ireland to import from Scotland. In such circumstances, the evaluation must
consider the variability of the price of energy both purchased and sold in addition to the
various scheme costs and avoided cost factors mentioned above.
Some schemes are predicated on seasonally varying two-way energy transfers, such as
those between the Nordel and UCPTE systems of Northern Europe. The economic analysis
of this type of scheme will be based on the impact of the interconnection on total costs in
both systems. This will require estimation of total system investment and operating costs, to
balance against the system gains in economic efficiency and reliability.
Finally, HVDC schemes may be used to implement interconnection between systems for
mutual support, that is reserve sharing, generation capacity saving, opportunity trading etc.
This was the concept behind the UK-France Cross-Channel scheme [53, 54] and is the basis
for the Italy-Greece interconnection and the possible Ireland-Wales and Arabian Gulf
interconnections. In these cases the economic evaluation must also take account of the
impact of the interconnection on the total capital and operating costs in the systems
concerned.
6.2 Economic Criteria and Decision Rules
The unbundling of electricity supply sectors and the liberalisation of electricity markets have
introduced additional dimensions to interconnection economics, because transmission is
regarded as a natural monopoly. The rules of the market will determine how the benefits of
interconnection may be realised by the various market participants. The promoter of the
proposed interconnection, however, will wish to ensure that commercial arrangements can
be put in place to deliver an adequate return on investment (ROI) for the interconnection.
In this liberalised market environment, new transmission links, as with other power system
facilities, may be built either by public utilities or by entrepreneurs through private finance
initiatives. The suggested economic criteria apply to both public and private finance using
the method of discounting. On this basis each alternative is evaluated according to its yield
or internal rate of return (IRR), otherwise known as the discounted cash flow (DCF) rate of
return. The comparison discount rate (or factor) for the DCF calculation is the weighted
average cost of capital, depending on the source of funds, debt, equity, and retained
earnings.
It should be noted that the ‘yield’ or IRR is the discount rate for which the net present value
(NPV) of a project is zero. It is therefore the rate that would cause an investor to be
indifferent to the project. It follows that a developer would look at the yield or internal rate
of return (IRR) and the time period of that return. Thus if the IRR is positive and also
exceeds the comparison or DCF discount rate, then it may be worthwhile to make the
investment. Moreover, in comparing project alternatives the one with the highest IRR above
the comparison rate is usually the preferred choice.
33
14.20newD
If the link is to be privately owned and is to carry firm power transfers based on energy
sales and use-of-system contracts, the yield is calculated over the contract period of say up
to 10 years, based on capital and O&M costs for the facility and the revenues expected. It is
thus important to check demand and pricing arrangements and how these may affect
revenues over the period necessary to achieve the required IRR margin above the
comparison discount rate.
The same basic approach would be applied if the link was owned by a conventional (public)
utility, but in this case the emphasis would be on levelising the NPV of life-cycle costs over
the book life to calculate the minimum revenue requirement for inclusion in the rate base.
The book life is typically 30 years since long term recovery is usually a prerogative of public
utilities, as they are often subjected to rate-base regulation. Revenue requirements are the
amounts that must be covered in order to provide compensation for all expenditures made
as a result of implementing an investment decision.
It should be noted that for private finance initiatives the time period for achieving the
required return has been historically limited to the period of, for example, a power purchase
contract, say five to ten years. This can be a significant consideration, particularly if the cost
of money is also high, because revenues cannot be guaranteed for long as a result of
regulatory uncertainty or variable demand. Furthermore, any delay in gaining revenue can
greatly reduce the IRR.
The DCF approach is generally recognised as a logical, and by many as the most logical, way
to analyse and compare alternative investment opportunities involving different time periods
and generating different cash (or benefit) flows. However, a choice often has to be made
between the two basic decision rules, viz: a positive NPV or highest IRR above the
comparison rate.
Public utilities often prefer the NPV rule, levelised as a life-cycle stream of costs, because
they have to minimise the revenue requirement in making investments to meet future
demand. Developers, however, prefer the IRR rule with a good margin over the comparison
rate of return, often taken as the discount rate, as they need to offer good returns to attract
funds. Nevertheless, the value of a revenue-producing project for a developer is the net
present value of the cash flow, and so in ranking alternatives he will select for further
analysis those with the highest NPV.
6.3 Economic Analysis Methodology
The flow chart of Figure 6.1 illustrates a typical power sector economic analysis process and
the inputs thereof. The chart is taken from a training manual on power system planning
produced by Tokyo Electric, and illustrates the so called ‘least-cost’ method for the economic
evaluation of power projects. According to this method, two or more alternative plans for
developing power facilities are formulated, the annual cost of each plan for a given period is
calculated and the plan is chosen which minimises the present worth sum of the cost
stream.
In comparing alternatives there will be only one choice which results in the least overall cost
and thus minimum revenue requirement. However, in making economic comparisons it is
the difference in results, which is important. An economic comparison is thus the valuation
of a choice against the next best alternative or ‘opportunity foregone’. This is most
efficiently identified by the application of a discount rate to represent the opportunity cost of
money and by taking into account local market prices (in the cost of money as well as
equipment and construction) in making the analysis.
34
14.20newD
The discount rate, as discussed earlier is used in the DCF calculation to establish the present
value (or worth) of an investment and is invariably the weighted cost of capital.
Estimated Demand Service Level
Supply Capability of Necessary Link

proposed DC Link Reliability
Formulation of Draft
Plans for Developing
DC Link
Period of Economic
Evaluation
Calculation of Annual
Cost of Each Plan
Conversion of Annual
Cost into Present Worth
Selection of Plan
whose Present Worth
is Minimized
Figure 6.1 Flow Chart of Scheme Development
The present worth (PW) (or value) methodology collects the varying annual costs (or
revenue requirements) of a given investment and moves these through time to become a
single, cumulative, payment at one point in time. An alternative approach is to convert the
varying amounts of annual costs into a constant amount that would have to be paid
(through revenues) yearly over the book life of the investment. This approach is known as
the levelised annual cost (LAC) method, in which the sum of the constant annual costs is just
equal to the cumulative present worth of the actual annual costs.
The cumulative present worth, or present value (PV), of a stream of actual annual costs, S1,
S2, - - SN, taken over a period of N years and applying a discount rate (x) is:
t=N
PV = Σ St . [1 /(1 + x)t]
t=1
in which the expression inside the brackets is known as the present worth factor in
compound interest tables, and is applied to deflate to the reference year (t=1) the given
annual cost. As noted in the PV formula, the sum of all such deflated annual costs is the
present worth of the investment.
35
14.20newD
Similarly, the levelised annual cost (LAC) required to amortise, or completely pay off, the
present value of all annual costs over the book life (N) of the investment is:
LAC = PV [x (1 + x)N / ((1 + x)N - 1)]
The expression inside the brackets in the above equation is known as the Capital Recovery
Factor (CRF), equal to the discount rate plus the sinking fund depreciation, also to be found
in standard compound interest tables. The CRF is often taken as the “Fixed Charge Rate” in
calculating the fixed component of firm (base-load) energy. For example, the capital
investment for an increment of generation costing US$1500/kW, amortised at a fixed charge
rate of 20% over a book-life of 30 years and operated at 70% capacity factor, can be
recovered with a levelised annual charge of circa US$50/MWh. This is often regarded as the
avoided ‘fixed’ cost of energy.
6.3.1 DCF spreadsheet examples
Each plan alternative can be developed as a spreadsheet that would be used to represent
the various cost streams. The spreadsheets of Tables 6.1 & 6.2 show the application of the
Present Worth (PW) method for the analysis of a regional power development scheme
involving five countries and carried out under the auspices of the World Bank. The PW
evaluation methods used are as follows:
• Total Present Worth (Table 6.1): - This table summarises for each option the total
present worth as the present value sum of all capital and operating costs to the plan
horizon. The alternatives are then compared on a net present value (NPV) basis to
identify the ‘least-cost’ alternative and the savings from the regional over the
independent development in each country, also to a total systems base.
• Cumulative Present Worth (Table 6.2): -This table collects, for alternative Plan D of
Table 6.1, all year-on-year costs in each category expressed as current US$
estimates. They are summed to the horizon year, to give category and system cost
totals, and with the latter factored year-on-year to give the cumulative present
worth.
The concept of least-cost development, as illustrated by the above flow chart and
spreadsheet examples, implies that an optimum plan can be identified from alternative
programs, which will satisfy forecast demand at an appropriate quality (reliability) of supply.
Taken over a 10 to 20 year planning horizon, the method determines the least-cost
development program and the long run marginal costs (LRMC) of electricity supply.
Furthermore, the spreadsheet can be organised to identify both the life cycle (NPV) cost
streams and internal rate of return (IRR), and so is common to the evaluation of both public
and private finance projects.
An example of a spreadsheet analysis for a private finance initiative is shown in Table 6.3.
This example is a desk study looking at alternatives for a mine-mouth power plant and
200km of transmission line for interconnection with a public utility power system network.
The spreadsheet is organised to show for one of the alternatives the stream of annual costs
and revenues, with the latter based on avoided capacity and energy costs to the public utility
over a 10-year power-purchase contract period. The objective of the desk study is to
compare the internal rates of return (IRR) of different options.
36
14.20newD
TABLE 6.1
COMPARISON BETWEEN INDEPENDENT DEVELOPMENT AND REGIONAL PLANS

ECONOMIC COSTS – CURRENT DOLLAR ESTIMATES (MILLIONS)
ITEM INDEPENDENT PLAN “A” PLAN “B” PLAN “C” PLAN “D”
Capital Cost 6,153 4,533 5,970 5,413 4,242
Operating Costs:
Thermal Costs 2,321 1,700 1,739 2,155 1,884
Hydro Costs 1,912 1,950 2,147 1,960 2,567
Purchases 811 1,457 1,781 1,923 1,605
Total Operating Costs 5,044 5,107 5,667 6,038 6,056
Unserved Energy Costs 48 1 7 7 1
Total System Costs 11,245 9,641 11,644 11,458 10,299
Net Present Value @ 10% 5,000 3,997 4,954 4,767 3,920
Savings of
Regional Plans
Current Dollars - 1,604 (399) (213) 946
Net Present Value - 1,003 46 233 1,080
37
14.20newD
TABLE 6.2
PLAN “D” REGIONAL INTERCONNECTION PROJECT

ECONOMIC COSTS – CURRENT DOLLAR ESTIMATES U.S.$ millions
Year Capital Thermal Hydro Purchased Total Total Present Worth Value
Cost O&M O&M Power Operating Cost System Cost (at 10%)
Factor $M
1992 138 138 1.0000 137
1993 197 197 0.9091 179
1994 52 52 0.8264 43
1995 18 71 69 28 168 186 0.7513 140
1996 1 76 73 31 180 181 0.6830 123
1997 170 78 80 33 191 361 0.6209 223
1998 180 83 84 36 203 383 0.5645 216
1999 285 87 91 26 204 489 0.5132 251
2000 405 94 96 40 230 635 0.4665 296
2001 592 98 102 45 245 837 0.4241 355
2002 880 97 109 108 314 1,194 0.3855 460
2003 621 105 116 113 334 955 0.3505 335
2004 484 108 212 122 442 926 0.3186 295
2005 188 117 223 148 488 676 0.2897 196
2006 32 127 235 157 519 551 0.2633 145
2007 0 168 247 141 556 556 0.2394 133
2008 0 182 262 149 593 593 0.2176 129
2009 0 191 276 195 662 662 0.1978 133
2010 0 202 291 233 726 726 0.1799 131
TOTALS 4,241 1,884 2,566 1,605 6,055 10,298 3,920
38
14.20newD
TABLE 6.3 CASH FLOW (Millions US$) AND RATE OF RETURN
Alternative 3B: 2 x 220MW Coal (US$1750/kW including transmission)
1 2 3 4 5 6 7 8 9 10
Year
Plant Cost 759

30% equity 227
70% debt 531
Principal 76 76 76 76 76 76 76
Operating expenses
Fuel costs 36 37 38 39 40 41 43
O&M costs 22 23 24 24 25 26 27
Taxes & Insurance 6 6 6 6 6 6 6
Interest 80 68 57 46 34 23 11
Revenues
Energy 219 225 232 239 246 254 261
Capacity 34 34 34 34 34 34 34
Investment Costs 228 76 76 76 76 76 76 76

Operating Costs 144 134 125 115 105 96 87
Total Revenue 253 259 266 273 280 288 295
Gross Income -228 33 49 65 82 99 116 132
Financial IRR% 23.33
Notes: 1. Fuel/O&M costs and energy revenues escalated at 3% pa from year 1.
39
14.20newD
It is assumed that the project is financed with 30% equity and 70% short-term debt, with
the latter refinanced at the end of a 3-year construction period for the balance of the
contract when an agreed revenue stream will apply. Refinancing is taken as a loan repaid in
seven equal instalments at an interest rate of 15%, which is also taken as the comparison
rate for the DCF calculation.
6.3.2 Cost Streams for Least Cost Ranking
The cost streams for transmission alternatives are aggregated with the system costs, say
from a development plan, to establish a set of costs for the overall least-cost development.
Cost estimates are converted to economic costs using shadow prices for the major cost
components, and all costs are expressed in constant or current currency (say US$) terms
according to which is most appropriate. The economic comparison is between the savings
from interconnected operation and the costs of establishing the interconnection.
The methodology employed to rationalise costs to a common base or currency relies on the
theory that the economic costs of major projects, in the currency of the country developing
the project(s), are greater than the shadow prices (or costs to the economy) due to the
foreign exchange rates and cross-border prices associated with the project(s). On this
premise the economic costs presented in the spreadsheet analysis of Plan D (Table 6.2) of
the regional (interconnection) study were multiplied by either:
• the “official exchange rate” at the time the expenditure was incurred when in current
dollars, or
• the “official exchange rate” in base year (1992) when presented in constant (1992) US
dollars.
The above approach is often used in dealing with the scarcity of different ‘country’ labour,
capital and foreign exchange resources. Market imperfections due to the availability of these
resources are accommodated in project analysis by applying shadow prices instead of local
market prices in the measurement of costs and benefits. In some circumstances market
prices will correctly reflect the scarcity value of resources, but other project or policy
constraints often operate to divorce market prices and economic values. For example, trade
taxes and duties that cause a divergence between the value of fuel at domestic and
international prices. Shadow prices are thus applied to correct for such distortions, by
ensuring the efficient allocation of resources despite the distortion.
Constant prices are often referred to as ‘real prices’ because they ignore inflation, whereas
current prices include inflation. It should be noted that inflation is not normally considered
in NPV analysis, subject to relative prices remaining unchanged, as the same value of NPV
will be calculated whether the calculation is made in real or inflationary terms. Analysis
using constant (real) prices are often favoured by economists (and politicians), to
demonstrate economic efficiency through the control of escalation, that is, price rises
resulting from resource depletion, increased demand, etc. Price control in these
circumstances is often by regulation.
In contrast to relative price changes, inflation signifies that the general purchasing power of
money is declining. It is thus often necessary to conduct analyses in current prices in order
to show that a project alternative is both financially viable as well as economically efficient.
Current prices, for example, will more directly demonstrate demand/price elasticities and
thus the ability to meet the minimum revenue requirement to service debts and meet
stakeholder obligations out of cash flow.
40
14.20newD
Regardless of the context, scenario or option in which a transmission interconnection is

considered, the investment choice depends not only on the costs directly associated with the
transmission project, but rather on the way in which the project affects the costs of the
entire power system. This means that the method of appraisal must be on the basis of the
total system lifetime costs, calculated for each year of the planning period, and taking into
account:
• investment costs of any new supply capacity - say via a transmission link;
• the total operating cost of the supply system, with (generating) plants operating in merit
order, according to the shape of the monthly/annual total load duration curve; and
• any other new supply resource costs associated with all the capacity on the supply system
in each year of the planning period.
Spreadsheet computer models or more complex production cost programs are often used to
schedule cost streams and enable operating cost minimisation according to economic
dispatch criteria, which may be through a Power Pool or by a major Utility.
For economic comparisons to be made as a project-specific exercise say for a private

investment initiative, it is first necessary to establish the likely revenue to be set against
investment and operating costs. In this exercise the avoided cost approach has merit
because it enables estimates to be made of revenues to be gained from sales into a
wholesale electricity market. In a deregulated network, market simulation techniques can be
used to forecast revenue streams.
6.3.3 Avoided Cost Concept
Avoided cost can often be regarded as the cost of self-sufficiency, and particularly so in
assessing options for the purchase of firm power from a remote source. Avoided cost is
therefore a useful mechanism in assessing the merits of bulk power transfers over long
distances, such as schemes to deliver energy from Inga to markets in Europe and South
Africa.
The theoretical basis for avoided cost derives from marginal cost determination of changes
in the use of electricity. In this respect, the cost of interconnections is considered a marginal
cost in the longer-term. Thus benefits or avoided costs for a change that is expected to last
for several years are determined by calculating the cumulative net present value of the
annual avoided cost in terms of capital investment and operations to a total systems base
over the useful life of the proposed interconnection.
Avoidable resources and their costs are defined as those which a Buyer would otherwise
have to provide to meet the next increment of firm capacity and energy demand if he did
not make firm power and energy purchases from a Seller. They are thus the marginal costs
of supply hitherto used to determine tariffs and revenues. Utilities levelise these costs by
amortising the generation investment (fixed) cost over 30 years, usually at a 20% fixed
charge rate and 70% capacity factor, and adding the variable (fuel + O&M) levelised costs.
Using this approach, the levelised avoided cost becomes the marginal price for the purchase
of firm power and energy.
Marginal cost analyses should be undertaken to determine sound pricing policies, based on
willingness and ability to pay, as these methods take account of the next increment of power
and energy to meet demand. It follows that demand forecasts must be realistic to ensure
41
14.20newD
that tariff structures are practical for sustaining the revenues necessary to achieve the
required rate of return.
6.4 Coping with Uncertainty
Inflation, as already noted, is not normally considered in NPV analysis, subject to relative
prices remaining unchanged, as the same value of NPV will be calculated whether the
calculation is made in real (constant price) or inflationary (current price) terms. However,
relative price stability is a factor that should be reviewed in undertaking sensitivity analyses
to test the ranking for robustness, when considering the risks and uncertainties of economic
choice.
Typically, these risks and uncertainties are mostly economic and include the values of
discount rate, inflation, border prices, construction and fuel costs, and project delays. In the
open market of today, the variation in demand and price add to the uncertainties. In
confirming least cost options, sensitivity analyses are often extended to include
non-economic costs and benefits that stem from concerns over the environment or
institutional changes in electricity supply. However, the basic NPV calculation should
incorporate the best estimates of variables and parameters that determine cost and benefit
streams. Risk analysis provides a better basis for judging the relative merits of alternative
projects – but it does nothing to diminish the risks. Some risks can be reduced by further
investigations, or by a flexible design of the project, but these measures may impose
additional costs.
Non-discounting methods, such as the pay-back method are sometimes considered in high-
risk situations, on the premise that the project that returns its outlay sooner is to be
preferred since there is less risk overall attached to it. The pay-back period for each project
can be determined by accumulating the cash flow items. As soon as the cumulative cash
flows become non-negative the project has ‘paid back’. However, the pay-back method is
discredited by its two major characteristics, namely, the timing of cash flows within the pay-
back period are ignored, as are cash flows outside what turns out to be the pay-back period.
A wide range of techniques for coping with uncertainty is to be found in the literature.
These include sensitivity analysis, scenario planning and probabilistic economic assessment.
Depending on the circumstances being examined, it may be necessary to generate a set of
separate year-by-year plans, each based on a different set of assumptions. Different
scenarios could be created for other values of discount rate, demand growth, fuel price, etc.,
as the weakness of the yield rule is that it considers only the rate of return on the project
and not the scale of profits.
Where there are uncertainties attaching to several input parameters, a technique of

probabilistic economic assessment may be appropriate. Statistical distributions are applied
to input parameters, and the result is a statistical distribution of expected net present value
or other economic parameters. However, this can be a complicated procedure and should
be applied with caution [55].
6.5 Non-Quantified Benefits and Costs
Any overall appraisal of the HVDC transmission proposal should also take account of any
costs and benefits, which have not been expressed in monetary terms. The non-quantified
benefits can include aspects such as energy source diversity, increased flexibility in system
operation, improved resilience against widespread failure, access to larger markets etc. By
42
14.20newD
breaking a system into non-synchronous parts, a HVDC link reduces the risk of spread of a
cascade event. Even if rare and unpredictable, such a benefit deserves consideration. Non-
quantified costs may relate to environmental impact, loss of autonomy, increased
technological dependence, etc.
The environmental effects of energy development need to be carefully evaluated, and in

comparing alternatives it is now usual to extend the analysis to include the non-economic
costs that stem from concerns about the environment. A particular benefit of
interconnection associated with hydro schemes is the realisation of integrated operation that
usually permits mitigation of the effects of energy production from fossil sources, in addition
to savings in thermal fuel costs. Such savings contribute to the reduction of carbon dioxide
releases into the atmosphere, replaced by hydro as a renewable energy.
Perhaps more importantly, the conjunctive operation of run-of-river and storage hydro plants
via interconnection enables significantly more energy to be produced from a given
hydrological scenario. This can apply within and/or between river systems, as demonstrated
in studies of the Zambesi and its major tributaries, and can be coupled with opportunities of
conserving water for other uses, for example during a drought. Such multi-purpose
schemes, applying the conjunctive operation principal via interconnection, provide a special
challenge to economic analysis. But they also provide opportunities to reduce the scale of
storage in a new hydro project, thereby limiting the inundation of agricultural land and
destruction of habitat, as well as the cost.
7. OTHER FACTORS AFFECTING A DECISION
This section addresses a variety of considerations and risk issues, which might influence a
decision whether or not to proceed with a transmission scheme.
7.1 Risks
Any major project involving large financing requirements and lengthy lead times imposes
some degree of risk. Managing the risk over the time from project conception to commercial
operation and beyond is the task of the purchaser or developer. This can be accomplished
by having access to experience and providing for several loops in the feasibility process so
that a refining of technical and economic assessments is achieved. The operational impact
of the project on the power system(s) must be well studied to ensure system performance is
not degraded. There is the temptation by the purchaser to cut costs throughout the
process. Such risks may be a false economy if poor performance or delays in reaching
commercial operation are the consequence. A summary of the risks which are faced when a
new DC link is being put into service include:
7.1.1 Technical Risks.
• The project is not optimally sized or located. Compromises are necessary and speed of
installation is often more crucial.
• The use of new technologies. Even when new technologies offer cost benefits and
improved performance, users may opt to purchase proven configurations or traditional
equipment if they are not confident in being first to try the innovation.
• An inadequate specification may cause extra expense at later stages in the construction
or during operation. There are temptations to cut costs early in the feasibility and
implementation phases by limiting the time for preparation of the specification and
43
14.20newD
contract or limiting the expertise to evaluate the work undertaken in these phases. The
inadequate specification and control may result in a poorer ultimate performance.
• Insufficient training of technical and operational staff and for maintenance procedures.
• The selection of the spares holding should be guided by a reliability study. It requires a
balance to minimise down time during operation, but to spend no more than a minimum
on parts which may never be needed.
7.1.2 Economic Risks
• During the course of the project, unforeseen increases in costs may arise due to pricing
changes, change orders, exchange rates or interest rates or inadequate controls in the
contract;
• The scope of work, both in the physical content of items included and in the commercial
warranties or constraints applied, can affect price variations between the time of
estimating and that of a call for binding tenders. Open and flexible market conditions
apply;
• Escalation may be significant in contracts that extend over a long time or where economic
stability is weak;
• Open electricity markets will cause uncertainties in the need for transmission capacity and
therefore to income from the project. Instead of thinking of the link alone, its value to
the whole network(s) should be considered: in terms of energy transmitted, capacity
added and ancillary services provided;
• There may be risks which erode benefits, e.g. due to fuel price changes, or other
investment decisions which reduce transmission utilisation;
• The owner or user may be exposed to a regulatory risk, if he is prevented from charging
enough to recover the investment;
• If the financier dictates the selection of the lowest cost bid against the recommendation
of the technical staff, he does so at the risk of some increase in technically caused costs
further into the contract or over the lifetime of the plant.
7.1.3 Social and Political Risks
• There are often environmental concerns raised when overhead transmission or cable is
required. Public concern over right-of-way, field effects and visual impact must be dealt
with and appropriate compromises and additional compensation or costs, such as
underground cable in place of overhead line, may be required to limit delay in the project;
• Workforce disruptions with either the supplier, the purchaser or the transporter;
• External disruptions such as theft, undersea cable damage by anchors or fishing
apparatus, accidents or sabotage in transportation;
• Political instability may erupt and restrict or destroy the construction or prevent the use of
the line or the receipt of the consideration.
7.2 Dependency and Interdependency
A primary motivation for interconnection is the purchase of firm energy for a contract period,
to enable demand to be met within a least-cost investment strategy. However, political
problems may arise if dependence is for an extended period or the purchase is a significant
portion of total demand. Each participant loses some autonomy, and the energy deficient
country in particular may have self-sufficiency concerns.
Participants who become net importers of electrical energy under such a strategy may be
said to support the development of their neighbours' energy resources to the exclusion of
44
14.20newD
their own, with associated political risks. While such concerns should be recognised in
evaluating alternatives, decision-makers must also recognise the risk of breakdown through
insufficient investment or the extra costs of electricity supplied by other than the least-cost
solution.
Nevertheless, imported energy is always a valid option in achieving least-cost investment. A

practical compromise is to plan to preserve a reasonable power balance by setting limits on
the dependency of any one country. A good "rule of thumb" is for each country organisation
to be responsible for at least 80 percent of its power and energy needs, and for load
dispatching in its own network.
Financial dependence applies when the power export revenues accruing to a supplier
country are necessary for the financial viability of its power sub-sector. This may represent
substantial risks for the supplier country, and particularly so if revenues must be in a
stronger and scarce third currency. Where financial dependence is perceived as a problem,
special lending for the importer country or possibilities for counter trade may be an effective
means of resolution. In any event, financial dependence is valuable as a deterrent to the
exercising of supply restrictions, provided agreements are equitable and controlled through
appropriate contract.
Related to the issue of dependency there may be, in some countries, an employment issue.
If interconnection is an alternative to the construction of a new power station with its
associated employment implications for construction, fuel purchase and operation, there may
be opposition to the interconnection project on the basis that it will destroy or export
employment opportunities. Interconnection may also affect the development of indigenous
resources and related issues of employment and energy dependence. In these
circumstances it will be necessary for the promoters to demonstrate the macro-economic
and other benefits of the scheme.
Another related issue is “nuclear”. Anti-nuclear campaigners in a country with no nuclear

generation may oppose interconnection with a system that incorporates nuclear generation
on the basis that the interconnection will permit the importation of “nuclear” electricity, such
generation being, in the objectors’ opinion, not acceptable. When a submarine HVDC link
between the Republic of Ireland and England and Wales systems was under consideration,
this particular objection surfaced in Ireland. Open market access may render this argument
publicly invalid.
7.3 Market and Institutional Arrangements
Whenever two electric power systems are to be interconnected it is necessary to establish

joint structures to oversee operation and management of the interconnected system and to
provide mechanisms for the trading of electricity.
The promotion of a scheme of interconnection is highly dependent on the market and

institutional arrangements governing the two systems. This can be exemplified by two
situations in Europe.
The Republic of Ireland-England and Wales submarine link mentioned above appeared
attractive and worthy of further consideration when examined in the 1980s by two
monopolistic, centrally-planned utilities - ESB of Ireland and CEGB of the UK. Following re-
structuring of the electricity supply industry in England and Wales, the scheme was no
longer attractive because the lack of central capacity planning in England and Wales meant
45
14.20newD
that the capacity benefit was less obvious and the associated revenue stream less
predictable. Also, it was not clear that the capacity and marginal trading benefits would in
fact accrue to the scheme’s promoters.
On the other hand, it would appear that the liberalisation of electricity markets in Europe has
created the circumstances favouring energy trading and the promotion of HVDC links from
Norway to Germany, The Netherlands, and possibly to the UK. Similarly, in Australia the
National Electricity Code includes provisions for the use of entrepreneurial investment in
interconnections and this has lead to at least one such HVDC project.
In general, a combined generation and transmission scheme to deliver electricity at a lower

cost than alternative sources can probably be promoted satisfactorily irrespective of market
structure. However, a scheme whose viability is dependent on the “traditional”
interconnection benefits of capacity saving, reserve pooling and marginal trading may be
difficult to promote in an unbundled environment, as it is difficult to ensure that the benefits
of the scheme can be realised by the promoters.
7.4 HVDC as a Barrier
Some HVDC interconnections between systems with the same frequency, especially back-to-
back schemes, may be perceived as a barrier to closer integration of the systems, or even
the countries. In some instances these barriers are seen as desirable (because they stop the
spread of voltage collapse) by at least one of the areas concerned, such as between Texas
and its neighbours in North America. Similarly the connections between the grid systems in
India by HVDC have benefits recognised by the parties concerned.
However, in the case of interconnections with countries where closer ties and closer
integration are desired, back-to-back links may be viewed unfavourably. The connection of
the CENTREL countries (Poland, Czech Republic, Slovakia and Hungary) to the western
European UCPTE system was seen as a symbol of the increasing acceptance of these
countries by Europe in a political as well as an energy sense. The development of the DC
connections further east to the countries of the Unified Power System (UPS) through HVDC
back-to-back links might be seen by some as maintaining a barrier against integration of
these countries, independent of any doubts about the wisdom of creating a yet larger single
system. Furthermore, the view has been expressed that HVDC links, where there are no
geographical bottlenecks, give rise to passage monopolies that are detrimental to companies
not involved in the ownership of the link and may restrict their freedom to do business.
On the other hand, the sheer size of a large AC system poses problems; e.g. Spain to
Poland, being one electrical system since 1996, does experience stability questions due to a
slow oscillation (about 0.2Hz) between the two extremes. Very difficult circumstances can
propagate or cascade through an AC system. The eastern and western systems of North
America remain unsynchronised, not for any political reason, but for this technical question
of size.
7.5 Supplies to communities en route
It may be that the communities or countries over which a transmission line passes require a
supply of energy. If this is so the cost of the whole may be increased. Alternatively is it
feasible to add a conventional supply on the same towers? Overlay an AC ripple current on
the DC conductor? Apply 50Hz or 60Hz on the overhead ground wire? Add a mini-DC
scheme on the same towers with ground return? Or simply provide a supply which bears no
46
14.20newD
relationship to the DC project, but is administered integrally with the HVDC scheme?
Whatever the technical solution, a supply may be essential to the acceptance, and therefore
to the containment of the risk to the project as a whole.
Technical options for local supplies from HVDC transmission lines are discussed in Appendix
I.8.
8. CONCLUSIONS
1 The economic assessment of new transmission includes the definition of several

technical and operational items, such as:
a) Operational criteria for the link, mainly if its revenue-earning operations are peak
power transmission or energy transmission or a combination,
b) The reliability criteria related to its security and availability, and
c) Different scheme configurations and characteristics.
2 The choice of the technology to be used in a long distance transmission is carried out
by developing the economic assessment for each possible alternative, such as HVDC,
EHV-AC or the independent development of generation in each area. It is important
for a fair assessment that the comparison be made using the same criteria.
3 For some applications, HVDC technology is the unique choice, as in long submarine
cable transmission and interconnection between asynchronous AC systems. For
these, an economic assessment is also valuable to establish the required investment,
the cost of the energy to be transported, the net income to the owner, etc.
4 In addition to the technical and operational inputs defined above, an economic

assessment requires the definition of financial and economic parameters such as:
a) procedure to be used in the evaluation of the different schemes;
b) economic evaluation of the station and transmission line energy losses;
c) information on the cost of equipment, transmission line, station construction and
services, materials, etc;
d) costs related to the environmental aspects of the alternatives;
e) costs of the operation and maintenance of the installations;
f) technical, economic and environmental benefits that the choice of DC may bring;
and
g) financial parameters.
4.1 The decision on the procedure to be used should consider the nature of the owner
(private or state owned, merchant developer, also known as Independent Power
Producer (IPP), etc).
4.2 As any other project, the implementation of a transmission link involves technical and
financial risks to the investor, that should be considered in the feasibility study of the
transmission link. In the case of an HVDC link, all care should be exercised to match
estimates made before tender to realistic tender values. One way to improve this is
at the planning stage to request the prospective station manufacturers to prepare
estimating prices based in a preliminary specification (mini-specification) prepared by
the purchaser and/or its consultant.
47
14.20newD
5 The Guide has been prepared to help the user in the preparation of an economic
assessment of an HVDC link by discussing in detail the above mentioned aspects and
many other pertinent items. It also includes some considerations on specific aspects
of interconnections in general.
ACKNOWLEDGEMENT
The Working Group acknowledges CIGRÉ WG 37.22’s Working Report of August 1998 from
which some suggestions have been incorporated here.
APPENDIX I - HVDC CONFIGURATIONS
Different types or configurations of an HVDC link are briefly presented here. Some basic
definitions for HVDC converter units are given first:
A 6-pulse converter unit consists of a three-phase full wave converter bridge with six
thyristor valves and the associated transformer. It is the basic converting element and is
operated either as a rectifier converting AC current into DC or as an inverter converting DC
current into AC.
Figure I.1
6, 12 and 24 pulse convertors
A 12-pulse converter unit consists of two three-phase bridge arrangements connected in

series on the DC side with their respective transformers, one with a star (wye) connected
winding, the other with a delta connected winding, producing a phase difference of 30°.
This is the most commonly used arrangement of a converter pole.
A 24-pulse converter unit consists of four three-phase bridge arrangements connected in

series on the DC side with their respective transformers, producing a 15°-phase difference
one from the other. This is not economical today and is not discussed further.
Typical diagrams of 6, 12 and 24-pulse converter units are presented in Fig. I.1.
48
14.20newD
I.1 Monopolar HVDC Systems
Monopolar HVDC systems have either ground return or metallic return.
A Monopolar HVDC System with Ground Return consists of one or more 12-pulse
converter units in series or parallel at each end, a single conductor and return through the
earth or sea, as Fig. I.2 presents. It could be a cost-effective solution for an HVDC cable
transmission and/or the first stage of a bipolar scheme. At each end of the line, it requires
an electrode line and a ground or sea electrode built for continuous operation. Thus,
possible interference with other structures, installations or systems and magnetic field
effects may be considered.
~ ~
T7863
Figure I.2
Monopolar HVDC system with Ground Return
A Monopolar HVDC System with Metallic Return usually consists of one high voltage
and one low voltage conductor, but could consist of two conductors at + and – half voltage,
as Fig. I.3 presents. The neutral is tied at one converter station to the station grounding
grid or to the associated ground electrode and at the other station through a capacitor or an
arrester to its grounding grid. Monopolar is used either as the first stage of a bipolar
scheme, avoiding ground currents, or when construction of electrode lines and ground
electrodes constitutes an uneconomical solution due to a short distance or high value of the
earth resistivity. If this arrangement is the first stage of a bipolar scheme, the neutral
conductor could be insulated from the high voltage at this stage. It may be more
economical, however, to build both conductors at the first stage and run them in parallel at
high voltage with ground return to save losses.
Both configurations require one or more DC smoothing reactors at each end of the HVDC
line located on the high voltage side, or on the low voltage side if the performance is
acceptable, and DC filters in the case of an overhead line.
49
14.20newD
~ ~
T7863
Figure I.3
Monopolar HVDC system with metallic return
I.2 Bipolar HVDC Systems
A Bipolar HVDC System consists of two poles, each of one or more 12-pulse converter
units, in series with electrode lines and ground electrodes at each end. There are two
conductors, one with positive and the other with negative polarity to ground for power flow
in one direction. For power flow in the other direction, the two conductors reverse their
polarities. It is a combination of two monopolar schemes with ground return, as Fig. I.4
presents. With both poles in operation, the imbalance current flow in the ground path can
be held to a very low value.
+
~ ~
~ ~
Figure I.4
Bipolar conventional DC system, one valve group per pole
This is a very common arrangement with the following operational capabilities:
50
14.20newD
• During an outage of one pole, the other could be operated continuously with ground
return.
• For a pole outage, in case long-term ground current flow is undesirable, the bipolar
system could be operated in monopolar metallic return mode, if certain DC arrangements
are provided, as presented in Fig. I.5. Transfer of the current to the metallic path and
back without interruption requires a Metallic Return Transfer Breaker (MRTB) and other
special-purpose switchgear in the ground path of one terminal. When a short interruption
of power flow is permitted, such a breaker is not necessary.
• During maintenance of ground electrodes or electrode lines, operation is possible with
connection of neutrals to the grounding grid of the terminals, but only if the imbalance
current between the two poles is held to a very low value.
• When one pole can not be operated with full load current, the two poles of the bipolar
scheme could be operated with different currents, supposing that both ground electrodes
are connected.
• In case of partial damage to DC line insulation, one or both poles could be continuously
operated at reduced voltage.
• In place of ground return, a third conductor can be added end to end. This conductor
carries unbalanced currents during bipolar operation and serves as the return path, when
a pole is out of service. The neutral of one terminal could be grounded, while the other
would float or be tied to the grounding grid through an arrester or a capacitor.
51
14.20newD
~ ~
M RTB
~ ~
Figure I.5
Bipole with monopolar metallic return for pole outage
I.3 Point-to-Point HVDC Overland Systems
Point-to-Point HVDC Overland transmission is the common application of power transmission

over long distances or asynchronous interconnection of two power networks.
For an evaluation process the following parts, consisting of a Point-to-Point Overland

System, may be considered:
• Converter Stations at the two ends of the transmission.

• HVDC Overhead Line.
• Electrode Lines, which could be medium voltage insulated overhead lines, connecting the
neutral point of the converter stations with the associated electrodes. It is noted that, if
the route is the same for a certain length, HVDC line and electrode line could be on the
same towers.
• Electrode Stations, which could be located in the ground or sea shore
Figure I.6 shows one end of a conventional bipole with filters and two convertors per pole at
a generating station. A single common busbar is shown, but in some cases this may be
impractical due to a high fault level. The busbar can be split into two or more sections.
Alternatively it is possible to have no busbar, with one or two generators “unit-connected” to
each valve group. This arrangement has advantages if the generators are built to absorb
harmonics and to operate at variable speed, but as stated in Section 2.3 no such scheme has
been built.
52
14.20newD
Aux
T7868
Figure I.6
Conventional Bipole termination
I.4 Point-to-Point HVDC Submarine Systems
For sea crossings longer than 50km of conventional AC cable, or 100km if XLPE or other new
cable is used, the use of AC is virtually impossible, as the capacitive currents become too
high. This is the application of DC for submarine transmission.
The main parts of a Point-to-Point HVDC Submarine System are:
• Converter Stations at the two ends of the transmission.

• Land Cable sections,
• Submarine cable section(s),
• Transition points to connect cables and lines, perhaps with disconnectors
• Electrode Lines.
• Two marine or shore electrodes, one anode and one cathode
• Overhead Line section(s); if there are overhead line sections, DC filters are required.
Technical limitations in today's cable design and manufacture dominate in the optimisation
of an HVDC cable transmission. For a solid type submarine cable or a mass impregnated
one, voltage and current limitations are in the order of 450kV DC and 1500 A respectively,
whereas there is no limitation for the distance. For a pressurised light oil filled cable, today's
voltage and current limitations are 600kV DC and 3000 A respectively, whereas the distance
must be below 80 km. There are certain advantages in the solid type choice due to
environmental and maintenance issues.
Another application is the use of long HVDC Underground Cables in densely populated urban
areas. In these areas, construction of overhead lines becomes expensive. The choice of
HVDC versus the AC alternative has the advantage of low losses, but the disadvantage of
53
14.20newD
the space required for the converter stations though this may be reduced in the case of
voltage-sourced convertors.
I.5 Back-to-Back HVDC Links
Back-to-Back HVDC Links are special cases of monopolar HVDC interconnections, where
there is no DC transmission line and both converters are located at the same site. For
economic reasons each converter is usually a 12-pulse converter unit. The valves for both
converters may be located in one valve hall. The control system, cooling equipment and
auxiliary system may be integrated into configurations common to the two converters. DC
filters are not required, nor are electrodes or electrode lines, the neutral connection being
made within the valve hall. Figure I.7 show different circuit configurations, with and without
a smoothing reactor.
~ ~
~ ~
T7867
Figure I.7
Back-to-back DC circuits
In these systems the optimisation of the whole station for the voltage and current ratings for
a given power rating, to achieve the lowest life cycle cost is straightforward. Generally, the
DC voltage rating is low and thyristor valve current rating high in comparison with HVDC
interconnections with overhead lines or cables. The main reason is that valve cost is much
more voltage dependent, as the higher voltage increases the number of thyristors. A low
voltage tertiary winding can be built in to the converter transformer for the AC filters and
compensation. Smaller reactive power switching steps can thus be achieved [36].
A large Back-to-Back HVDC system could be comprised of two or more independent links so
that the loss of one convertor unit will not cause loss of full power capability.
Considering an HVDC transmission between two asynchronous power systems, the optimum
location of the terminals should be considered. The choice is between HVDC overhead lines
and a Back-to-Back link at the border of the two existing grids. Back-to-Back links have a
54
14.20newD
cost benefit considering terminal costs only, provided the capacity of the existing grids can
handle the new load. If system reinforcements are necessary, using a Back-to-Back link with
a long AC feeder might assist.
1.6 AC and DC in Parallel
Environmental restrictions are being imposed on the expansion of power systems and
especially on the construction of new transmission lines. As an alternative to new lines, the
enhancement of power transmission capacity of the existing lines may be considered. One
way of achieving this is by converting an existing AC line to DC, with possible alterations to
the insulators. For double circuit AC lines, conversion of one AC circuit to a DC leads to a
parallel operation of AC and DC lines on the same transmission towers. Studies performed
for the close coupling of two parallel AC and DC lines running on conventional AC towers
show no adverse effect on the operation of the DC link [21].
I.7 Switching Possibilities on DC Side
For bipolar systems several possible DC switching arrangements may be provided, which
increase the availability of the overall HVDC interconnections:
• Capability of connecting any line to any converter pole or to neutral through an

arrangement with three DC buses (two pole buses and a neutral bus), as figure I.8
presents.
P N P
D C Line
Fully Insulated
Spare
(O ptional)
D C Bus
T7869
Figure I.8
DC switching scheme for a fully insulated spare conductor
55
14.20newD
Figure I.9
Valve Groups in parallel
T7870
Figure I.10
Valve Groups in series
• Capability of connecting the two valve-groups in parallel, as figure I.9 presents. It is

noted that the 12-pulse converter unit of each group would need to be insulated for full
line voltage. This is an alternative staging of a DC link, which is intended to minimise the
line losses at the early stage.
• Capability of connecting the two valve-groups in series, as figure I.10 presents. It is
noted that the 12-pulse converter unit of each group would need to be rated for full line
current. This is an alternative staging of a DC link where the independent switching by
by-pass of one group while continuing the passage of current in the other is important.
56
14.20newD
• Capability at the junction of overhead and cable sections of a bipolar DC line to connect any of the
overhead line poles to any cable.
• For more than one bipolar line, paralleling of converter poles on to one line or intermediate
switching of the lines may also be provided.
There are other possible switching arrangements also.
I.8 Multi-terminal HVDC Systems
Multi-terminal HVDC (MTDC) operation is a viable strategy that can be incorporated into planning
alternatives. The potential applications along with the possible configurations are discussed in [23].
Three main possible configurations are mentioned in this chapter, the first two refer to parallel-
connected 12-pulse converter units, rated for full direct voltage and the last one to series-connected
12-pulse converter units, rated for full direct current. More explicitly the following conceptual
classifications are:
• Radial MTDC System, where each converter station is connected to a single DC line and for
some of the distance the energy flows through a common DC line. No part of the DC system can
be disconnected without significant energy flow change between the AC systems.
• Meshed or Ring MTDC System, where each converter station is connected to more than one
DC line. Any part of the DC system can be disconnected without energy flow change between the
AC systems.
• Series Connected MTDC System, where all the convertor stations are connected in series to a
common ring shaped DC line. No part of the common DC line can be disconnected without
interrupting energy exchange between the AC systems.
In practice the few multi-terminal DC systems that exist are radial, i.e. there is a tapping point on a
point-to-point two terminal system. Variants that can be considered are
• Positive and negative pole inverters at separate locations, as figure I.11 suggests
• Rectifiers or inverters at separate locations with parallel connection, as figure I.12 presents
• A series connected small tap, as figure I.13 presents [26
T7873 T7873
Figure I.11
Pole inverters at separate locations
57
14.20newD
T7874
Figure I.12
Parallel-connected multi-terminal
T7875
Figure I.13
Series-connected small tap multi-terminal
58
14.20newD
I.9 Intermediate Supplies
Power taps to feed small Intermediate Supplies improve the attractiveness of

transmission, mainly from the benefits to the economic development of the regions as a
whole. These taps are distinguished from the Multi-terminal stations both by their field of
application (feeding small loads or isolated networks) and by their main requirements
(variable size, low impact on the main HVDC transmission system reliability, no interference
with the main system control as each draws the current it requires).
Figures I.11, I.12 and I.13 suggest means of feeding small intermediate supplies. In
particular, series HVDC tap schemes, are postulated in [26]. The development of Voltage-
sourced convertors using IGBT devices for DC may be suitable to supply intermediate loads.
There may be upper limits of the DC voltage for this technology [41].
APPENDIX II - EXPERIENCE REVIEW
Power systems develop in many different ways and some examples of the variety of reasons
for using HVDC are given below. A common factor to be observed in these examples is that
both an economic case and the corporate will to complete the project are prerequisites to a
HVDC project. Where possible, economic comparisons are included in these descriptions.
II.1 Cross Channel Link
The 2000MW France-UK Cross Channel link between the networks of EdF and The National
Grid Company was fully commissioned in 1986 [49,53,54]. The economic argument in
favour of the link was primarily to gain capacity reserve by both parties to the exchange.
Each country could defer 1500MW of peaking generation, receiving that level of reserve
across the link. The cost to each country of the link was less than such generation. The key
performance target set at the link planning stage was to achieve 95% all year round
availability. This was recognised as being onerous and much effort was made to design
plant and develop submarine cable embedding techniques with this aim in mind. In
particular:
- Proven technology was a prerequisite of terminal plant and cable designs.
- The development of cable embedding techniques in the channel sea bed was also a pre-
requisite following the poor performance of the first Cross-Channel link (160MW) due to
frequent damage to the cables by shipping.
- Maintenance and spares provision was the subject of detailed analysis at the planning
stage to ensure the target availability would be met.
In the event performance has exceeded expectations and target availability levels have been
met. Since 1991 the 95% availability target has been consistently exceeded with values
between 95.8% and 97.9% and exchanges have exceeded 16TWh per year. The Cross
Channel link has thus had the highest load factor of any link in the world, an average of
99.4% in the period 1992-97. It will be clear that the link has been used as an energy
transporter (firm and secondary) in addition to the original incentive at construction of the
saving of new generation capacity.
59
14.20newD
In practice, the implementation hurdles were considerable in the days of government-owned

supply authorities in both countries. About 10 years were involved in the initial planning,
design feasibility, embedding technique development, local public inquiries and negotiations
between the transmission companies and their respective Governments. About 5 years were
involved subsequently in the construction and commissioning of the link with the full
2000MW entering service in October 1986.
The scale of investment for the 2000MW Cross Channel link was about 3GFF on the French
side and about £400M on the English side, both in money of the day, totaling about £700M,
and it has been said that subsequent energy trading repaid the investment every 2.5 years.
II.2 East and West Europe
In the late 1980s the need to exchange electrical energy between east and west Germany
was recognised. The power systems then operated to different criteria and a back-to-back
DC connection was planned. With the unification of the country in 1990, a single electrical
system became desirable and the back-to-back project was cancelled. Once prerequisites of
primary and secondary power-frequency control were equalised and the necessary AC
transmission was constructed, the east and west German systems were synchronised in
September 1995.
Power exchange between Western Europe (UCPTE) and the CENTREL system of Poland, the
Czech Republic, Slovakia and Hungary formerly took place through three back-to-back links
(Duernrohr 550MW, Vienna South-east 600MW and Etzenricht 600MW) [39]. Due primarily
to the political will to integrate these electrical systems with those of Western Europe and
also to the desire to take further advantage of the experience of interconnected operation
gained with the back-to-back links, these four countries were synchronised with Western
Europe in 1996. It was again necessary to establish common operational standards in
power-frequency control, and this was preferred to the alternative of different standards and
power flow control through the DC links, which are therefore no longer used.
II.3 Cahora-Bassa
The only HVDC link with a convertor inside South Africa is the link from the Cahora Bassa
dam in Mozambique to the Apollo substation near Pretoria. This link was built in the early
1970’s during the closing years of the Portuguese administration to transmit up to 1800MW.
The lines within Mozambique were, however, largely destroyed during the civil war from
1975 until 1993. No power was transmitted between 1980 and 1997. The lines, convertor
stations and power station have been rehabilitated, and the link was returned to service in
1998. Because an AC line was built from Cahora Bassa to Bindura in Zimbabwe to enable
Zimbabwe to benefit from Cahora Bassa as well, a parallel AC path was created to the HVDC
link via Zimbabwe and Botswana. This has necessitated the design of a new master power
controller as part of the rehabilitation of the control systems [56].
II.4 Quebec-New England
The province of Quebec maintains a power system that is not synchronous with its
neighbours. There are many reasons, the original being technical: to withstand a fault on
the Churchill Falls-to-Southern Quebec 735kV system. During such an event the generators
accelerate, creating a stress to stability, which must then be handled by the remaining
transmission after the fault is cleared. Though weakened by the removal of a faulted
section, the system must regain stability by transmitting more power than before the fault.
60
14.20newD
To achieve this, it is necessary to accelerate the load areas so that they catch up to the
generators. If the load areas were synchronised with other systems, too large a system
would be created to permit this acceleration and stability would not be regained.
Despite the separation, Quebec is a good example of the trend towards interconnecting of
networks. There are six HVDC installations (five can be connected simultaneously), with a
total capacity of 4000MW. Two of the installations form a long-distance link while the others
are back-to-back convertors.
Labrador
Quebec
James
Radisson
Bay
Ontario
Nicolet
Montreal
MA Sandy
Pond
Atlantic
USA
Ocean
Figure II.1
Quebec to New England HVDC Transmission
The Quebec-New England 1500km transmission link (Fig II.1) was built to provide a direct
link from hydro resources near Radisson in Quebec to Sandy Pond in south Massachusetts.
A convertor station at Nicolet, close to the load areas of southern Quebec, is able to operate
on the same transmission line in a three-terminal DC system, supplementing the existing
735kV AC connections from James Bay to southern Quebec.
The AC system that collects the generation at Radisson can be either isolated or
synchronized to the AC transmission system. If it is synchronised, dynamic islanding of the
generating units from the AC system is necessary in extreme disturbances where instability
61
14.20newD
of the AC system threatens also the power carried by the DC. The ability to invoke this
islanding mode alone made HVDC more attractive than the construction of AC transmission
feeding a back-to-back station at the interconnection point between Quebec and the US.
Operational flexibility is gained by running synchronised, but the power the DC can carry
must be limited to the ability of the AC system to recover from certain fault events. If the
AC system is degraded, DC power is further restricted. Load dynamics during post-fault
voltage and frequency disturbance also affect the DC loading. The policy subsequent to the
installation of Quebec-New England HVDC has been to improve the AC system ability so that
synchronised operation becomes the norm.
Early in January of 1998, the entire power system in southern Quebec was subjected to an
extraordinary ice storm. More than 125 transmission lines were damaged around the major
load centre. These breaks in the system led to cyclical load shedding operations and calls
for reduced electricity consumption during peak periods in the affected load area. Additional
contingencies were suffered by the south-west James Bay network preventing this
alternative transmission to the load centre. The demand could not be met without the
injection of power near the load centre. The HVDC link was used to supply Nicolet in multi-
terminal operation with the source system in islanded mode. Providing power to the inverter
at Nicolet played a key role in the recovery from this serious event, as the station was
equivalent to a local generating power source in the affected area. This operating mode
would have not been possible without the HVDC link and its multi-terminal nature.
II.5 India
The Indian power network is divided into five Regional Electrical grids, Northern, Western,
Southern, Eastern and NorthEastern. The development in each region is governed by the
availability of natural resources such as gas, coal and hydropower. Despite planning on the
self-sufficiency of each region, there is a trend towards the interconnection of four regions
(North, West, South and East) through inter-regional links, aiming to tap the benefits of
exchange of power from surplus to deficit regions. This has helped the optimisation of the
load–generation balance, fuel savings, reserve capacity reductions and emergency support
provisions.
In order to achieve a controlled power exchange between the regions and also to improve
the dynamic performance of the grid, the best mode of interconnection between two regions
was the asynchronous link through a Back-to-Back (BtB) HVDC link. Three BtB HVDC links,
connecting North with West at Vindhyachal (2 units of 250MW), West with South at
Chandrapur (2 units of 500MW), South with East at Vishakhpatnam (1 unit of 500MW) [57]
are in operation. A fourth at Sasaram (1 unit of 500MW) between East and North is
planned. Seeing the prospects of increased exchange of power between the regions,
augmenting these BtB links may also prove to be viable.
For bulk power transmission from a number of large thermal power stations, two long
distance HVDC schemes, one in the Northern region between Rihand and Dadri and one
between Chandrapur and Phadge in the Western region (both 1500MW ±500kV capacity)
are in service. A 2000MW ±500kV, 140km long HVDC project between Talcher in the East
and Kolar in the South is planned also. The prospects of a new HVDC link of 3000MW
capacity from an independent power producer, i.e. a non-utility station, in the Eastern region
purely on techno-commercial viability, are high. A few more similar lines are under study
and prospects for their being materialised are quite good.
62
14.20newD
APPENDIX III
MINI-SPEC INFORMATION FOR PRELIMINARY DESIGN OF HVDC SCHEME
Although an estimating price can be given for only a minimum of replies to the points below,
the more data that is given, the more precise a design can be.
PART 1 SCHEME TYPE
Select type of system envisaged:
1.1 Tick one: Point to point, two terminals
Multi-terminal, three terminals
Multi-terminal, more than three terminals
Back to Back
1.2 Select one from each line, adding distances to 1.2.1:
1.2.1 Overhead Line…………… km Submarine Cable……… km Both
1.2.2 No neutral path Earth return Metallic return
1.2.3 One pole One bipole More than one bipole
1.2.4 DC side switching between poles Required Not required

1.2.5 DC side switching between Required Not required
bipoles
1.2.6 DC side filters Required Not required
1.2.7 One convertor per pole Two convertors per pole, Two convertors per
series pole, parallel
1.2.8 Three convertors per pole, series Three convertors per pole, More (specify)
parallel
63
14.20newD
PART 2 AC SYSTEM FOR EACH TERMINAL
Terminal 1 Terminal 2
2.1.1 Voltage, nominal kV
2.1.2 Maximum continuous kV
2.1.3 Minimum continuous kV
2.2.1 Frequency, nominal Hz
2.2.2 Maximum continuous Hz
2.2.3 Minimum continuous Hz
2.2.4 Maximum short time Hz
2.2.5 Minimum short time Hz
2.3.1 Maximum short circuit level MVA
2.3.2 Minimum short circuit level MVA
2.4.1 Insulation levels, BIL kV
2.4.2 Insulation levels, SIL kV
2.4.3 Creepage distance mm/kV
2.4.4 Clearance distance mm/kV
2.5.1 Distortion and/or TIF limits for ac system

2.5.2 Which harmonics are to be assessed?
It is preferred that only harmonics of 6n ±1, where n is any positive integer, are assessed because
others are largely governed by the ac network.
2.5.3 Harmonic impedance data of system available?
Yes/No
2.6 What are the design constraints (number of feeders, security criteria, current user practice, etc) for
the ac switching stations?
2.7.1 Max var absorption MVAr
2.7.2 Max var generation at busbar MVAr
2.8 Max var bank switching size MVAr
2.9 Maximum permitted step voltage change %

arising from filter switching
2.10 Maximum temporary (<1sec) overvoltage kV

that existing equipment can withstand
2.11 Existing negative sequence voltage %
2.12 Existing harmonic voltages %
2.13 System solidly grounded?
2.14 Radio interference limits
64
14.20newD
Give results of ac system disturbance studies including transient voltage and frequency
variations.
PART 3 DC SYSTEM
3.1 Nominal power rating (at inverter output terminals) MW
3.2 Overload capability… MW

With ambient temperature °C
Duration seconds
3.3 Minimum power for continuous MW
operation
3.4 Power flow required in both directions? Yes/No
3.5 Describe time scale of any stages of development
3.6.1 Capitalised cost of losses, fixed $/kW
3.6.2 Capitalised cost of losses, variable $/kW
3.7 Target energy availability %
3.8 Target reliability, pole forced outages per year FOR
PART 4 GENERATORS (if applicable)
Location
4.1 Rating MW
4.2 Voltage KV
4.3 Harmonic absorbing capability
4.4 Power factor
4.5 Transient reactance, direct axis %
4.6 Sub-transient reactance, direct axis %
4.7 Maximum reactive power absorption MVAr
4.8 Generator transformers: rating MVA
4.9 Voltage ratio KV
4.10 Reactance %
4.11 Winding type
PART 5 AUXILIARY SUPPLIES for each terminal
5.1 Preferred auxiliary voltage kV
5.2 Frequency Hz
5.3 Source existing from AC busbar?
5.3.2 If existing, please indicate expected
reliability (outages/year)
5.4 Start up of auxiliaries manual or
automatic?
65
14.20newD
PART 6 CONTROLS AND TELECOMMUNICATIONS
6.1 Locations from which instructions will be received

6.2 Control philosophy: automatic or operator responsible?
6.3 Expected power controls: rate of change of power,
modulation to control frequency, etc?
6.4 Telecommunication type
6.5 Commissioning stage telecommunication to be
available? If so give type.
PART 7 SITES
Scope of Supply: Tick to include, X to exclude Terminal 1 Terminal 2

7.1 AC switchyard to DC connection
AC switchyard for AC filters
AC switchyard for outgoing lines and transformers
7.2 Electrode
Electrode Line
7.3 Civil Works
Engineering
Transport to site
Installation
Commissioning
Station communications facilities, giving type
PART 8 SITE CONDITIONS
8.1 Ambient temperature, nominal °C
8.2 Maximum °C
8.3 Maximum wet bulb °C
8.4 With coincident dry bulb °C
8.5 Minimum °C
8.6 Maximum wind speed km/h
8.7 Altitude m
8.8 Rainfall mm/yr
8.9 Isokeraunic levels Days/yr
8.10 Air pollution, ESDD
8.11 Seismic level, max vertical force g
8.12 Maximum loading gauge for ports m
and route
8.13 Maximum transport weight T
66
14.20newD
REFERENCES
A selection of the extensive literature on HVDC is referenced when specifically relevant to

this document. For a full bibliography, the reader is referred to the summaries given in the
five Annotated Bibliographies published in the USA [refs 7 to 12].
1. G. D. Breuer, E. M. Hunter, “The place of HVDC in the Power Transmission field”, IEEE
reprint of two DC Transmission sessions, Winter Power Meeting, New York, January
1963.
2. E. W. Kimbark, “Direct Current Transmission, Vol 1. Wiley Interscience, 1971.
3. A. Gavrilovic, “Some views on the choice between AC and DC alternatives”, IEE
Conference Publication No 107, 1973, p298.
4. C. T. Wu (editor) “AC-DC Economics and Alternatives – 1987 Panel session report”.
IEEE Paper 90WM 194-1 PWRS.
5. N. G. Hingorani, “Higher voltage DC Transmission ; A Power Electronics Workhorse”,
IEEE Spectrum April 96.
6. Oak Ridge National Laboratory Report for U.S. Department of Energy, “HVDC Power
Transmission Technology Assessment”, April, 1997. ORNL/Sub/95-SR893/1.
7. An annotated Bibliography of High Voltage Direct Current Transmission, 1969-1983,
Bonneville Power Administration.
8. An Annotated Bibliography of High Voltage Direct Current Transmission, 1984-1989,
Western Area Power Administration, Bonneville Power Administration.
9. An Annotated Bibliography of High Voltage Direct Current Transmission, 1989-1991,
Western Area Power Administration, Bonneville Power Administration.
10. An Annotated Bibliography of High Voltage Direct Current Transmission and Flexible AC
Transmission (FACTS) Devices, 1991-1993, Bonneville Power Administration, Western
Area Power Administration.
11. An Annotated Bibliography of HVDC Transmission and FACTS Devices, 1994-1995, US
Department of Energy, Western Area Power Administration.
12. An Annotated Bibliography of HVDC Transmission and FACTS Devices, 1996-1997, EPRI
and Bonneville Power Administration.
13. W. P. Lewis, “Issues and options for interconnection in Southern Africa”, Journal of
Energy in Southern Africa, published by SA ERI, Capetown, August 1994.
14. J. Arrillaga, “High voltage Direct Current Transmission” (2nd edition), IEE Power and
Energy series, 29.
15. T. Weaver (ed), “FACTS Overview”, IEEE Power engineering Society and CIGRE, April
1995, published by IEEE as 95 TP 108.
16. G.N. Alexandrov & G.V. Podporkyn, “Improvement of the efficiency of 35 to 220 kV
lines”. IEE Conference Publication No. 345, 1991, page 226.
17. Campos Barros J.G., et al, “Direct Connection of Generators to HVDC Converters: Main
Characteristics and Comparative Advantages,” ELECTRA, No 149, August 1993, pp.
18-39
18. Guide for preliminary design and specification of Hydro Stations with HVDC unit-
connected generators, JWG 11/14.09, CIGRE Brochure No 116, August 1997.
19. A. Clerici, G. Valtorta, L. Paris, “AC and/or DC substantial power upgrading of existing
OHTL corridors”, AC or DC Power transmission, IEE Conference Publication No 345,
1991, p 220
20. H. L. Thanawala, G. R. Moore, M. H. Baker, “Discussion on Embedding of DC
Transmission in AC Networks”, CIGRE symposium “Working Plant and Systems Harder”,
London June, 1999, paper 400-08.
21. K. Ramesh, T. Adhikari et al., “Parallel operation of AC and DC Lines Running on the
Same Tower”, CIGRÉ Report 14-08, 1988.
67
14.20newD
22. Guide for Upgrading Transmission Systems with HVDC Transmission, WG 14.11, CIGRE
Brochure No 127, August 1998.
23. W. F. Long, J. Reeve, “Considerations for Implementing Multi-terminal DC Systems,”
IEEE Transactions on Power Apparatus and Systems, Vol. PAS-104, No. 9, September
1985, pp. 2521-2530.
24. F. Mazzoldi, J-P. Taisne, C. J. B. Martin, B. A. Rowe, ”Adaptation of the Control
Equipment to permit 3-terminal operation of the HVDC link between Sardinia, Corsica
and Mainland Italy”, IEEE Summer Power Meeting, 1988.
25. D. McCallum, G. Moreau, J. Primeau, D. Soulier, M. Bahrman, B. Ekehov, “Multi-terminal
integration of the Nicolet Converter station into the Quebec-New England phase II HVDC
Transmission System”, CIGRE Paper 14-103, 1994, Paris,
26. J. P. Bowles, H. L. Nakra, A. B. Turner, “A Small Series Tap on an HVDC Line,” IEEE
Transactions on Power Apparatus and Systems, Vol. PAS-100, February 1981, pp.
857-862.
27. A. Gole, R. Menzies, “Analysis of a Small Series Tap on an HVDC System Using Forced
Commutation,” Thyristor and Variable Static Equipment for AC and DC Transmission,
London, 1981, IEE Conference Publication No. 205, pp. 137-140.
28. Guide to Planning DC Links Terminating at AC System Locations having Low Short
Circuit Capacities, WG 14.07
Part I - AC/DC Interaction Phenomena, CIGRE Brochure 68, June 1992
29. Part II - Planning Guidelines, CIGRE Brochure 115, 1997.
30. Oak Ridge National Laboratory for the US Department of Energy: “HVDC Power
Transmission Electrode Siting and Design”, ORNL/Sub/95-SR893/3.
31. D. Carlsson, S. Nyberg, J. O. Sjodin, L. Eriksson, “Deep hole electrode for HVDC
transmission”, Proc IEEE/Royal Inst of technology, Sweden, Power Electronics, June
1995.
32. General Guidelines for the design of Ground Electrodes for HVDC links, WG 14.21,
CIGRE Brochure, 1997.
33. Determination of power losses in High-Voltage Direct Current (HVDC) convertor stations.
IEC 61803-1999.
34. Protocol for Reporting the Operational Performance of HVDC Transmission Systems,
CIGRÉ Report 14-97 (WG 04) 21, 1989, Revised 1997.
35. A survey of the Reliability of HVDC Systems throughout the World during 1997-1998,
CIGRÉ Report 14-102, 2000, Paris.
36. D.J. Christofersen, V. Kahle, C.V. Thio, An Example of Reporting HVDC Thyristor Valve
Performance Using 1983-1994 Average Unavailability Data, CIGRÉ International
Colloquium on HVDC and FACTS, Johannesburg, South Africa, Sept., 1997, page 6.2.
37. Guide for Reliability of HVDC Converter Stations, IEEE P1240, draft, May 1999.
38. McNeill HVDC convertor station. R.P. Burgess, I.E. Barker, GEC ALSTHOM Technical
Review No 5, 1991, p3.
39. G. Lochner & P. Daehler, “HVDC back-to-back intertie with long transmission lines
feeding a relatively weak AC system”. Thyristor and Variable Static Equipment for AC
and DC Transmission, London, 1981, IEE Conference Publication No. 205, p 204 .
40. L. Carlsson, A. Persson, T. Jonsson, G. Liss, P. E. Bjoklund, “New concepts in HVDC
Converter Design”, CIGRE Paper 14-102, 1996, Paris.
41. G. Apslund, K. Erikson, H Jiang, J Lindberg, R Palsson, K Svensson, “DC Transmission
based on Voltage Source Converters”, CIGRE Paper 14-302, 1998, Paris.
42. Commutation failures, causes and consequences, W G 14.05. CIGRE Brochure No 103,
November 1995.
43. Coordination of Controls of Multiple FACTS/HVDC links in the same system, W.G. 14.29,
CIGRE Brochure No. 149, 1999.
68
14.20newD
44. M. Takasaki, T. Hayashi, “Subsynchronous oscillations analysis of power system with

power electronics applied facilities”, CIGRE Symposium paper 310-03, Tokyo, 1995.
45. Cross-modulation of Harmonics in HVDC schemes, W G 14.25, CIGRE Brochure 143,
June 1999.
46. J. Arrillaga, C. P. Arnold, B. J. Harker, “Computer Modeling of Electrical Power Systems”,
John Wiley and Sons.
47. P. Dey, G. Orawski, “Optimisation of Direct Current Overhead Lines”, IEE Conference
Publication No. 107, 1973, p 189.
48. R. D. Rosevear, “Cables for high voltage DC systems”, Power Technology International,
1990, p129.
49. J. B. Yates, R. Arnold, “Demand patterns make submarine cross channel link economic”,
Modern Power Systems, February 1985.
50. A. Rainey, M. Watson, C Harvey, “Public inquiries and consents process for the Moyle
Interconnector project”, 6th International IEE Conference Publication No 423, 1996, p39
51. H. Khatib, “Financial and economic evaluation of projects with special references to the
electrical power industry”, IEE Power Engineering Journal, February 1996.
52. H. Khatib, “Financial and economic evaluation of projects”, IEE Power engineering series
book, No 23, 1997.
53. H.B. Dreyfus, F.D. Jenkin and H.H.R. Taylor, “Planning considerations for additional
cross channel interconnection”, High Voltage DC and/or AC Power Transmission. IEE
Conference Publication No 107, Nov. 1973, p141.
54. P.H. Ashmole & M.G. Dwek, “Technical investigation for a cross channel interconnection
reinforcement”, High Voltage DC and/or AC Power Transmission. IEE Conference
Publication No 107, Nov. 1973, p145.
55. F. M. Wilkes, “Capitalised Budgeting Techniques”, a Wiley Interscience publication.
56. W. Bayer, K. Habor, D. Povh, D. A. Jacobson, J. M. G. Guedes, D. A. Marshall, “Long-
distance transmission with parallel AC/DC link from Cahora Bassa (Mozambique) to
South Africa and Zimbabwe”, CIGRE Paper 14-306, 1996, Paris.
57. J. D. Wheeler, J. L. Haddock, J. Loughran, “Chandrapur back-to-back HVDC link; first
steps towards an Indian national grid”, GEC ALSTHOM T&D Review 2/1995, p28.
69
Le CIGRÉ a apporté le plus grand soin à la réalisation de cette brochure thématique numérique afin de vous
fournir une information complète et fiable.
Cependant, le CIGRÉ ne pourra en aucun cas être tenu responsable des préjudices ou dommages de quelque
nature que ce soit pouvant résulter d’une mauvaise utilisation des informations contenues dans cette brochure.
Publié par le CIGRÉ

21, rue d’Artois
FR-75 008 PARIS
Tél. : +33 1 53 89 12 90
Fax : +33 1 53 89 12 99
Copyright © 2000
Tous droits de diffusion, de traduction et de reproduction réservés pour tous pays.
Toute reproduction, même partielle, par quelque procédé que ce soit, est interdite sans autorisation préalable.
Cette interdiction ne peut s’appliquer à l’utilisateur personne physique ayant acheté ce document pour
l’impression dudit document à des fins strictement personnelles.
Pour toute utilisation collective, prière de nous contacter à [email protected]
The greatest care has been taken by CIGRE to produce this digital technical brochure so as to provide you with
full and reliable information.
However, CIGRE could in any case be held responsible for any damage resulting from any misuse of the
information contained therein.
Published by CIGRE
21, rue d’Artois
FR-75 008 PARIS
Tel : +33 1 53 89 12 90
Fax : +33 1 53 89 12 99
Copyright © 2000
All rights of circulation, translation and reproduction reserved for all countries.
No part of this publication may be produced or transmitted, in any form or by any means, without prior
permission of the publisher. This measure will not apply in the case of printing off of this document by any
individual having purchased it for personal purposes.
For any collective use, please contact us at [email protected]

0186 Economic Assessment of HVDC Links

Uploaded by

Copyright:

Available Formats

0186 Economic Assessment of HVDC Links

Uploaded by

Document Information

Original Description:

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

0186 Economic Assessment of HVDC Links

Uploaded by

Copyright:

Available Formats

186

Members of the Working Group:

J.F. Allaire (Canada) W. Peter Lewis (UK)

2. GENERAL CONSIDERATIONS OF AC AND DC TRANSMISSION

3.1 Identifying alternatives for interconnections

4. TRANSMISSION SYSTEM COSTS

4.1 Check List of Cost Areas

5.1 Technical Benefits

6.1 Factors to be Considered

6.3 Economic Analysis Methodology

7. OTHER FACTORS AFFECTING A DECISION

APPENDIX I – HVDC CONFIGURATIONS

I 1. Monopolar HVDC Systems

APPENDIX II – EXPERIENCE REVIEW

II 1. Cross Channel link

APPENDIX III – MINI-SPEC

III 1. Scheme Type

Interconnection of high voltage transmission (grid) networks can enable electricity to be

Appendix I details alternative configurations of DC transmission, Appendix II some

2 GENERAL CONSIDERATIONS OF AC & DC TRANSMISSION

2.2 The Benefits of Interconnection

(c) Reduction in Reserve Capacity. A margin of reserve capacity has to be

(g) Environmental Benefits. Many of the above factors have consequential

(i) Technology Exchange. Shared systems tend to encourage standardisation of

2.3 Technical Considerations

When independent, asynchronous systems are to be interconnected, there are some

2.3.1 Long Distance Bulk Power Transmission

AC transmission becomes limited by:

2.3.2 Interconnection by AC or HVDC

If two or more independent systems are to be interconnected by a synchronous AC link,

For submarine interconnection AC cables generate an increasingly wide variation of voltage

2.3.3 HVDC Multi-terminal Systems

When power is to be transmitted from a remote generation location across different

3.1 Identifying Alternatives for Interconnections

• costs and benefits of each versus individual system expansion;

In order to generate points of comparison, a rigorous process of technical and economic

a) Power rating and overloading requirement. Once the objective of the

b) Equal demands on the performance parameters of all the alternatives should be

- Reactive power exchange by the interconnection at both ends. Identical

- Harmonic filtering requirements. It is well known that DC and power

In some cases, modifications may be necessary to existing communication

- Overvoltage considerations. The limits on overvoltage should not differ

e) Intermediate Loads. A long-distance transmission project, whether AC or DC, may

The assessment process establishes the interconnection as a viable alternative to other

In some instances some of the different interconnection options may be of different

3.2 Technical Assessment

A process of technical assessment is necessary, whether a link is AC or DC or a hybrid, to

3.2.1 Performance and Cost Analysis

The process in brief leading to a feasibility assessment follows:

1. Electrical Performance, AC and DC Systems

• Harmonic distortion is examined to establish filtering requirements under steady

• Subsynchronous resonance effects should be looked at to ensure that the DC link

5. Consider particular aspects of HVDC, as outlined below in parts 3.2.2 to 3.2.9.

3.2.2 HVDC Electrodes

The following, developed when considering HVDC specifications, may be useful.

• To select average generation cost per kW

A loss of kW at 0 MW flow, @ $/kW = $

An assessment of reliability influences the economic evaluation. Reliability calculations will