P
PRIOR TO THE MID-1960S, WHEN MERCURY ARC VALVES WERE USED AS THE MAIN
ac-dc/dc-ac conversion equipment, the increase of the dc transmission voltage level from ±400 kV (3/133kV groups in series) to ±450 kV (3/150-kV groups in series) did represent at that time a huge technological effort by the power equipment industry. Thyristor power semiconductor technology for ac-dc/dc-ac
conversion was first applied in the early 1970s, and many HVDC schemes did immediately apply the voltage level of ±500 kV for dc transmission, with little evidence of too great a technological challenge.
Then, by the late 1970s, the necessity to transmit 6,300 MW over 900 km led the Brazilian electricity sector managers to decide upon adopting the HVDC transmission alternative (for the
Paraguayan-side 50-Hz generators energy supply) in a geographically parallel transmission scheme
with 765-kV ac circuits (for the Brazilian-side 60-Hz generators). At that time, FURNAS, the Brazilian utility in charge of the transmission installation, had prepared a set of technical specifications for
the HVDC equipment, so as to enable manufacturers to bid the most economical alternative between
© ARTVILLE
march/april 2007
1540-7977/07/$25.00©2007 IEEE
IEEE power & energy magazine
61
Also, in terms of the total project
cost in $/kW, for HVDC projects with
Total Cost (Investments + Capitalized Losses)
transmission lines longer than 1,500
km, the transmission line would likely
represent the dominant cost compoDirect Investments = F( kV, Conductors
nent—above 50% of the total investSection and Number)
ment cost of the project—which brings
up the importance of designing the line
2
Cost of Losses = F (Power, and 1/kV )
with accurate optimization studies. In
fact, the predominance of the line cost
Voltage kV
on the total project cost will work in
favor of the 800-kV option.
figure 1. Selection of optimal voltage level: Direct investment plus losses.
The transmission line cost has two
main components:
✔ the investment in the equipment, including conducOptimal Voltage Level f(min)
tors, insulators, tower structures, engineering, environmental and construction costs, and other
hardware equipment
✔ the capitalized costs of power losses.
Length and
Both components will show values that will vary from
Power Rating
country to country. However, the cost optimization, which
Increase
will determine the transmission voltage level, for a specified power transmission rating would be to find an optimal
Voltage kV
point ( fmin ) or region on an economic analysis of the type
shown in Figure 1.
figure 2. Optimal voltage as length and power increases.
As the power rating (P) becomes higher, the optimal voltage (V) will also increase, mainly due to the benefit of reduc±500 kV to ±600 kV. The latter was chosen as being the ing power losses that shows a factor related to the square of
most economically attractive and adopted by FURNAS.
inverse voltage (P/V2 ), such as shown in Figure 2, which can
Since the 1980s no other HVDC scheme required a be obtained for a given power rating and line length of the
larger amount of power or a longer transmission distance project. Therefore, as an indicator, beyond 1,500 km and
to justify the employment of a voltage level above ±500 3,000 MW, the alternative of ±800 kV is probably the most
kV. However, the thyristor technology above ±500 kV was attractive from an overall economic analysis, considering the
made available about 25 years ago, or in other words, one total cost (investment plus capitalized loss components).
Considerations similar to those that led FURNAS to adopt
full technological cycle behind.
More recently, many requirements of employing bulk the transmission voltage of ±600 kV discussed above are
power transmission corridors are being seriously considered, now leading China and India to seriously examine economisome of them already at the edge of an official commitment, at cal voltage levels of ±800 kV.
the time this article is being prepared. Voltage levels above
±500 kV are again “on stage,” with the range under considera- Transmission Configurations
tion being either ±600 kV or ±800 kV. What has been com- for HVDC Lines
monly shared among experts in the field nowadays is that the What has been said before reflects the economic analysis
33% voltage level rise—from the available 600 kV to a new of a new project. However, another economic-derived
800 kV-would represent a smaller technological step than was aspect will show up in relation to the environmental pertaken 25 years ago in Brazil or 40 years ago with the transition mits and reports needed in preparation to build and operate
from mercury arc technology to solid-state thyristors. With the a new transmission line, which can be quite demanding in
need for 800 kV on the horizon, this article discusses impor- terms of time and resources; i.e., it indeed represents a cost
tant aspects to consider as well as technological challenges that component to be considered in the financial model of the
project. This aspect will act in favor of concentrating more
need to be overcome for successful implementation.
power in fewer projects. Or, in other words, as the effort to
Choosing 800-kV DC
be ready to build a new scheme is considerable, it is better
When power transfer ratings above 3,000 MW per bipole and dis- to maximize the amount of power that can be transmitted
tances beyond 1,500 km are being considered, the voltage level of over the new corridor. In terms of comparing 500-kV, 600±800 kV should be examined as a competitive alternative, as it kV and 800-kV transmission tower configurations, Figure
may be more economical than either ±600 kV or ±500 kV.
3 describes typical tower heights and horizontal distances
Cost $ or Unit Cost $/kW
Cost $
Optimal Voltage Level f(min)
62
IEEE power & energy magazine
march/april 2007
for each voltage level. This comparison shows for the three
voltage levels that differences in tower height (H) and horizontal spacing (D and R) are not so remarkable. This confirms the additional advantage to utilize higher voltage
levels for new transmission projects.
The Necessity of Power Evacuation:
Cares to Be Taken
If a fault occurs in the line, rendering one conductor
pole unavailable, HVDC schemes can temporarily continue to carry at least half of the power through the other
conductor pole, using the ground or another conductor
as the path for current return. Additionally, in many
projects, each of the pole conductors is designed with
temporary double power-carrying capacity, so as to
overcome this limitation.
If a fault occurs in the station (valves or transformers), the
resource available would be to use the overload capability of
the (remaining) converter station equipment, so as to reduce
the percentage of power not available. This feature, although
it might lead to additional costs, is an important part of the
engineering studies of a project.
The reliability studies to be carried out during the system
planning phase should recognize the likelihood (frequency
and duration) of possible faults in the HVDC link, in order to
improve the level of the information on how much power
shall be considered as being unavailable. Several factors need
to be addressed:
✔ line faults: monopolar or bipolar, permanent or transient nature
✔ station faults: critical equipment, switching possibilities, spare units
✔ load factor of the HVDC link: this is of special impor-
tance when the link is connected to a remote generation station, and the generation has a nonuniform, or
seasonal, load shape.
As counterparts of these factors, the project owner should
consider:
✔ contingency power capacity of each line pole
✔ overload capability of the converter stations.
All these factors should be analyzed and carefully balanced against the function of the HVDC project, whether it is
going to be a transmission or an interconnection project. Normally, transmission projects are designed to supply power to
a load area. Interconnection projects deal with energy transfers between areas and therefore exhibit different roles and
may be designed with specific performance targets.
Another aspect of importance when dealing with new
HVDC schemes, and especially if they employ 800 kV, is the
possibility of having many HVDC lines feeding the same electrical area—the HVDC multi-infeed issue. Research on this
aspect started in the early 1990s by the Electric Power
Research Institute (EPRI)/University of Wisconsin-Madison
and Centro de Pesquisas de Energia Electrica (CEPEL). More
recently, CIGRÉ Working Group B4.41 has addressed this
issue, establishing performance indicators to predict situations
more prone to multiple commutation failures. A technical
report from this working group is expected to be made available by the end of 2007.
HVDC Performance Statistics
and Design Targets
HVDC statistics of failures have been continuously and consistently analyzed and treated by one CIGRÉ Advisory Group
D
+
−
H
kV
500
600
800
D (m)
13
15
20
H (m)
47
51
58
R
RoW (m)
65
80
100
R
figure 3. Transmission electrical distances for 500-kV, 600-kV, and 800-kV dc.
march/april 2007
IEEE power & energy magazine
63
Beyond 1,500 km and 3,000 MW,
the alternative of ±800 kV is probably the most attractive
from an overall economic analysis, considering the total cost.
(AG04—HVDC System Performance) since the 1970s. One of
the main parameters of these statistics is the resulting HVDC
forced energy unavailability (percentage in time where the energy is not available). As those statistics gather information from
almost all operating schemes, it is not an easy task to summarize the operational indices based on average values, since some
schemes show particularities that may lead to high unavailability times, due, for instance, to lack of spare equipment; at the
same time, many dc schemes do not show any failures.
Some recent results issued by this working body, considering a representative number of schemes and time observation period, can be summarized in Table 1. As the issue of
large converter transformer failures has been a critical and
negative operational factor, the overall statistics are greatly
affected if transformer failures are or not considered. Table 1,
for comparison purposes, also includes reliability design
indices specified in most recent schemes.
For transmission line reliability, one interesting experience
comes from the FURNAS Itaipu HVDC lines. Twelve-year
(1988–2000) statistics reported a total of 54 events (either
monopolar or bipolar) with a total duration time of 25 hours,
or, around two outages/year and 1.9 hours/year as average.
One should have in mind that these lines crossed a region of
high values of extreme wind speed, which has been responsible for several tower breakdowns over the years.
Some recent HVDC schemes recently implemented in
southern China specified a maximum of 6 and 0.1 for
monopolar and bipolar outages/year, respectively. The first
years of operation of those schemes have led to operational
performance indices that are compatible with those specified.
CIGRÉ AG04 statistics show that average duration time of
station outages is in the order of 10 hours/year.
The numbers and arguments presented here, in a summarized form and not in great detail, provide arguments for a
technical discussion on how system reliability constraints
can influence the configuration of the HVDC scheme, and
especially those new schemes of 800 kV. In summary, one
can say that:
✔ the number of forced bipolar line outages is very small;
usually these faults are associated with tower breakdown (due to extreme winds for instance) and therefore
most line faults happen at one pole (due to insulation or
switching misoperation, for instance); this is why the
issue of operation with one pole with ground return (via
electrodes) becomes so important in HVDC schemes
✔ also, for converter group faults (within stations), a full loss
of the whole station is a much rarer event than losing one
pole (in the case of one converter per pole) or a portion of
a pole (in the case of two series converters per pole)
✔ the statistics presented above would have to be used
with different criteria, depending on if the link will have
the role of either power supply (MW-based) or energy
interconnector (MWh-based), with the latter being able
to be designed under more relaxed requirements
✔ however, even knowing that the number of expected
average hours of unavailability is reduced, the possibility of losing a power block of 3,000 to 6,000 MW,
even being a rare event or typically of short duration, is
a subject of major concern to the power grid; perhaps,
rather than doubling equipment as a safeguard, system
measures to alleviate the impact upon the grid, such as
load transfer, special switching in the network as well
as in the dc station, generation re-dispatching, etc.,
could be analyzed
✔ also, to allow monopolar operation, the HVDC system
designer should carefully dimension and plan the
ground electrodes, as this element plays an important
role when only one conductor pole is available. There
is general consensus within the
HVDC community that a better
table 1. Summary of operational and design targets
for availability of HVDC schemes.
knowledge of the land path
(crossed by the line) geology and
Overall Energy Availability of HVDC schemes (10 year
(a) 98.5%
geomorphology is vital to guaranrecords) considering: (a) overall performance;
(b) 99.5%
(b) excluding transformer failures
tee that operation with current
Forced Energy Unavailability—Specification Targets
0.5% (or 99.5%
return through ground would not
of availability)
cause any adverse interference with
Scheduled Energy Unavailability—Specification Targets
1% (or 98.5% of availability
telecommunication or pipelines, or
considering forced and
cause transformer saturation in
scheduled)
nearby transformers
64
IEEE power & energy magazine
march/april 2007
The fact that there are some clear technological developments
to be accomplished should not preclude the go
ahead of this alternative.
✔ if a fault occurs in the converter station, some other
strategies may apply, mainly those derived from converter switching (paralleling) or making use of spare
units—especially those related to converter transformers, which have been behaving as the critical equipment in many HVDC schemes, especially those with
higher power ratings.
Electrical Design Considerations
of Transmission Lines
Operating experience with HVDC transmission lines in the
voltage range of ±400 kV to ±600 kV around the world,
along with the results from research studies carried out at
voltages between ±600 kV and ±1,200 kV, has established
the technical feasibility of transmission lines at ±800 kV.
Electrical design of ±800 kV transmission lines requires consideration of the following aspects:
✔ corona
✔ air insulation
✔ insulators.
The vast amount of knowledge and experience gained in
designing and operating ac transmission lines at voltages up
to 800 kV, and some even at 1,100 kV, cannot be applied
directly to the case of dc lines because of basic differences in
the electric field and space charge environments in the vicinity
of ac and dc transmission lines. The alternating nature of
electric fields produced by ac lines leads to the space charges
created by corona being confined within a narrow region
around the conductors. In contrast, the steady nature of electric fields produced by dc lines makes the space charge generated by corona fill the entire space between the conductors
and ground. These differences in the electric field and space
charge environment have a large influence on the corona and
insulator performance, and to a lesser extent on the air insulation performance, of dc transmission lines.
The current state of knowledge of the corona, air insulation, and insulator performance of dc transmission lines is
reviewed below and any additional research studies required
for the design and operation of future ±800 kV transmission
lines are highlighted.
Corona Performance
Corona performance of both ac and dc transmission lines is
generally defined in terms of corona losses (CLs), radio
interference (RI), and audible noise (AN). However, for dc
march/april 2007
lines, the corona-generated space charge environment,
defined in terms of the ground-level electric fields and ion
currents, is also an important design consideration. One of
the earliest comprehensive investigations on dc corona was
carried out in Sweden. The studies were made on a test line
at voltages up to ±600 kV. The measurements included CL
and RI for different conductor configurations, mainly under
fair weather conditions, since it was observed that, unlike
in the case of ac corona, RI levels for dc corona were lower
under conditions of rain than in fair weather (although the
corona losses in rain are higher, as in the ac case, compared to those in fair weather).
Research and development necessary for the design and
operation of the Celilo-Sylmar ±400-kV transmission line
were carried out by the Bonneville Power Administration
(BPA) at the HVDC test center at The Dalles, Oregon. Following completion of the ±400-kV transmission line, the test
center was modified as the EPRI-HVDC Project for studies
in the range ±400 kV to ±600 kV. The studies included
long-term measurements and statistical analysis of CL, RI,
and AN from a number of different conductor configurations
and, for the first time, also of ground-level electric fields and
ion currents. The results of the 4-year project, specifically
directed at investigation of the phenomena, characteristics,
and requirements of dc transmission lines in the ±400 kV to
±600 kV range, were summarized in a reference book by
EPRI called Transmission Line Reference Book HVDC To
±600 kV, popularly known as the Green Book.
Following publication of the Green Book, EPRI sponsored a large project at the Institute de Recherche d’Hydro
Quebec (IREQ) to study the technical feasibility of HVDC
transmission in the range of ±600 kV to ±1,200 kV. A large
amount of statistical data was obtained on the corona performance of three bundles consisting of four, six, and eight
subconductors, selected for operation at nominal voltages of
±750 kV, ±900 kV, and ±1,050 kV, respectively. Based on
the results obtained in different seasons and weather conditions, some empirical formulas were derived for CL, RI,
and AN. A comprehensive study of the ground-level electric
field and ion environment was also carried out. Finally, controlled psychophysical studies were carried out to determine
design criteria for RI and AN from HVDC lines.
A study, sponsored jointly by EPRI and the U.S. Department of Energy (DOE), was carried out at the research center
at Lenox, Massachusetts, initially conceived to cover the
IEEE power & energy magazine
65
voltage range of ±600 kV to ±1,200 kV, but subsequently modified to cover the voltage range of only ±400 kV to ±600 kV.
Special techniques were developed in this study to measure the
space charge densities in the vicinity of the line and particularly
the space charge due to large ions (charged aerosols) carried
downwind from the line. The results of this study were documented in EPRI’s HVDC Transmission Line Reference Book and
a tentative methodology was proposed for corona design of
transmission lines in the range of ±400 kV to ±600 kV.
In the EPRI studies at IREQ, a full-scale tower structure
corresponding to a bipolar dc line was simulated in the highvoltage laboratory and tests were carried out for different
conductor-tower clearances, with mixed (i.e., switching
impulse superimposed on dc) positive voltages applied to one
pole with the other pole either grounded or with a negative
bias voltage applied to it. In the EPRI study at Lenox, the
influence of preexisting space charge on the flashover characteristics under the application of mixed voltages was studied.
Air Insulation Performance
Insulator Performance
Data on flashover and withstand characteristics of conductortower and conductor-ground air gaps, under the application of
lightning and switching overvoltages, are required for determining the minimum clearances necessary on HVDC transmission lines. A good proportion of the results obtained from
studies carried out for ac transmission lines can be used for
the design of dc lines. In the EPRI-BPA study, a flashover
tower and a paved outdoor area were constructed at the
Dalles test facility to carry out studies of various air gaps
associated with dc transmission lines and substations.
As in the case of ac lines, flashover at normal operating voltage due to contamination and/or wet weather conditions has
been considered the most critical factor influencing the selection of insulator strings on HVDC transmission lines. Some
of the early studies to understand the dc withstand characteristics of polluted insulators were carried out in Sweden.
In the EPRI-BPA study, artificial contamination tests
were carried out on a large variety of line insulator strings
in a specially built fog chamber. The main objective of the
EPRI-IREQ study was to develop specifications of the
UHV Gr id
ºôÃËúµç
ÃË
úµçI
ÎÚ³ľ
¾Æë
Ö÷
Íø
Nort
h e as
Northeast
ÎýÃË
Îý
úµçII
t
ÎýÃË
¹¹þÃÜ
µç³§
¡Á
Nort
North h
i n a
China
¹þÃÜ
¹þ
¶þ¡¢Èý³§
°²Î÷
°²
ÃÉ
Î÷úµçIV
ÃÉ
Î÷úµçII I
Nort
h w e
Northwest
°×Òø
°×
ÓÀ
µÇ
260
0 ¾±̧ß
Äþ¶« 26
Î÷Äþ
ƽ
Á¹
ǬÏØ
À-Î÷Íß
¹Ùͤͤ
100
10
0
ÉÂ
±±
ɱ±Ãºµç 220
Òǿ¨ ¶«
s t
400
400
μ±±
À¼
ÖÝ
¶«
¶«
150
ÀÖ
ɽ
ÑÅ
½-Ë®
µç
´¨ Î÷
Î÷Ë®
µç
300
ÖØ
Çì 360
360
ͼÀý
ÐìÖÝ
ÖÝ
160
160
270
270
ÑØ
º£ºËµç
Á¬ÔÆ̧
Û
270
500kV½»
Á÷
500
Kv
°ÙÍò ·ü ¼¶½»Á÷
³üÖÝ
©
170 Ì©ÖÝ
170
130
70
170 ÎÞÎý150ÉϺ£Î÷±±
170
ÄϾ©
¾©
26
260
0
170 ÉϺ£Î÷ÄÏ
100
440
ÑØ
º£ºËµç
º¼±±
Î人 440
Îߺþ 290 24
240
0
»´ÄÏúµç
»´
220
220
800
KvÖ±Á÷
¡À800kV¼¶
750
Kv
750kV½»
Á÷
½ð
Ȼ
280
³¤É³300
380
380
400
400
ÄÏ
Äϲý
360
360
Ea
½ð
ɳ½-II ÆÚ
Ú
¸ÓÖÝ
Ch
South
So
China
280
280
300
30
0
ÐìÖÝ
Ã
µçºµç
450
300
300
4500
½ð
ɳ½½-IÆÚ
Çൺ
36
360
0
140
140
i n ¾£
aÖ Ý
300
0
¶÷Ê© 30
ÑØ
º£ºËµç
250
340
¼Ã
ÄÏ
Ô¥±±
פÂíµê
300
ÑØ
º£µçÔ´
300
300
240
½ú
ÖÐ
úµç
ÈýÏ¿µØ
ϵçÕ¾ 270
270
ÑÅ
Áú½-Ìݼ¶
¼¶
ÌÆɽ
120
Ìì ½ò
280
280
ʯ¼Ò
ׯ
ׯ
400
ÄÏÑô
ÑÅ
°²Central China
Ce
ntr
a l Ch
´¨ Î÷Ë®
µç
300
300
Î÷°²¶« 330
¡Á
470
Tibet
42
420
0
½ú
ÖÐ
300
460
46
0
½ú
¶«ÄÏúµç 100
½ú
¶«ÄÏ
½ú
¶«ÄÏúµçII
300
¶«ÄϽ¼
Ti
ÉòÑô
480
150
±±¾©
¶« 150
450
450
ÃÉÎ÷
ÃÉ
Î÷úµçII
ÃÉÎ÷úµçI
µçV
ÃÉ
Î÷úµç
ÄþÏÄúµç
ÕÅ
Ò´
480
480
Ch
200
200
280
28
0
s t ¸£ÖÝ
170
17
East
China
410
ÑØ
º£ºËµç
ÎÂ
ÑØ
º£ºËµç
i n a
Ȫ
ȪÖÝ
Taibei
figure 4. Plans for a ±800-kV (red lines) future UHV grid in China.
66
IEEE power & energy magazine
march/april 2007
Another important aspect with new HVDC schemes, and especially
if they employ 800 kV, is to clearly define whether the link will have
the role of power supply or energy interconnector.
HVDC power supply to determine the 50% flashover voltage of heavily contaminated insulators with an accuracy of
5%. Following completion of the EPRI project, additional
studies were carried out at IREQ on dc source requirements for pollution tests. In the EPRI study at Lenox,
extensive studies were carried out on the pollution performance of different types of line and station insulators
and on the performance of wall bushings.
Design Considerations
The design of a new HVDC transmission line for satisfactory corona performance requires two sets of information:
first, analytical or empirical methods for predicting the
corona performance of the proposed line configuration, and,
second, design criteria for acceptable corona performance.
However, the information presently available from either
research studies on experimental lines or measurement surveys on operating dc transmission lines may not be adequate for deriving accurate empirical formulas for
predicting the corona performance, particularly for ±800
kV. However, the information on acceptable levels of RI,
AN, and ground-level electric fields and ion currents may
be sufficient to establish tentative design guidelines.
Information currently available on the flashover and
withstand characteristics of air gaps is adequate for the
design of ±800-kV transmission systems. The selection of
insulator strings for reliable operation under conditions of
pollution requires knowledge of the nature and severity of
pollution along the route traversed by the transmission line.
This information can be obtained by conducting pollution
surveys along the line route. The number and type of insulators in the string may then be obtained either using available data on the required specific leakage distances or
through actual tests in a pollution chamber.
Need for Additional Research Studies
The review presented above underlines the importance of
research and development studies before undertaking the
design, construction, and operation of any large-scale HVDC
transmission system. Research studies would also permit
optimization of the new systems taking due account of any
local environmental constraints, operation and maintenance
requirements, and new developments in material and equipment technologies.
Currently available experimental data may not be adequate
for predicting the corona performance, particularly the
march/april 2007
ground-level electric fields and ion currents, of new ±800-kV
transmission lines. There is need for studies on experimental
lines with different conductor bundles as well as for longterm measurements on existing HVDC transmission lines in
order to develop accurate methods for predicting the corona
performance of future HVDC transmission lines.
Although currently available information may be adequate for the air insulation design of ±800 kV transmission lines, some questions remain unanswered, such as
the influence of electrodes and of any preexisting space
charges on the flashover characteristics of line and station
air gaps. Studies are also required for developing safe
operating and maintenance practices for ±800 kV transmission lines.
Further research is needed for optimal selection of insulators and ensuring reliable operation of ±800 kV lines. Studies are also required to evaluate the effect of nonsoluble
materials on the flashover characteristics of insulators and the
behavior of nonceramic insulators stressed with direct voltages and their aging characteristics.
Research Topics Relevant to
800-kV HVDC Systems
The technical challenges that should be faced when designing
new ±800-kV HVDC systems are summarized below.
Converter Transformers
This equipment constitutes a key parameter for any new
HVDC project, and especially when dealing with ±800 kV
due to the large amount of power that can be interrupted due
to a transformer outage. CIGRÉ AG 04 statistics reveal that
many failures have been occurring, albeit concentrated in a
few HVDC schemes (i.e., apparently some projects have
shown to be less robust regarding transformers). Also, these
statistics show that causes of failures arise from different
sources, such as mechanical, insulation, oil contamination,
etc. There is no direct relation of failures with voltage levels,
but they tend to be concentrated in large power projects.
Higher Power Handling Capability of Thyristors
The existing thyristor technology, referred to as 5 in. diameter,
may need to be expanded to a larger width (6 in.) to accommodate the current level required by some of the 800-kV projects under consideration; such development, although not
negligible in terms of technological effort, should not be considered as a critical aspect.
IEEE power & energy magazine
67
The adoption of ±800 kV as the nominal voltage for new HVDC
applications brings along some technological challenges that should
be examined by the electrical power industry and academia.
External Insulation
An investigation of technical problems associated with
HVDC converter stations at voltages above 600 kV has
revealed that one of the critical considerations in implementing ±800 kV systems is the design and development of outdoor dc wall bushings.
Operating experience has shown that for systems above
±400 kV, the problems with the performance of dc wall
bushings become increasingly more serious. Increasing the
specific leakage distance was found to be not sufficient to
improve the performance of bushings. Laboratory investigations have shown that nonuniform wetting of bushings under
rain is a more critical condition than pollution for the
observed flashovers.
Extensive investigations have, therefore, been carried out
to understand the mechanism of bushing flashover in
nonuniform rain. Results of these investigations clearly show
that increasing the specific leakage path while keeping the
same insulator length does not improve the performance of
wall bushings under nonuniform rain. The use of hydrophobic coatings and of booster sheds has been shown to be
effective in improving the performance of wall bushings.
The use of silicone rubber or other similar materials rather
than porcelain is also being seriously considered. However,
there is still a need for research studies to evaluate the effectiveness of new methods and materials in improving the performance of dc wall bushings for operation at ±800 kV.
Electrodes
As already mentioned above, the electrode design for ±800 kV
systems would play a relevant role in the configuration
design and selection. The electrode design should allow
ground return with minimal (preferably no) adverse factors
since this mode of operation is critical for reliability.
Configurations
Another interesting design aspect is related to the configuration of the HVDC scheme. Due to the size of power to be
handled, the series or parallel arrangements, one or two
valve groups per pole, should be compared with great care;
also, the possible staging (expansion from the base configuration) of the project might be a key factor in determining
the best configuration.
Availability of Test Facilities
This is another interesting aspect that seems to be fulfilled as
the 800-kV market appears to be so promising (see next section). Leading manufacturers have formed internal project
teams or cooperative efforts to rapidly develop testing facilities necestable 2. Possible 800-kV DC Projects in China.
sary to test new project orders.
Project
Yunnan-Guangdong, Yunnan-Eastern
China, Yunnan-Central China
Jinshajiang-Shanghai
Jinshajiang-Jinhua in Zhejiang
Jinshajiang-Quanzhou in Guang
DonGuangdong Guang Dong
Yalongjiang-Chongqing
Yalongjiang-Suzhou in Jiangsu
Humeng in Inner Mongolia-Liaoning
in Shenyang
Humeng in Inner Mongolia-Beijing
Eastern Ningxia-Nanking in Jiangsu
Hami in Xinjiang-Zhengzhou
in Henan
Tibet-Guangdong, Tibet-Central
China, Tibet-Eastern China
Kazakhstan-China
Far East Hydroelectricity in
Russia-Shenyang in Liaoning
68
IEEE power & energy magazine
Transmission
Distance
Capacity
Commissioning
>1,500 km
24.8 GW
2010–2020
About 2,000 km
About 2,000 km
>2,000 km
2010–2020
2010–2020
2010–2020
2010–2025
2010–2025
2015–2020
>2,000 km
9.8 GW
About 2,400 km
12.6 GW
10.8 GW
2015–2020
2015–2020
2015–2020
>2,000 km
35 GW
2015–2025
>2,000 km
2015–2025
2015–2025
New 800-kV HVDC
Market Perspectives
Since 2004, there has been an
increasing interest in HVDC
applications involving voltage
levels above 500 kV. In 2005 and
in early 2006, China and India
have announced their decision to
go for ±800 kV. Brazil is also
considering it, but the configurations envisaged might lead to
lower than 800-kV solutions.
Additionally, applications in the
Southern Cone of Africa have
created an interest in considering
voltage levels higher than 500-kV
dc in projects to explore the Inga
River potential.
march/april 2007
Studies are needed on experimental lines with different
conductor bundles and for long-term measurements
on existing HVDC transmission lines.
Table 2 summarizes the potential applications for new
800-kV dc schemes in China for the next 20 years. Figure 4
describes the location of these schemes within the Chinese
grid. Also, in India, the Northeast-Agra 800-kV 6,000-MW
project is currently very close to becoming reality.
Conclusions
The adoption of ±800 kV as the nominal voltage for new
HVDC applications brings along some technological challenges that should be examined by the electrical power
industry and academia with great care. However, to transmit
large blocks of power (over 3,000 MW/bipole and beyond
1,500 km of line length), the option of ±800 kV dc is likely
to be the most cost effective alternative. The fact that there
are some clear technological developments to be accomplished should not preclude the go ahead of this alternative.
They are feasible to be dealt with in due time to match market needs and do represent lower gaps than those in past
decades. These are good challenges. HVDC technology will
improve compliance with these market needs.
For Further Reading
C.A.O. Peixoto, “Itaipu 6300 MW HVDC transmission system feasibility and planning aspects,” presented at the U.S.
DOE Conf., Phoenix, AZ, USA, Mar. 1980.
I. Vancers, D.J. Christofersen, A. Leikbukt, and M.G. Bennett, “A survey of the reliability of HVDC systems throughout the world during 2003–2004,” in Proc. CIGRÉ 2006,
Paris, France, Paper B4-202.
N. Knudsen and F. Iliceto, “Contribution to the electrical
design of HVDC overhead lines,” IEEE Trans. Power Apparatus Systems, vol. PAS-93, no.1, pp. 233–239, 1974.
EPRI, Transmission Line Reference Book HVDC To
±600 kV. Palo Alto, CA: EPRI, 1976.
EPRI, HVDC Transmission Line Reference Book. EPRI
Report TR-102764, 1993.
P.S. Maruvada, Corona Performance of High-Voltage
Transmission Lines. Ltd., Baldock, UK: Research Studies
Press, 2000.
Biographies
Marcio Szechtman received his B.Sc. and M.Sc. from the
University of Sao Paulo in 1971 and 1976. He joined
CEPEL, the Brazilian Power System Research Center, in
march/april 2007
1976 until 1996, when he concentrated his career in the
field of HVDC system studies and modeling. He participated in the Itaipu system studies and more recently in the
2,000-MW Garabi back-to-back Argentina-Brazil electrical interconnection. Since 1997 he has been acting as a
consultant for technical, economic, and regulatory studies,
concerning ac and dc transmission and distribution projects. He is an IEEE/PES Fellow and has a long history
with CIGRÉ Study Committee B4 (formerly SC 14)
(HVDC and Power Electronic Equipment), serving as
chairman since 2002.
P. Sarma Maruvada obtained the B.E. in electrical
engineering from Andhra University and M.E. in high-voltage engineering from the Indian Institute of Science. He
received the M.A.Sc. and Ph.D. in electrical engineering
from the University of Toronto. He worked as a researcher
and research manager at the Hydro Quebec Institute of
Research (IREQ) on electrical performance of transmission
lines for 29 years and is currently working as a consultant.
Dr. Maruvada made significant contributions to the understanding of corona performance and the development of
design criteria of ac and dc transmission lines. He is a Fellow of IEEE (1983) and an Honorary Member of CIGRÉ
(1998) and has published over 60 papers on the corona performance and the electromagnetic environment of ac and
dc transmission systems and a book on Corona Performance of High Voltage Transmission Lines. He was chairman of CIGRÉ Study Committee 36 on Power System
Electromagnetic Compatibility (1990–1998). Dr. Maruvada
received the 2003 IEEE Herman Halperin Electric Transmission and Distribution Award.
R.N. Nayak received his B.Sc (electrical engineering)
from REC Rourkela and M.Tech. from IIT Kharagpur, India.
He is currently working as executive director of engineering,
quality assurance and inspection for Power Grid Corporation
of India Ltd. and is associated with concept to commissioning of transmission projects including EHVAC and HVDC.
This includes system planning, design, engineering, procurement, quality management, and implementation. His experience also covers design, engineering, and implementation of
state-of-the-art load dispatch, communication systems, and
telecommunication networks. He is convener of the CIGRÉ
Study Committee B4 Working Group on Technological
p&e
Assessment of 800 kV HVDC Applications.
IEEE power & energy magazine
69