IEEE HVDC Guide Outline-July22-2012

Proposed Outline of HVDC Overhead Line Design Guide
July 2012

HVDCTransmission Systems
Intention of Guide (For AC designers already know how to design lines)
Introduction
HVDC Line Performance
DC versus AC ratings
Reliability
Single Pole (Monopole) operation
Overhead line versus underground cable
System performance Wayne Galli

System Design:

Insulation Coordination (tower clearance, clearance to ground) Next presentation (backup)
- Switching surges
- Contamination
- Lightning
- Tower clearances
- Clearances to ground

Lightning and Grounding
- Acceptable performance
- Calculation Methods

Electrical Effects Next Chapter (Winter meeting presentation)
- Corona & Field Effects
- Calculation Methods (AN, RI)
- Inductions
- Rules and Regulations (Maximum values and limit at edge of ROW)
- Co-use of Towers/ROW

Conductor System - Next Chapter (Winter meeting presentation)
- Current Rating Determination
o Selection of weather conditions for ratings
- Hardware Testing
o Selection of criteria
o Corona Performance
o Corrosion Performance
o Current and non-current carrying components
- Conductor Selection
o Optimization using Life Cycle Costing evaluation

Mechanical Performance
- Weather loading,
- Aeolian vibration
- Conductor galloping

Component Design:

Insulators
- DC insulators versus AC Insulators (DC insulator characteristics)
- Types of Material
- Comparison advantages & disadvantages
- Determination of leakage distance (charts of contamination levels)
- Testing and Evaluation
- Selection of insulators
- Reference standards

Spacer Dampers
- Types
- Spacing and damping capability

Other Hardware (Falls under conductor systems)
- Connectors
- Suspension clamps
- Guys

Conductor (Phase & Ground Wire)
- Types
- Selection of conductor

Structure
- Types
- Selection of type and configuration

Foundation
- Types
- Selection of foundation

Electrode
- Types of electrodes
- Comparison suitable type of electrodes
- Location
- Components
- Performance criteria

Electrode line
- Voltage requirement
- Current requirement
- Line configuration
- Reliability
- Metallic ground return
-

Introduction
Transmission of electric power by high voltage direct current(HVDC) was launched in 1954 when the first
commercial transmission link between the Gotland and the mainland of Sweden was put into
service.Since then, HVDC technology has advanced dramatically, and close to 100 HVDC transmission
systems have been installed around the world. HVDC is better known for its capability of transferring
large amount of power between two points and its potential for other applications has been overlooked
until recent years.
Today the most common application of HVDC is bulk power transfer over a long distance. HVDC
transmission has the advantage of lower line losses and lower transmission line costs. These savings are
off-set by the costs of conversion stations and the conversion losses. For a 500 kV HVDC transmission
system carrying 3 GW of power to be competitive with the AC option, the distance had to be longer than
500 miles in the 1970s. Due to relatively lower converter costs and higher line loss costs comparing to
the other costs, this breakeven point is now around 300 miles. The highest voltage for DC transmission
was at 600 kV until recently when an ultra-high-voltage direct-current (UHVDC) 800 kV transmission
system was built in China. In countries such as China, India, Brazil and South Africa where large amounts
of power are regularly transferred over long distances, HVDC could become the backbone of their
transmission systems.
The second most common application of HVDC has been back-to-back asynchronous interconnection.
Two power systems operating at different frequencies are tied together using two HVDC converter
stations and a short or no overhead line between them. This scheme can also be adopted to isolate
systems within the power grid, when required, from affecting one another and thus provide stability to
the grid. There are several of these types of back-to-back interconnections in US, not just for
asynchronous frequencies, but also for reasons of reliability.
With the advent of Voltage Source Converter (VSC) based technology, opportunities for HVDC
application have opened up. Conventional HVDC converters make use of phase commutated converter
(PCC) based technology. The advancement of solid state devices has increased the power capability of
ratings but not controllability. VSC technology uses gate turn off devices, like IGBTs (Insulated Gate
Bipolar Transistors), which have higher switching frequencies, therefore allowing for better voltage
control. This technology is used for the following applications.
Urban in-feeds
Constrained right-of-ways
Improved voltage stability
Underground and sea cable transmission
Wind farm integration

Some recent examples are:
2002: Cross Sound Cable HVDC Light for controlling power exchange
2007: Sharyland Link between ERCOT and Mexico for asynchronous networks
2009: NordE.ON1 linking an offshore windfarm to the German mainland
There is an increasing amount of wind farms that need to be connected to the AC transmission network.
Integration of wind farms using HVDC Voltage Source Converters (VSCs) have emerged as the
technology of choice.Power can be fully controlled using VSC based technology so that the intermittent
electricity supply from a wind farm cannot disrupt the grid. At the same time, the wind farm is isolated
from the ac network by the dc link. The voltage in the wind farm is therefore not affected by changes of
the voltage in the ac network that may be caused by switching actions or remote faults.
Other than bulk power transfer, HVDC technology can be applied to improve power system reliability, to
connect an AC system to renewable power sources, to control and improve power flows. The maximum
capacity of a VSC is currently at about 1,000 MW. The improvement of the capability of VSC will open
up even more opportunities for the application of HVDC technology.

HVDC Line Performance
The conventional wisdom concerning HVDC line design is that it is a simple extension of AC
design. In a physical sense, HVDC lines are similar to AC transmission lines in that both use
bare stranded conductors and both must meet or exceed minimum specified electrical clearance
to ground and to other conductors at operating voltage.
A direct comparison of performance between HVDC and AC should consider the level of
maintenance required to maintain acceptable outage rates. HVDC lines are made of essentially
the same materials as AC lines of a comparable voltage level. Normal weathering of
components of foundations, structures, conductors, splices, clamps, insulators, and shield wires
should be similar in a 400 kV HVDC line and a 345500 kV HVAC line.
However there are several differences between AC and DC lines, generally DC insulators require
greater leakage lengths, DC lines have higher power flow capability and controllability, DC also
allows for the flexible operation of lines with various ground return arrangements. The cost of
construction and the power flow on HVDC circuits is, however, much higher than with AC
circuits so that design mistakes are amplified. Figure 1 shows an AC and DC line running
parallel to each other.

Figure 1: DC line on the left, AC line on the right

DC line outage rates also tend to be lower than that of AC lines in general. In specific instances
where the DC line outage rate was initially higher than a comparative AC line due to component
or equipment issues, once the issue was corrected the DC line outage rates dropped to below the
AC line outage rates.
Some of the differences of an HVDC circuit compared to an HVAC circuit are:
- Failure of an HVDC line can be quite different from that of an HVAC line. Flashover
of any of the three-phase conductors of a single-circuit AC line causes breaker operation
and takes the line completely out of service. A bi-pole HVDC line is somewhat like a
double-circuit AC line in that the poles of the HVDC line can be operated independently,
just like the AC circuits. The failure of one pole or circuit allows continued operation
although at a reduced power level.
- A bi-pole HVDC line needs only two conductors, instead of the three required by a
single-circuit AC line or the six conductors in a double-circuit AC line. This normally
requires a smaller right of way and a less visually obtrusive tower. Figure 2 shows
schematically the tower configurations for 1200 MW (two circuits AC, bipolar HVDC)
and 15002000 MW transmission at EHV AC single circuit or monopolar HVDC by
alternative tower designs.

Figure 2: Comparison of physical size for HVAC and HVDC circuits with comparable power flow
capacity [1]
- The conductors in an HVDC line do not experience skin effect, so electrical losses are
lower for a given power flow.
- DC lines offer power flow control, whereas HVAC lines are electrically passive.
Depending upon the placement of the HVDC circuit within the AC transmission system,
this advantage can be critical to system reliability during system emergencies.
Monopole
HVDC
Single AC
circuit
Bipole
HVDC
Two AC
circuits
- HVDC cables have no length limit as do AC cables, so HVDC cables can be used for
long-distance transmission of power, either underground or undersea (submarine).
- There is no need for frequency synchronization between systems linked by DC, so
HVDC circuits can be used to link systems that are not necessarily in frequency
synchronization. The line portion of the circuit is not necessary to claim this advantage
since a back-to-back converter station works similarly.
- The addition of an HVDC circuit does not increase the short circuit currents on existing
AC switchgear.
- An HVDC link can be relied upon as part of a systems generation reserve.

HVDC performance studies have often focused on the converter station, information on DC line
performance is therefore rarely available. However the performance levels of DC lines and
converter stations needs to be separated in order to better understand the issues which impact
upon DC line performance levels.
HVDC Line Performance Survey
In 2010 EPRI conducted a survey regarding the performance of various HVDC schemes around
the world, the results were published in 2011 [2]. The aim of the survey was to get an initial
understanding of various issues related to HVDC transmission line performance. A secondary
aim was to determine areas where further research and study is warranted. A more detailed
survey done in conjunction with CIGRE study Committee B2 is to be formulated based on this
initial survey. Highlights of the observations of the survey responses are shown below.
Design Parameters
The voltages ranged from 400 to 600 kV with one and two pole designs and two, three, or four
conductor bundles. All the lines are greater than 440 miles (708 km) long with no intervening
taps. Most of the lines have extensive end-point grounds, figure 3 shows a) a horizontal ground
electrode configuration and b) a vertical ground electrode configuration.

Figure 3: Ground electrode configurations: a) Horizontal and b) - Vertical

a) b)

Planning Issues
From a planning perspective, the ability of HVDC lines to operate temporarily with a single pole
is helpful. For one HVDC line, single-pole operation has in use for over 10 years.
In HVDC lines with separate structures for each pole, it appears that most consider the loss of
both poles as an n-2 contingency. In most cases, the operation of the HVDC line with a single
pole using earth return was anticipated. This is consistent with normal planning criteria for two
parallel single-circuit AC lines where the loss of both lines (on separate structures) is usually
seen as an n-2 contingency.
In the case of an HVDC line with both poles on a common structure, the loss of both poles would
be considered an n-1 contingency. Given the much higher power flow on HVDC lines and their
limited presence in most AC systems, this may be a very important limitation on the use of
HVDC.
Operations
The maintenance costs for HVDC lines appear to be similar to HVAC lines. In the case of one
utility, where there are three 765-kV HVAC lines roughly paralleling the two HVDC bi-pole
lines for 800900 km, the outage rate for the HVDC lines is about half that of the HVAC lines.
DC line outage rates are in general lower than that of comparable AC line outage rates. One
utility responded that the DC line outage rates monitored over a 14 year period was about a 1/5
of a comparable AC lines outage rates over the same time period.
Perception and Annoyance
Most of the problems with noise, interference, and human discomfort appear to have been
thoroughly reviewed before construction of the lines. A few cases of audible noise problems
have occurred. There were no reports about human perception of ion currents. EPRI is currently
undertaking studies on this topic under its HVDC program.
Development of Very Reliable HVDC Lines
Historically, the high cost and unreliable nature of converter stations dominated any concerns
about the reliability of HVDC lines. The literature and manufacturers data indicate that DC
converters are becoming more reliable and less expensive. As this occurs, overhead HVDC
circuit reliability will come to be driven by line reliability, much as it is for AC circuits.

HVDC Line Design

Similar to the design of AC overhead lines, the design of a HVDC overhead line can be
separated into two main categories:

1. System Designs
2. Component Designs

The system design covers the performance of the overall line or system and includes the
following topics:
- Insulation Coordination
- Lightning and Grounding
- Electrical Effects (Rules and Regulations)
- Other Electrical Designs (Ampacity)
- Mechanical Performance (Weather loading, vibration, galloping)
- Conductor Selection (Life Cycle Costing Method)

The component design covers the performance of the individual components. The performance
of each component is coordinated with one another and must meet the system design
requirements. The component design covers the following topics:
- Insulators
- Spacer Dampers
- Other Hardware
- Conductor (Phase & Ground Wire)
- Structure
- Foundation
- Electrode
- Electrode line

The guide covers voltages up to 600kV. Information above this voltage is not readily available.

Insulators

Effects of Contamination on HVDC Insulators
One of the key issues that impacts the insulation design of HVDC lines and substations is the
contamination performance of the external insulation. Since the beginning of overhead power
transmission it was noted that the performance of external insulation is adversely affected when
the insulating surface is contaminated with airborne deposits such as marine salt or industrial
pollution. These deposits may form a conductive or partially conductive surface layer on the
insulator when wet, resulting in discharges and in severe cases flashover of the insulators.
Under DC energization, it is important to properly design the external insulation to withstand the
contamination conditions. This is illustrated in Figure Error! No text of specified style in
document.-1Figure Error! No text of specified style in document.-1Figure 1-1 which shows a
comparison of the required insulation lengths for HVAC and HVDC systems to withstand
lightning and switching overvoltages as well as the effects of insulator contamination.

Figure Error! No text of specified style in document.-1
A Comparison of the Requirements for Switching Lighting and Contamination for HVAC and
HVDC Systems. See Table Error! No text of specified style in document.-1Table Error! No text of
specified style in document.-1Table 1-1 for the Assumptions Used.
It is apparent from Figure Error! No text of specified style in document.-1Figure Error! No
text of specified style in document.-1Figure 1-1 that on HVAC systems, the insulation lengths
are in most cases determined by either switching or lighting overvoltages. Insulators with a
suitable leakage distance to arcing distance ratio are then selected to meet any contamination
performance requirement. This is in contrast with HVDC systems where Figure Error! No text
of specified style in document.-1Figure Error! No text of specified style in
0
1
2
3
4
5
6
7
8
300 400 500 600 700 800 900 1000 1100 1200
Maximum System Voltage (kV)
R
e
q
u
i
r
e
d

S
t
r
i
k
i
n
g

D
i
s
t
a
n
c
e

(
m
)
0
1
2
3
4
5
6
7
8
300 400 500 600 700 800
R
e
q
u
i
r
e
d

S
t
r
i
k
i
n
g

D
i
s
t
a
n
c
e

(
m
)
Lightning
1
.
4

p
.
u
.
1
.
8

p
.
u
.
2
.
6

p
.
u
.
V
e
r
y

H
e
a
v
y
H
e
a
v
y
M
e
d
iu
m
L
ig
h
t
V
e
r
y
L
ig
h
t
Lightning
V
e
r
y
L
ig
h
t
L
i
g
h
t
M
e
d
i
u
m
H
e
a
v
y
1
.
7
5

p
.
u
.
HVAC HVDC
0
1
2
3
4
5
6
7
8
300 400 500 600 700 800 900 1000 1100 1200
R
e
q
u
i
r
e
d

S
t
r
i
k
i
n
g

D
i
s
t
a
n
c
e

(
m
)
0
1
2
3
4
5
6
7
8
300 400 500 600 700 800
R
e
q
u
i
r
e
d

S
t
r
i
k
i
n
g

D
i
s
t
a
n
c
e

(
m
)
Lightning
1
.
4

p
.
u
.
1
.
8

p
.
u
.
2
.
6

p
.
u
.
V
e
r
y

H
e
a
v
y
H
e
a
v
y
M
e
d
iu
m
L
ig
h
t
V
e
r
y
L
ig
h
t
Lightning
V
e
r
y
L
ig
h
t
L
i
g
h
t
M
e
d
i
u
m
H
e
a
v
y
1
.
7
5

p
.
u
.
0
1
2
3
4
5
6
7
8
300 400 500 600 700 800 900 1000 1100 1200
R
e
q
u
i
r
e
d

S
t
r
i
k
i
n
g

D
i
s
t
a
n
c
e

(
m
)
0
1
2
3
4
5
6
7
8
300 400 500 600 700 800
R
e
q
u
i
r
e
d

S
t
r
i
k
i
n
g

D
i
s
t
a
n
c
e

(
m
)
Lightning
1
.
4

p
.
u
.
1
.
8

p
.
u
.
2
.
6

p
.
u
.
V
e
r
y

H
e
a
v
y
H
e
a
v
y
M
e
d
iu
m
L
ig
h
t
V
e
r
y
L
ig
h
t
Lightning
V
e
r
y
L
ig
h
t
L
i
g
h
t
M
e
d
i
u
m
H
e
a
v
y
1
.
7
5

p
.
u
.
HVAC HVDC
document.-1Figure 1-1 shows that the insulation design is clearly dominated by the
contamination performance requirement.

Table Error! No text of specified style in document.-1
Assumptions Used in the Calculation of the Clearance Curves Presented in Figure Error! No text of
specified style in document.-1Figure Error! No text of specified style in document.-1Figure 1-1

L
i
g
h
t
n
i
n
g

- Flashover rate of 0.6 flashover/100 km/years.
- Tower footing resistance of 20 ohms with a soil resistivity of 400 ohm-meters.
- Calculations based on a ground flash density of 6.0 flashes/km
2
/year.
Note that the lightning curves are relatively flat, since the lightning requirements should be relatively
constant with system voltage. Tower heights increase and coupling factors decrease with increasing
system voltage. These effects, along with the increase in power frequency voltage, combine to produce
a gentle increase in the curve.
S
w
i
t
c
h
i
n
g

S
u
r
g
e

- For HVAC a Gaussian stress distribution and for statistical overvoltages E
2
of 2.6, 1.8, and 1.4 per
unit. (E
2
of 2.6 per unit represents a typical value for high-speed reclosing of breakers without a pre-
insertion resister; 1.8 per unit represents a typical value for high-speed reclosing with a single pre-
insertion resister; and 1.4 per unit represents a value for a breaker with possibly one or two pre-
insertion resisters or with controlled closing.). A line with 500 towers is assumed.
- For HVDC a typical slow front stress level of 1.75 p.u. is assumed. This is associated with the
overvoltages occurring on the un-faulted pole of a bipolar HVDC line during a ground fault on the
other pole.
- Each of the curves sweeps sharply upward, portraying the plot of the strike distance as a function of
the CFO.
P
o
w
e
r

F
r
e
q
u
e
n
c
y

The power frequency voltage requirements are shown as a function of the IEC contamination levels of:
Contamination level ESDD (NSDD = 0.1 mg/cm
2
) Unified Specific Creepage Distance
Very Light < 0.004 mg/cm
2
22 mm/kV
Light > 0.004 mg/cm
2
28 mm/kV
Medium > 0.02 mg/cm
2
35 mm/kV
Heavy > 0.1 mg/cm
2
44 mm/kV
Very Heavy > 0.5 mg/cm
2
55 mm/kV
- For HVAC the use of ceramic insulators with a 170 mm spacing and 380 mm creepage is assumed
for Very Light and Light contamination levels. For Medium to Very Heavy the use of ceramic
insulators with a 170 mm spacing and 350 mm leakage are assumed.
- For HVDC the use of ceramic HVDC insulators with a 170 mm spacing and 545 mm creepage is
assumed for all contamination levels.
- The use of V-strings is assumed.
Note: The Unified Specific leakage Distance is defined as the leakage distance of the insulator divided by
the maximum voltage which will be applied across the insulator. For AC cases this would be the
maximum phase to ground systems voltage, i.e.
3
m
U
.
Figure Error! No text of specified style in document.-1Figure Error! No text of specified
style in document.-1Figure 1-1 also shows that while the insulation length requirements for
switching and lightning is comparable between HVAC and HVDC, the resulting insulation
distances to satisfy the contamination performance requirement is significantly higher for HVDC
than for HVAC. In areas with significant contamination levels this may require large insulation
dimensions, which may influence, and in some cases dictate, the conceptual design of the
project. Choices that may be impacted include:
- The routing of the lines and siting of the converter station, to avoid contamination conditions
- The use of underground cable instead of overhead lines to minimize the exposure of external
insulation surfaces to contamination.
- Utilizing indoor switchyards and converter stations to protect the external insulation surfaces
from contamination and/or wetting.
- The choice of particular insulator assemblies or conductor configurations for the transmission
line or special layouts of the converter stations to accommodate long insulation distances or
special insulation solutions.
It is clear that these choices may have a significant cost impact and it is therefore important to
take the necessary care when doing the insulation design. The importance of this subject is
underlined by the fact that there are now both Cigr and IEC working groups which aim to
formalize an approach for the selection and dimensioning of high-voltage insulators for use in
contaminated conditions. Although this work is not complete yet, there have been significant
progress and the aim with this report is to provide an overview of the present state of the art.
This report therefore contains sections dealing with:
- Differences in the behavior between HVDC and HVAC insulation in contamination
conditions. With special emphasis on why there is a need for longer insulation lengths when
energized with HVDC.
- Experiences with composite insulators, which should include aspects regarding aging and the
possible use of lower insulation distances when the insulation surface is hydrophobic.
- Developments in Cigr and IEC with regards to the establishment of international guidelines
for the selection and dimensioning of insulation.
- An overview of subjects and research questions where more work is needed.
The report is structured around the elements of the contamination process. Briefly, insulators
installed in contaminated environment, e.g. close to the ocean, factories, or other sources of salt
such as agricultural activities or roads (due to salting in winter time), may be covered with a
conductive surface layer during wet conditions. This could result in discharge activity, known as
dry band arcing, which, in severe cases, could lead to flashover. Important parameters that
define the insulator performance are:
- The contamination deposition process
- Insulator flashover characteristics
- Long-term (aging) performance
All these aspects are brought together to derive a simplified dimensioning process, which is
presented later.

Performance of Insulators of Different Materials
The electrical stresses that HVDC places on the electrical insulation are very different to that
under AC energization. Electrical discharges on DC energized insulators tend to be more severe
because of the absence of voltage zeros of the applied voltage. This results in a requirement for
longer insulation lengths. Also the electric field (E-field) along the DC insulators can be very
different to that of the AC. Under DC energization the voltage distribution along a clean and dry
insulator is determined by the internal resistance along the insulator. This voltage distribution is
easily disturbed by a change in surface conditions, unlike the condition under AC where the
capacitive influence of the E-field is much stronger. It is therefore found that small changes in
the surface conditions of DC insulators may result in large changes in the E-field along the
insulator string. On disc insulator strings, for example, the occurrence of single disc
flashovers has been reported. This is when a single disc in an insulator string experiences
flashover while the rest of the insulators are under stressed. On composite insulators the E-field
along the insulator may also be severely distorted due to the presence of space charge.
From the above it becomes clear that AC practices and experiences cannot be directly applied to
DC insulator designs which therefore require specific consideration. In this section the focus is
on the long-term performance of DC insulators and the issues that need to be considered when
selecting and dimensioning DC insulators.
Porcelain Disc Insulators
Most recorded in-service failures of DC porcelain disc insulators have been caused by corrosion
of the zinc-alloy sleeve in the insulator pin which causes it to swell. This in turn subjects the
porcelain dielectric to hoop stresses, eventually resulting in cracks and a mechanical failure of
the insulator. This phenomenon is however restricted to areas with a high contamination severity
and continual high humidity. On modern DC porcelain insulators this problem has been solved
by employing a pure zinc sacrificial sleeve and by the application of a thin polymer coating to
the pin in the area where it is in contact with the cement to block circulating electrolytic current.
Other possible failure mechanisms such as thermal runaway or ion migration in the disc has not
been observed under normal service conditions.
Toughened Glass Disc Insulators
Spontaneous bursting of glass discs on HVDC lines resulting in significant failure rates on pre
1988 installations. These failures were associated with ion movement in the glass material,
especially around inclusions. The ion concentrations in the disc results in a distortion of the
mechanical stresses in the glass and ultimately in a spontaneous shattering of the glass shell.
These failures prompted the development of a special high purity glass for HVDC applications.
The glass used in these insulators is characterized by minimal in homogeneities and much higher
resistivity glass.
In locations with severe contamination it was also found that glass insulators may be subjected to
erosion from arcing activity. In some cases these erosion tracks may be deep enough to
precipitate the shattering of the glass disc. Laboratory tests have shown that glass insulators are
more prone to such erosion than porcelain. Erosion of glass insulators are a sign that the
insulator strings are under dimensioned for the conditions present and it can therefore be solved
by re-insulating with appropriately dimensioned insulators.
DC glass insulators are also normally fitted with zinc sleeves on both the cap and pin, and the pin
cement interface is usually protected by a non conductive coating to block electrolytic current.
Polymer Insulators
Service experience collected during the early 1990s has shown the service experience on
composite insulators on HVDC systems have generally been good. This results should, however,
be seen against the very limited number of insulators (i.e. less than 1000 units) contained in the
sample and their relatively short service life (i.e. less than 10 years) at the time of the survey.
This survey highlighted however some instances of severe erosion in high contamination areas
and corrosion of the end fittings. In more recent reports excellent service experience of HVDC
composite insulators after 25 years of service has been claimed.
Aging in Contaminated Conditions
From service experience it is known that erosion damage has only been observed at sites with a
high contamination severity. Most DC systems with external insulation are however located in
areas with a moderate to low contamination severity, thereby avoiding excessive erosion stresses.
In more recent applications of composite insulators the trend is to reduce the leakage distance
utilized on polymer insulators to about 75% of that which would be specified for glass and
porcelainError! Reference source not found.. This would increase the stress on the material
with regards to its erosion performance. In China this has been recognized to such an extent that
a new multi-stress test has been developed to replicate Chinese contamination conditions.
In a comparison of the tracking performance of polymer insulators under AC and DC
energization when subjected to single and multi-stress tests has shown that under the same stress
(i.e. DC voltage equal to the r.m.s. AC voltage) that the erosion and deterioration incurred is
more severe (both in extent and erosion depth) under DC energization than for AC. This finding
has also been confirmed by inclined plane tests.
From the collected service experience, however, it is not apparent that polymer insulators run a
greater risk of erosion damage as only a very few cases have so far been recorded. This may
partly be explained by the longer leakage distance required on DC systems as compared with AC
to obtain an acceptable performance, which would also reduce the electrical stress on the
material. On the other hand the applicability of especially the inclined plane test to DC
insulators has been questioned. During the inclined plane test the insulators hydrophobicity is
intentionally broken down, while in service this is a primary factor which limits erosion damage
by blocking leakage current.
Aging in Areas with Little to No Contamination
On AC systems it is now well established that the primary aging mechanism on polymer
insulators under clean conditions (i.e. little to no contamination) is corona in combination with
water on the insulator surface. It is believed that the ionized air combine with the water to form
a weak acid that can either directly attack the material or cause it to loose its hydrophobicity.
This normally results in cracks in the housing material and eventually the insulator may fail
mechanically due to a brittle fracture. Also corona plumes (under dry conditions) from the end
fittings or grading rings, when in direct contact with the housing material may lead to the
premature aging and cracking of the housing material. One of the primary ways to inhibit such
aging is the application of corona rings to limit the E-field to below the threshold value.
To date this aging mechanism has not received much attention on HVDC systems but it has been
confirmed that water induced corona occur on HVDC insulators and that it results in a loss of
hydrophobicity. Based on this evidence it can be concluded that there is also a need to grade the
electric field along DC insulators with corona rings to below the threshold for aging due to water
induced corona. Only the design of such corona rings is not as simple as the E-field along the
insulator is determined by the resistivity of insulator materials, the resistivity of the surface layer
and it may be temporarily distorted by the presence of space charge.
Knowledge Gap
The above provided a broad overview of the aging effects present on all types of insulation. For
porcelain and glass insulators the problems are well known and to a large extent already covered
by existing standards. The situation is different for polymer insulators where service experience
with this insulator type is not well publicized. It is therefore difficult to draw firm conclusion on
the most appropriate ways to test the insulators to ensure its long term performance. The
following aspects should be further investigated:
- Research is needed to better understand the aging mechanisms on polymer insulators under
DC energization to determine appropriate design limits.
- There is a need to develop a methodology for the design of corona rings for HVDC polymer
insulators.

Dimensioning Process
This section provides an overview of the insulator design process for HVDC insulation with
respect to contamination. It is recommended that an exhaustive approach be followed to
minimize uncertainties of the input data and its impact on the final design. Consequently such an
exhaustive design approach may need to include the following activities for each different set of
conditions and constraints (e.g. different sections of the line, pollution environments, etc.). A
general overview of the process is provided in Figure 2-1.
There are four activities that must be performed:
1. Candidate insulator selection
2. Site severity determination
3. Selection of the leakage distance
4. Qualification

Selection of candidate insulating solutions: The process is initiated by identifying possible
insulation solutions. In the initial phase this may be based on a simplified design procedure as is
explained later in this section or on past experience. The term insulating solution is used here
in its broadest sense to include:
1. The selection of insulator material and profile,
2. the choice of outdoor or indoor substation configurations,
3. the implementation of mitigation measures (e.g. washing, or coating) as part of the
insulation design or,
4. setting operational constraints such as operating temporarily at reduced voltages during
critical pollution or climatic events.

Site severity assessment: The aim with contamination severity assessment is to obtain an
accurate picture of the contamination severity of area concerned based on data collected over a
relatively long period. The initial assessment is usually based on:
1. Collected performance data on existing lines or substations, preferably DC energized, but
AC data could also be useful.
2. Identification of the type (i.e. Type A or B as defined in IEC 60815-1) and composition
(i.e. type of salts, non soluble deposits etc.),
3. Measurement of the quantity of the pollution present,
4. Characterization of the climate, specifically identifying if there is a prolonged dry season.
5. An assessment of geographical, topological and geological features to identify possible
contamination sources and
6. A survey of present and foreseeable future pollution sources and land use.

Figure 2-221
An Overview of the Insulator Design Process for HVDC Insulation with Respect to Contamination [Cigr Forthcoming].

1
Candidate insulator selection (eg, selection of material,
profile)
Decreasing confidence
2 Design severity determination for the candidates
Information from exisiting d.c.
installations in the area (or
similar)
or
Test station data from d.c.
energised insulators
or
Extrapolation of data from a.c.
installations or test station or
pollution monitoring
or Qualitative severity estimation
3 Selection of creepage distance
On the basis of existing
applicable insulator data from
the field
or
On the basis of existing
applicable insulator data from
laboratory
or
Evaluation by testing where
previous data is not
available/applicable
4 Qualification
Prequalified by previous
experience
or Full scale test or
Agreement to use dimensional
interpolation/extrapolation
or
Agreement to use severity
interpolation/extrapolation
Note: Phases 1-3 may need to be iterated
Insulator design process
For critical installations (e.g. converter stations) the above information may not be accurate
enough, thus resulting in a need for a more detailed assessment. This preferably includes setting
up experimental stations at representative locations with a selection of DC energized insulators
to get an estimate of long term pollution accumulation.
Selection of leakage distance: The required insulator dimensions (notably the leakage distance)
are determined from available service experience or test results. If that is not available
representative tests may be performed on the candidate insulator types to determine the statistical
flashover properties. Under representative test is understood any laboratory test, which is
designed to imitate the natural contamination conditions as closely as possible by replicating the
(1) pollution severity (i.e. ESDD and NSDD) (2) its composition (i.e. type of salt, and non
soluble components) (3) the uniformity of the deposit and (4) wetting conditions.
The above information is then used to do a statistical risk evaluation of the insulation design,
including any mitigation measures, to verify compliance with the required performance criterion.
This should include an assessment of the impact of (1) the number of insulators exposed to the
same conditions, (2) the frequency of pollution events and (3) may include additional safety
factors to cater for any uncertainties in the input data.
Qualification of the insulation design: This is the last step in the process whereby the chosen
insulation design is evaluated either by testing or by a comparison with past experience. An
important part of this process is to obtain agreement and approval of the insulation design from
all stakeholders.
The Simplified Design Method
The simplified design method can provide useful orientation at the start of a project to identify
the range of preliminary solutions. It can also be an effective tool to analyze the outage
performance, and the adequacy of the insulation solutions of existing systems. It is however
important to note that the simplified method has serious limitations which may result in either an
over- or under dimensioned insulation. Aspects that may affect the accuracy of the design are:
- For DC the contamination performance is the dominating factor determining the size (i.e.
axial length of the insulators). Thus any uncertainty in the estimation of the pollution
severity may directly impact the required insulator length. This is fundamentally different to
AC where insulator lengths are rarely impacted by the required contamination performance.
- Higher contamination levels call for a greater increase in the leakage distance on DC
insulators than for AC. Any error in the severity estimate has therefore a larger impact on the
DC insulation dimensions than for AC.
- A further complication in the DC design process is the effect of the electrostatic attraction
which can result in significantly higher levels of pollution accumulation on the DC energized
insulators when compared with AC energized or non-energized insulators. This ratio may
range between 1 and 10 and therefore introduces a large uncertainty in the estimation of the
pollution severity when general environmental or AC-specific site severity methods are used.
- Other uncertainties in the input data which may affect the outcome of the design process are
the non-uniformity of the pollution accumulation on the insulators (e.g. top to bottom ratio,
radial differences and distribution along the insulator) as well as the type and composition of
the contamination present.

An overview of the simplified dimensioning method is given in Figure 2-3Figure 2-3Figure 2-2.
In the following paragraph, each step is briefly discussed for sites characterized by pre-
deposited contamination.
The first step in the simplified dimensioning process is to determine the reference site
contamination severity. This is preferably done by measuring the contamination severity on DC
energized insulators to obtain the most representative results. The measurements should include
Equivalent Salt Deposit Density (ESDD) and Non-Soluble Deposit Density (NSDD)
measurements, which are performed so that the top to bottom ratio and distribution of
contamination along the insulators is also quantified. It is also valuable to perform a chemical
analysis on the contamination present to determine the dominant salts present.
As an alternative, it is also possible, but not recommended, to base the site severity assessment
on measurements on AC energized or non-energized insulators. In such a case it is necessary to
estimate the contribution of the electrostatic field on the accumulation on DC energized
insulators. This is done with the DC/AC accumulation factor K
p
. This factor is selected by
considering the generation and transportation of contaminants around the DC station. General
guidance on the selection of K
p
is as follows:
- K
p
is between 1 and 1.2 in areas where the wind speed is the dominant factor that determines
the amount of pollution carried in the air or in areas where high wind speeds prevail.
Examples of such areas are those with pre-deposited contamination which is close to the
coast or in the desert.
- K
p
range from 2 to 3 in areas characterized by human activity such as mining, industry, roads
etc.
- K
p
falls in the range 1.3 to 1.9 in areas which falls between the above two categories and
which is characterized with extended wind still or low wind conditions. These areas may be
some distance from the coast or from pollution sources associated with human activity.
- K
p
is between 1 and 1.2 in areas where the contamination deposit and wetting occurs
simultaneously.
It should be noted that K
p
can be higher than the values given above when the site location is
characterized by extended dry periods and can be lower when there are frequent natural cleaning
events such as rain.

Figure 2-332
An Overview of the Simplified Insulator Design Process for HVDC Insulation with Respect to
Contamination [IEC/Cigr Forthcoming]
Once the site severity measurements are available then these natural contamination level needs
to be converted to an equivalent laboratory test severity. With this correction it is recognized
that artificial testing differs from natural contamination in a number of important aspects, which
is:
- Type of Salt: Laboratory testing is mostly performed with marine salt (NaCl) whereas
natural contamination layers may often contain less soluble salts such as gypsum (CaSO
4
).
Presently, however, there is no generally applicable method to quantify this effect other than
performing specific flashover testing on insulators with natural contamination.
- Amount of non-soluble material present in the contamination layer: The standardized
laboratory test usually subjects the insulator to a contamination layer with an NSDD=0.1
mg/cm
2
. In service the NSDD levels may typically vary from 0.01 10 mg/cm
2
. Through
Measurements from
AC installations or
on non-energized insulators
as per IEC 60815
Conversion of values to
equivalent dc energized**
Correct the severity for the non-
uniformity of the pollution layer
Correct for
electrostatic attraction
Correct the severity for NSDD
to a reference value of 0.1 mg/cm
2
Correct the severity for type of salt
(Not practical at present)
Correct the severity for the
diameter effect on accumulation
Measurements from
DC test site or installation
ESDD
dc
Top/bottom
ESDD/NSDD
Pollution composition
Correct the severity for any
Statistical data correction
** Eventually taking
account of climate data
(e.g. wind velocity)
Site DC severity
Number of events
Number of insulators
Type of insulator
For each candidate
Insulator type
Required/Design DC severity
. .
Preliminary estimation of the
required USCD
DC
based on
insulator type and material
Correct the USCD for the
effect diameter on flashover
Design USCD
DC
Measurements from
AC installations or
on non-energized insulators
as per IEC 60815
Conversion of values to
equivalent dc energized**
Correct the severity for the non-
uniformity of the pollution layer
Correct for
electrostatic attraction
Correct the severity for NSDD
to a reference value of 0.1 mg/cm
2
Correct the severity for type of salt
(Not practical at present)
Correct the severity for the
diameter effect on accumulation
Measurements from
DC test site or installation
ESDD
dc
Top/bottom
ESDD/NSDD
Pollution composition
Correct the severity for any
Statistical data correction
** Eventually taking
account of climate data
(e.g. wind velocity)
Site DC severity
Number of events
Number of insulators
Type of insulator
For each candidate
Insulator type
Required/Design DC severity
. .
Preliminary estimation of the
required USCD
DC
based on
insulator type and material
Correct the USCD for the
effect diameter on flashover
Design USCD
DC
mathematical manipulation the ESDD level can be corrected as follows:
33 . 0
106 . 0
1 . 0
|
.
|
\
|
=
NSDD
K
NSDD

These corrections result in an estimate of the equivalent site severity at a NSDD=0.1 mg/cm
2
,
which is more-or-less independent of the insulator type which will be utilized.
The next step is to adjust this basic site severity characteristic for each of the candidate insulators
considered. This is done as follows:
- The non-uniformity of the contamination layer: The equivalent ESDD for a uniformly
contaminated insulator can be derived from field measurements and the top to bottom ratio
(T/B) as follows:
( )
( ) | |
33 . 0
1
10 /
/ 38 . 0 1
/ 59 . 0 1
59 . 1

+
= B T LOG
B T
K
B T

The non-uniformity of the contamination deposit along the string is taken into account by
utilizing the average value of the ESDD/NSDD measurements taken along the string.
- Insulator diameter: Larger diameter insulators collect less contamination than small
diameter insulators. The amount of contamination on insulators with a large diameter (i.e. an
average diameter, D larger than 115 mm) can be estimated from measurements on disc
insulators as follows:
35 . 0
115
|
.
|
\
|
=
D
K
D

- Statistical considerations: This correction factor is chosen to obtain a sufficiently low risk
for flashover. This takes account of the number of insulators simultaneously exposed to the
same contamination event, the frequency of contamination events etc. into account.
For line insulators: 4 . 1 =
S
K and
For substations or installations with less than 50 insulators: 0 . 1 =
S
K

The correction factors mentioned above are utilized as follows to determine the required design
DC severity (SDD):
Measured S D B T NSDD P
ESDD K K K K K SDD =
/

The design DC severity corresponds to the contamination severity at which representative
laboratory tests can be performed. At this stage it is also possible to make a first estimate of the
leakage distance (unified specific leakage distance: USCD) that will be required for the project.
The following equations, derived from previous laboratory results, can be used:
- For all hydrophilic insulators:
33 . 0
115 SDD USCD
Basic DC
=
-
- For all hydrophobic insulators:
25 . 0
65 SDD USCD
Basic DC
=
These equations are valid for line insulators and substation insulators with a relatively small
diameter. Insulators with a large diameter generally have a lower flashover voltage than
insulators with a smaller diameter, thus requiring longer leakage distances. The following
correction factor can be used to correct for the effect of the insulator average diameter on
hydrophilic insulators, D [mm]:
30 . 0
115
|
.
|
\
|
=
D
C
D

At present it is not considered necessary to correct for the effect of diameter on hydrophobic
insulators.
For installations at high altitude an additional correction factor can be considered to adjust the
leakage distance for the lower flashover voltage under low air density conditions. The following
equation is proposed for hydrophilic insulators and a height above sea level, H[m]:
|
.
|
\
|
=
8150
35 . 0
H
a
e C
For hydrophobic insulators this effect can conservatively be estimated as:
|
.
|
\
|
=
8150
H
a
e C
Finally the required unified specific leakage distance is then determined as:
Basic DC a D DC
USCD C C USCD =
In areas where the contamination and wetting occurs simultaneously the contamination
severity is determined by monitoring the leakage current activity on the insulators and to
compare peak values with those obtained from Salt-Fog tests. The severity of the site is
expressed as the laboratory severity (in terms of the salinity of the salt water used [kg/m
3
]) which
would result in the same level of leakage current which is observed at the site. This is named the
Site Equivalent Salinity (SES).
Since there is a direct link between the site severity assessment and the laboratory test (i.e. Salt-
Fog) the simplified design process becomes even simpler as there is no need to correct for the
type of salts, NSDD and uniformity of the contamination deposit. At present it is proposed to
correct for:
- Diameter effect: Where the same equation as for pre-deposited contamination is utilized.
- Statistical considerations: Again the same correction as for pre-deposited contamination is
utilized.
The equivalent salinity (ES) is thus calculated as follows:

SES K K ES
S D
=
The basic required unified specific leakage distance can be estimated as follows:
- For all hydrophilic insulators:
33 . 0
15 ES USCD
Basic DC
=
- For all hydrophobic insulators:
25 . 0
15 ES USCD
Basic DC
=
As for pre-deposited contamination it is also necessary to adjust the basic required specific
leakage distance for the effect of diameter and altitude. The same equations can be used thus:
Basic DC a D DC
USCD C C USCD =
Other Considerations
In the foregoing section an overview is given on how to determine the insulator dimensions
based on the flashover performance of the insulators. This does not take into account of any
long-term aging effects if present. At present there is not enough documented service experience
available to give general guidance. There can be large variations in the composition of insulator
housing materials on offer from the different manufacturers and these materials may have very
different abilities to withstand the service stresses placed on it by the DC energization. It is
therefore advisable to collect as much relevant service experience as possible on the particular
insulator make that is a candidate for installation on the HVDC system. In addition it is also
necessary to consider E-field grading along the insulator to avoid the aging effects observed on
AC insulators in clean areas. Indications are that similar aging mechanisms may be at work on
DC energized insulators.

Electrode

HVDC lines can operate in bipolar or monopolar mode. When a DC line is operating in
monopolar mode, a return path is required. The return path can either be a dedicated metallic
return or an electrode. HVDC systems utilizing electrodes have been successfully designed and
put into commercial operation. Some of the earlier HVDC systems utilizing electrodes have
been operated for extended periods of time using the electrode return with no adverse effects.
However, in recent years, concerns regarding the operation of electrodes have been raised.
This guide reviews the latest techniques in establishing a ground electrode, the environmental
constraints and regulations in establishing ground electrodes, challenges and advantages of land
and ocean termination as well as condition assessment.
Introduction
The most cost-effective way of designing a monopolar HVDC systemis to construct it with a
high-voltage conductor and an earth return or sea return, depending on the application. In such a
system,the dc currentpath is through the high-voltage conductor, and the return path is through
the earth or sea return.In monopolar HVDC links, an electrode connection is required at each
terminal. The electrode either carries dc current into the earth or receives dc current from the
earth, so in principle, the earth is utilized as the current return path for the dc current. If earth or
sea return is not desired or acceptable, then a metallic conductor dedicated for current return
referred to as metallic returnisimplemented.
Current return through the earth saves the extra cost of the metallic return and reduces the power
losses as a result of the smaller ground resistance path when compared to the resistance of the
metallic return. The cost of adding an extra conductor for metallic return in the case of an
overhead line differs from the cost of adding a dedicated metallic return cable for monopolar
cable applications. The cost impact for each case needs to be evaluated during the planning phase
of the project.
In the normal operation of a bipolar HVDC transmission system, currents in the positive and
negative poles are equal and in opposite directions, therefore current between converter stations
is essentially zero and practically limited to the tolerances in the control and measuring systems
of the two poles, typically in the range of 10 amperes. During single-pole operation, earth return
could be utilized, or the conductor of the out-of-service pole can be utilized as a metallic return
for the return current. However, to ensure uninterrupted power transfer in one pole during a
sudden block or trip in the other pole, a bipolar system should be equipped with an earth
electrode or a dedicated metallic return.
Earth electrodes perform an important function for either monopolar or bipolar HVDC systems.
Almost 30 years ago, EPRI produced a detailed design manual [1] for high-voltage direct current
(HVDC) earth electrodes. Since then, changes and improvements have been made in earth
electrode design and operation. This guide provides an overview of HVDCground electrodes
including the following specific topics related to design and operation of ground electrodes.
- Latest ground electrode techniques
- Environmental constraints and applicable regulations for ground electrode operation
- Challenges and advantages of ground electrode ocean termination
- Condition assessment of ground electrodes
Converter Configurations
Monopolar HVDC System Configurations
Monopole with Electrode Return
A monopole system with earth electrodes is shown inFigure 1-1. In this configuration, there is
only one high-voltage conductor, and the current return path is through the electrodes. However,
this configuration may not be acceptable in some situations due to environmental concerns. This
configuration has been applied in some older cable systems.

Figure 1-441
Monopolar HVDC with Earth Return
Monopole with a Dedicated Metallic Return
This type of system avoids the concerns raised due to permanent earth electrode current. In this
case a second conductor of the same current rating of the main conductor but at much lower dc
voltage is needed as shown in Figure 1-2.

Figure 1-552
Monopolar HVDC with Dedicated Metallic Return

Bi-polar HVDC System Configurations
1. In a bipolar HVDC system, there are two poles of opposite polarity, and the typical operation
mode is with equal current between the two poles, which means no current will flow in the
electrodes, as shown in Figure 1-3.

Figure 1-663
Bipolar HVDC System with Ground Electrodes
2. In the event that one pole is out of service, the operation of the second pole can continue
either using the electrode as a return current path, or by using the metallic conductor of the
out-of-service pole as a return path, which is referred to as metallic return operation. The
metallic return operation is the preferred mode of monopolar operation, because no current
will then be carried by the earth electrode.
3. Shown in Figure 1-4 is the starting point of a bipolar system with balanced current operation.
Any current in the electrodes would be due to a very small amount of unbalance between the
two poles. Note that the circuit breakers shown in red color are closed.

Figure 1-774
Bipolar System with Balanced Current Operation
I
d

I
d

I
d
+

-

Pole

I
d
Figure 1-5 shows the same system with one pole taken out of service, while the remaining pole
continues its operation with the full dc current flowing in the electrodes.

Figure 1-885
Monopolar Operation Using Ground Return
4. The Ground Return Transfer Switch (GRTS) is then closed; this then connects the
metallic return in parallel with the electrode path, as shown in Figure 6.

Figure 1-996
Monopolar Operation Using Ground Return and Metallic Return

MRTB
GRTS
I
d

I
d1

I
d2

I
d
= I
d1
+ I
d2

MRTB
GRTS
I
d

I
d

5. The next step is to open the Metallic Return Transfer Breaker (MRTB) and complete the
transfer of current from the electrodes to the metallic path, completing the process as shown
inFigure 7.

Figure 1-10107
Monopolar Operation Using Only Metallic Return
The MRTB is a special commutating breaker that can force the current from the electrodes to the
metallic return (see Figure 1-8).

Figure 1-11118
Components of MRTB
Operation of the MRTB is as follows:
1. As the contact S opens, and develops an arc voltage, it generates an oscillation in the circuit
L,C, and S at the natural frequency of the loop, which is known and is part of the design.
2. As the current in S decreases by being shunted to the L and C branch, the arc voltage of S
increases due to its negative current voltage characteristic.
S
L
C
I
Zno
I1
MRTB
GRTS
I
d

I
d

3. A current oscillation is set in L and C, which grows in magnitude with time until its
magnitude equals the current to be interrupted, creating a current zero in S, and allowing it to
extinguish and recover.
Electrode Rating
Existing electrodes in various parts of the world are rated between 800 A and 4000 A [2,3].
However, there is no definition available for the current rating of the electrode. Depending on the
requirements, the electrode rating could either be:
- the maximum current that the electrode could handle under various operating scenarios, or
- the continuous rating of the electrode, so that maximum current for temporary overload could
be higher than the rating.
Reversible Electrodes
Current transfer in a monopolar system is always in one direction. Therefore, the electrode that
injects current into the earth (anode electrode) and the electrode that collects current from earth
(cathode electrode) are fixed.
Current transfer in a bipolar system is also in one direction. However, when a bipolar system
operates in monopolar mode, the electrode that injects current into the earth and the electrode
that collects current from the earth depend on whether the positive pole or the negative pole is in-
service. Electrodes connected to bipolar systems should be able perform as an anode as well as a
cathode depending on the polarity of the out-of-service pole. Electrodes that could perform as
anode or cathode are referred to as reversible electrodes.
Electrode Line and Reliability Consideration
A survey on existing earth electrodes shows that the length of electrode lines varies from 8 km to
85 km [3]. The lower limit of the length of electrode line is dictated by the influence of the
electrode electrical field on the ac grid at the converter station and the upper limit of the length of
the electrode line is influenced by many factors, including finding a suitable area with low
resistivity, proximity to infrastructure such as the ac grid and other metallic structures, and the
availability of suitable land sites. Figure 1-9 shows a bipolar system in steady-state operation
with equal dc currents in the two poles. The current flow into the electrode line at steady state is
essentially zero and practically limited to the tolerances in the control and measuring systems of
the two poles, typically in the range of 10 amperes. Upon the failure of the electrode line (open
circuit) during bipolar operation, the system can continue operating with the converter station
high-speed neutral bus ground switch (NBGS) closed. Operation with the station ground mat is
shown in Figure 1-13Figure 1-13Figure 1-10. This mode of operation can continue until any
event that leads to the loss of a pole takes place. Under this condition, the healthy pole must be
removed from service to avoid any dc currents in the station ground mat. Electrode lines are
generally reliable and designed with two separate conductors in order to avoid this situation.

Figure 1-12129
Bipolar System in Steady-state Operation with Healthy Electrode Line

I
d

I
d

Electrode
line
Open NBGS
Electrode
Station ground
mat
I
d

I
d

Open circuited Electrode
line
Closed NBGS
Station ground
mat
Figure 1-131310
Bipolar System in Steady-state Operation with Broken Electrode Line
Electrode Classification
Electrode stations can be categorized into three categories based on the location of the electrode
as shown inError! Reference source not found..
Table 221
Classification of Electrodes Based on Location [7]

Type of Electrode Description
Land Electrode Located on the land away from the sea or freshwater lakes
Shore Electrode Located on a shore against (salt) seawater. Shore electrodes can belocated
either on the beach without direct contact with seawater at a short distance (< 50
m) from the waterline or in the water, but protected by a breakwater
Sea Electrode

Located (typically on the seabed) in the water at some distance (> 100 m) from
the coastline
Current Blocking Devices
DC currents between ground electrodes cause potential gradients on the surface of the earth. As a
result of these earth potential gradients, dc currents may enter the neutrals of transformers.
Current in transformer neutrals will lead to core saturation and may also cause corrosion. Such
dc currents must be minimized or eliminated. Transformer neutral blocking devices can be
applied, as shown in Figure 1-14Figure 1-14Figure 1-11; however,these solutions can be
expensive depending on the number of transformers involved.

Formatted: English (U.K.)
Formatted: Font: (Default) Times, 12 pt,
English (U.K.)
Figure 1-141411
Schematic of Diagram of Transformer Neutral Blocking Device [9]
References
[1] HVDC Ground Electrode Design, EL-2020, Research Project 1467-1.
[2] International Engineering Company Inc. for Electric Power Research Institute
(EPRI), 1981.
[3] Summary of Existing Ground Electrode Designs, 1998.
[4] M.R. Nielsen on behalf of CIGR Working Group 14.21.
[5] Compendium of HVDC Schemes Throughout the World.
[6] CIGRE Advisory Group B4.04, 2005.
[7] HVDC Ground Electrodes Technical Report S90-003.
[8] Swedish Transmission Research Institute (STRI), 1990.
[9] J. C. Gleadoe, B. J. Bisewski, and M.C. Stewart, DC Ground Currents and
Transformer Saturation on the New Zealand HVDC link, International Colloquium
on High Voltage Direct Current and Flexible AC Power Transmission Systems,
October 1993, New Zealand.

Advancement in Present Ground Electrode Technologies
Electrical Resistivity and Electrical Field Calculation
The selection of a proper ground electrode site is of paramount importance for the reliable
operation of HVDC systems. A properly selected ground electrode site can not only ensure
smooth system operation but also rule out the need for mitigation measures against adverse
ground current effects. It is prudent to carry out a thorough soil investigation at the proposed
location. For such investigations, it is practical for the power utility to make use of the
experience of other organizations / research institutes involved in geological explorations such as
oil and gas exploration companies and other organizations involved in geophysical research and
mapping.
In selecting a particular site, the utility should perform geographical and geophysical surveys that
consider various factors, namely
- Electrical resistivity of earth and water
- Thermal properties such as thermal conductivity and thermal capacity
- Porosity (water and gas permeability)
- Penetration of moisture and influx of water
- Buried and earthed metallic structures within the area of influence
- Electrical infrastructure within the area of influence
- Environmental/land use/landowner considerations
- Accessibility
The current practices of geographical and geophysical surveys are discussed in detail in the EPRI
ground electrode design manual [1]. This section highlights several technological advancements
in the area of electrical resistivity measurements and electrical field calculations.
Low surface electrical resistivity in the local electrode area (i.e., earth or seabed and sea) is
important in order to keep the step-and-touch voltage in the close vicinity of the electrode within
safe limits. The deep earth layers in a larger area are more important in reducing the electrical
field distribution, which in turn, keeps the voltage gradient in the general area of ground
electrode within acceptable limits. A low-voltage gradient ensures that the ground current does
not cause corrosion of buried metal structures and does not enter the neutrals of transformers
installed in the area of the electrode station. For this reason and also to optimize the cost of the
ground electrode, it is important to know the electrical resistivity of the earth up to a depth of 20
km, covering an area with a radius of 10- 20 km from the ground electrode site.
Galvanic and inductive methods have been used to estimate the ground resistivity since the early
days of the ground electrode design [1, 2]. The modern versions of these techniques are:
1. High-resolution multielectrode DC resistivity imaging techniques suitable for shallow
resistivity measurements
2. Magnetotelluric techniques suitable for deep resistivity measurements.These methods were
used in recently completed projects [3, 4].
High-resolution Multielectrode DC Resistivity Imaging Techniques
The high-resolution multielectrode DC resistivity imaging method is an active source method
and provides high-resolution images of the electrical resistivity structure up to depths of several
hundreds of meters. Obtaining the resistivity structure up to few kilometers using this method is
possible, but typically not used, because the large amount of current that needs to be injected into
the ground, employing large electrode separations, is unrealistic. Multielectrode DC resistivity
imaging is a fully automated technique that uses a linear array of multiple current and potential
electrodes connected to a multicore cable. The current and potential electrodes are organized
according to a preprogrammed electrode array configuration. A sequence of measurements of
potential difference are measured and recorded for a number of ordered combinations of current
electrodes injecting a preprogrammed amount of current. Commercially available software is
used to process these measurements into an image of ground resistivity-depth profile.
High-resolution multielectrode DC resistivity imaging methods can also be used for shallow
resistivity measurements [3]. One candidate electrode site occupying an area of approximately
600 meters by 600 meters was covered with seven multielectrode DC profiles at 100 meters
separation to image the shallow structure up to a depth of 150 meters. A multielectrode
resistivity imaging system with 80 electrodes at 10-meter intervals was used, and this setup gave
a total profile length of 790 meters and a penetration depth of more than 120 meters. A typical
resistivity image of shallow ground obtained during the data processing is shown in Figure Error!
No text of specified style in document.-15Figure Error! No text of specified style in
document.-15Figure 2-1. In these measurements, a Wenner-Schlumberger configuration was
employed.

Formatted: Font: Not Bold, English (U.K.),
Check spelling and grammar
English (U.K.)

Typical Resistivity Image Obtained from High-resolution Multielectrode DC Resistivity Imaging
Method [3]
Magnetotelluric Technique
The magnetotelluric (MT) technique is also a fully automated technique.Two basic types of MT
methods available for deep resistivity measurements arenatural source MT and controlled source
MT.
In the natural source MT technique, natural electromagnetic (EM) waves are used to determine
the earth's electrical resistivity structure from a few hundred meters to several hundred
kilometers deep, depending on the frequency of the signal. These natural sources are mainly
located in the magnetosphere and ionosphere, separated from the earths surface by the
nonconductive atmosphere. Because the earth is a conductor, these natural sources induce
secondary fields in the earth.
In the controlled source MT technique, a current source of variable frequency is used to inject
the current into the ground. Depending on the frequency of the injected current, different depths
of penetration of the current into the earth takes place. The basis for the MT theory is provided
by Maxwells equations, which relate electrical and magnetic fields. The deep resistivity
structure is determined by measuring five components of time series data consisting of three
magnetic field and two electric field components [3].
Deep resistivity investigations were carried out using wide-band natural source MT equipment in
India [3]. This equipment consisted of a six-channel data acquisition unit, three highly sensitive
magnetic induction coils, and a GPS module. Commercially available software was used to
analyze the time series data and generate an image of the ground resistivitydepth profile. In one
candidate site, an area with a 10 km radius from the earth electrode site is covered by about 13
MT measurements, each with a time series recording of one day to achieve the desired depth of
investigation, assuming moderately conducting ground conditions. This was done to ensure that
resistivity data of up to a 5-second period with a good signal/noise ratio is obtained. A typical
resistivity image of deep ground obtained during the data processing is shown in Figure Error!
document.-16Figure 2-2.

Typical Deep Earth Resistivity Image Obtained from MTResistivity Imaging Method [3]

The ground electrode site selection process for the Caprivi Link in Namibia [4] also employed
the multielectrode DC resistivity imaging technique for shallow earth resistivity imaging and the
MT technique for deep resistivity imaging. The three-dimensional cube shown in Figure Error!
document.-17Figure 2-3 represents a 27x27x20 km earth resistivity model developed for the
electrode site.
Once the ground resistivity data is available, as shown in Figure Error! No text of specified style
in document.-17Figure Error! No text of specified style in document.-17Figure 2-3,
Good Conductivity up to the
depth of 4150m except the
top layer of about 100m
Formatted: Font: Not Bold, English (U.K.),
Check spelling and grammar
English (U.K.)
Formatted: Font: Not Bold, Check spelling and
grammar
Formatted: Font: (Default) Times, 12 pt, Not
Bold, Check spelling and grammar
grammar
commercially available finite element programs could be employed to estimate the potential
distribution in the area of the ground electrode station due to the dc current injected/received at
the electrode station.

Example of Three-dimensional Earth Resistivity Model That Was Employed to Estimate Potential
Distribution due to the Earth Electrode in the Area. The Color Scale Shows the Logarithm of
Resistivity in Units of m [4]
Figure Error! No text of specified style in document.-18Figure Error! No text of specified style in
document.-18Figure 2-4 depicts the iso-potential contours for a current injection of 1000 amperes
at the ground electrode [4].

Contours of Potential Distribution Estimated for a 1000 Amperes of Hypothetical Current Injection
at the Ground Electrode Station Using Three-dimensional Resistivity Model [4]
Finite Element Modeling Software Programs
Modern software programs such as the Current Distribution Electromagnetic Interference
Ground and Soil Structure Analysis (CDEGS) program [5, 16] have interfaces to accept field
Formatted: Font: Tms Rmn, 12 pt, English
(U.K.)
Formatted: Font: Tms Rmn, 12 pt, Not Bold,
English (U.K.), Check spelling and grammar
Formatted: Font: Tms Rmn, 12 pt, English
(U.K.)
measurements of resistivity data or resistivity models compiled by other software programs.
Once the resistivity model is available, a three-dimensional finite element model of the electrical
and magnetic field is developed. Quantities such as electric and magnetic field strengths, as well
as step-and-touch potentials required for electrode design, could be obtained from these
programs. Additionally, modules are also available for electrode size optimizations.
In addition to estimating the resistivity profile and electromagnetic field distribution, software
programs are utilized to calculate chlorine distribution caused by a sea or beach electrode. Figure
Error! No text of specified style in document.-19Figure Error! No text of specified style in
document.-19Figure 2-5 shows the predicted chlorine concentration distribution for an upgraded
SACOI beach electrode at Punta Tramontana, Italy.

Representation of Chlorine Concentration Distribution [6]
Design of Ground Electrodes
The design criterion for electrode stations is typically 50 years. A CIGRE technical brochure for
ground electrodes design published in 1998 [7] and the EPRI ground electrode design manual
published in 1982 [1] provide detailed explanations of ground electrode design. This subsection
on design of ground electrodes provides a summary of ground electrode design including the
latest developments in the area of electrode design.
Design of Land Electrode
The design aspects of land electrodes have not changed much since the early days of their
application. Some of these design aspects are common for all land electrodes, and some are
specific for given location and electrode design. This subsection discusses five aspects listed as
important in [7].
Heating of Soil
Heating of soil close to the electrode surface is an important quantity to consider in electrode
design. The current industry practice [7] is to design the electrode so that the maximum
temperature is limited to the boiling temperature of the water.
grammar
The boiling temperature of water depends on the water pressure; therefore, the boiling
temperature depends on the height of the water column and the altitude of the electrode site.
Typically, a soil temperature of 85C is considered as the threshold, and temperatures in excess
of this could cause steam formation [7]. The danger of steam formation is that the steam trapped
inside the soil might develop excessive pressures that may cause the electrode to explode. In
general, the overheating of soil reduces the moisture content, and drip irrigation techniques may
be adopted to maintain the soil moisture level for prolonged ground return mode of
operation.However, if the electrode is buried deep below the earth surface and there is water
column above it, the boiling temperature increases and a suitable correction to the maximum
temperature must be made.
The temperature rise of the electrode surface is calculated by using a formula first derived by
Kimbark [8]. This formula assumes a uniform heat conductivity and earth resistivity, and relates
the temperature rise at the electrode surface to the potential of the electrode using earth electrical
resistivity and thermal conductivity. The adopted industrial practice to estimate earth resistivity
and thermal conductivity is to use a soil sample extracted close to the surface of the electrode.
A CIGRE brochure [7] shows that this formula provides a pessimistic estimation of the
temperature rise due to the simplified assumptions, and it suggests the use of a correction factor
of 5 for continuous operation of ground electrode with burial depth less than 3 meters, and a
correction factor between 5 and 1 for deeply buried electrodes with burial depth between 50
meters and 500 meters.
Moisture Content of Soil/Electric Osmosis
Electrodes cannot be successfully operated in dry lands such as dry sand or hard rock areas.
Watering (irrigation) systems to maintain moisture content have been applied to ground electrode
sites. Such practices have been successfully applied in the Nelson River scheme in Canada and
Rihand-Delhi scheme in India.
The current density at the surface of a land electrode must be limited to 1 A/m
2
to avoid electro-
osmosis (i.e., movement of water in the direction of electrical field) [9]. The Rice Flats electrode
of Pacific Intertie used a current density of 0.5 A/m
2
, whereas in the case of Danish shore
electrodes, a current density of 5 to 8 A/m
2
is used because the presence of water is ensured by
the location of electrodes (below the sea level, 20 meters towards the beach) [7].
Geometric Layout
The resistivity profile at the ground electrode station indicates what geometric type of electrode
is suitable and whether one single electrode or a parallel combination of a number of
subelectrodes is required. Land electrodes could be categorized as vertical or horizontal
electrodes based on their geometric layout.
Vertical Arrangement (Borehole Electrodes)
A vertical electrode design is used for electrode stations with conducting layers at some depth.
In addition, vertical electrodes require less land space, and hence this design is an attractive
option for HVDC schemes associated with high currents. This is because a maximum current
density associated with the soil properties and material is used in electrodes. In the case of
vertical electrodes, longer electrodes buried deep into the soil could be used to maintain current
density, whereas horizontal electrodes need a higher cross section, which means a greater land
area to limit the current density.
Vertical electrodes consist of several subelectrodes with each subelectrode having a depth of 50
to 200 meters. Each subelectrode has an inner conductor surrounded by a backfill of a
conductive layer.Figure Error! No text of specified style in document.-20Figure Error! No
text of specified style in document.-20Figure 2-6 shows a typical cross section of a vertical
electrode.

Cross section of Typical Vertical Electrode Arrangement [11]
Horizontal Arrangement
Similar to the vertical arrangement, a horizontal arrangement also has an inner conductor
surrounded by backfill of a conductive layer. Typically, the active part (inner conductor) of the
horizontal electrode is buried about 2 meters below the earth surface. A cross section of a typical
horizontal electrode is shown in Figure Error! No text of specified style in
document.-21Figure Error! No text of specified style in document.-21Figure 2-7.

Cross section Through Horizontal Land Electrode [11]
Many different configurations, depending on the specific design, are possible for horizontal
electrodes.Refer to Appendix A for different configurations. In general, the configurations can
be linear, ring-shaped, or star-shaped arrangements. Symmetrical ring configurations have the
advantage of even current density along the electrode circumference, provided that the ground
resistivity in the area of ground electrode is uniform. All the other configurations have uneven
current distribution and hence unequal current density. Linear electrodes have a higher current
density at the two ends. High current density in an area increases the step voltage along the
surface of the earth. This could be mitigated by burying a part of the electrode at a greater depth
compared to the other parts of the electrode. For example, outmost tips of a star-shaped land
electrode at Bog Roy are buried to a greater depth than the other parts of the electrode to mitigate
high step voltages [12].
In addition to their different shapes, electrodes can be categorized as continuous and
noncontinuous configurations based on the continuity of the inner conductor. Continuous
conductor configurations have continuous backfilling trenches.
Noncontinuous conductor configurations have continuous or noncontinuous backfilling trenches
depending on whether a clear separation of the electrode is required. The inner conductor must
be connected to a common feeding point such as a bus bar where the ground electrode line is
terminated. Typically, the inner conductor is connected to the common point at a number of
equally spaced points along the circumference of the electrode. The idea of noncontinuous
configurations is to subdivide the electrode into separate parts so that irregularities in the current
feeding into each segment can be controlled separately. In some schemes, a small resistance is
added in series with the segments of electrode to reduce the current density around certain
segments. It is also possible to have irregular current distribution as a result of local resistivity
variations. This situation could cause high current densities, causing electric osmosis or heating
of soil in local areas. Segmenting of electrodes is also useful for maintenance where the segment
under maintenance could be disconnected while other segments are in-service.
Material Selection
Almost all the land electrodes built so far have an inner conductor surrounded by some form of
backfill containing carbon to give good contact between soil and the electrode. A CIGRE
brochure that provides guidelines on ground electrode design [7] describes inner conductor
material and backfill material containing carbon, as shown in Error! Reference source not
found..
Table Error! No text of specified style in document.-331 Inner Conductor and Backfill Material [7]

Conductor Material
1 Steel or "mild" steel rods or tubes, 30-40 mm in diameter. The steel conductor is mostly covering
the length of the electrode continuously.
2 SiCrFe rods, commonly 45 mm in diameter, length 1.25-1.75 m. These electrode bars are normally
only part covering the total length of the coke filling, 30-50 percent, which means that the coke
column is used also for longitudinal flow of current. There is a certain risk of an unequal current
density on the outside of the coke filling.
3 Graphite rods, commonly 100 mm in diameter, 1.2-2.4 m in length. As for the SiCrFe rods, the
graphite rods only cover part of the length or depth covered by the coke.
Backfill Material
1 Coke breezesmall particle solid residue left by the cracking process of petroleum refining.
2 Coke Result of distillation of bituminous raw coal.
3 In one electrode station the carbon material is described as graphite powder emulsion.
Step and Touch Voltages
Step voltage is the voltage difference across a step of humans or animals. Touch voltage is
defined as the voltage between the ground surface and any object such as a fence that might be
touched by a person standing close to the object. The permissible touch and step voltages [13] to
which an individual may be subjected to when the electrode is carrying DC current are
determined, based on standards such as IEC standards 60479-1 and 60479-2 and IEEE standard
80. Typical values of potential gradient and touch and step potential are given below. It is
important to note that, during the design and operation stages, due consideration should also be
given to local regulations regarding safety and interference in addition to the conditions at the
electrode site and the general vicinity of the electrode station.
- Touch voltage is typically 20 V. Excessive touch voltages can be mitigated by distance or
section insulation of the objects such as fences.
- Potential gradient on the surface is typically 220 V/m. Excessive potential gradients can be
mitigated by depth of burial.
- Step voltages at the electrode site aretypically 28 V/m. Excessive step voltages can be
mitigated by depth or fencing for areas exceeding such limits.

Both step voltages and potential gradients at a given location near the electrode station can be
accurately calculated by using commercially available software, provided that the ground
resistivity profile in the area of electrode station is known accurately.
Design of Sea Electrodes
In sea electrodes, the current is transferred directly from the electrode to the seawater. Sea
electrodes are connected to the coast using a cable, which increases the project cost. On the
other hand, sea electrodes do not cause electric osmosis problems because the electrode is fully
immersed in the seawater.Existing sea electrodes can be categorized into three construction
models based on the material used in the active part of the electrode [7]. These three models are:
- Sea electrodes using titanium
- Sea electrodes using graphite or silicon chromium iron (SiCrFe) rods
- Sea electrodes using bare copper
Sea electrodes using titanium as the active part (anodic operation)
At least four sea electrodes make use of a titanium anode:
- Fenno-Skan anode Dannebo
- Baltic cable anode Smyget
- Kontek anode Bogeskov
- Grita anode Corf strait
Anodes in the Fenno-Skan, Baltic Cable, and Kontek electrodes consist of titanium meshes built
by adding 20 m
2
modules. Each of 20 m
2
module with dimensions of 1.22 meters by 16.5 meters
makes a subelectrode. Each subelectrode can be rolled in a cylindrical coil of 1.22 meter length
and about 0.8 meters diameter for handling and transportation purposes. Forty 20 m
2
modules
are in each of the Baltic Cable anode and Fenno-Skan anode, making the total meshed area 800
m
2
. In the Kontek anode, there are one hundred 20 m
2
modules, making the meshed area 2000
m
2
. The titanium meshes are covered with special thin layer (5-20 m) of metals resistant to
corrosion. The metal coating on the titanium mesh in the Kontek anode is described as several
layers of precious metal oxides, which has the advantage of a very high percentage ofgas
emissions of oxygen instead of chlorine. Chloride development is low, because of the catalytic
coating and a low current density,which is 2.5 A/m
2
. Titanium meshes are mechanically
protected by polypropylene tubes, fiber concrete, or a layer of natural stone backfill. In the
Baltic cable and Kontek anodes, the current density at the mesh is about 2.5 A/m
2
, and the
resistivity of seawater is about 0.8 m. The Dannebo anode electrode of Fenno-Skan also has
the same physical size as the Baltic cable; however, the seawater has a resistivity of about 1.13
m.
The Grita anode consists of 39 bars of titanium coated with noble-metal oxides, which is a
material widely tested for cathodic protection and has a very high corrosion resistance as well as
a long life expectancy.
Sea electrodes using Graphite or Silicon Chromium Iron (SiCrFe) as active parts (for reversible
operation)
At least two sea electrodes use graphite as an active part of the anode, and at least one electrode
uses SiCrFe rods:
- Ris electrode of the Konti-Skan scheme
- Grosoysoyla electrode of the Skagerrak scheme
- Santa Monica electrode of the Pacific Intertie Scheme
The Ris electrode in the Swedish side consists of 30 horizontal graphite electrodes laid at the
seabed at a depth of 7-10 meters. The electrodes are individually covered with a concrete cover
with coke backfilling. Figure Error! No text of specified style in document.-22Figure Error!
No text of specified style in document.-22Figure 2-8 shows a diagram of a typical horizontal
graphite electrode in a linear arrangement. The Grosoysoyla electrode in Norway consists of 61
graphite subelectrodes placed in a wooden structure with coke backfill. Some of these are
connected in series to provide a better current distribution. The Santa Monica electrode consists
of a linear array of 24 horizontal electrode elements made up of SiCrFe rods suspended 0.5 to 1.0
meters above the ocean floor and located within concrete enclosures. Both graphite and SiCrFe
electrodes can be used in reversible operation (i.e., anodic or catholic operation) [7, 14].

Linear Submarine Electrode (Anodic Operation) [11]
Sea electrodes using bare copper conductors as the active part (for catholic operation)
Five electrodes in Europe use copper as the active part in the cathode:
- Baltic Kathode1 electrode on the German side of the Baltic Cable scheme
- Pampriniemi electrode on the Finnish end of the Fenno-Skan scheme
- Graal-Mritz electrode on the German side of the Kontek scheme
- La Torraccia electrode on the Italian mainland side of the Sacoi scheme
- Otranto Cape electrode on the Italian side of the Grita scheme
Although it is possible to use less expensive materials than copper as the cathode, copper has
been the choice due to the possibility of making reliable clamp connections suitable for the sea
environment by compression or welding. Although many physical configurations are possible,
circular or elliptic shape loops have been used frequently. Error! Reference source not found.
tabulates the parameters of some of the above listed electrodes.
Table Error! No text of specified style in document.-442 Parameters of Electrodes Use Copper as
Active Part [7]
Electrode station BalticKathode1 Pampriniemi
Graal-
Mritz
LaTorraccia
Copper cross-section [mm
2
] 300 300 400 300
Copper cable radius [mm] 11.5 11.5 13.0 11.5
Total electrode length [m] 5620 4500 5100 600
Surface of copper [m
2
] 406 325 417 43.4
Rated current [A] 1364 1250 1500 1000
Current density [A/m
2
] 6.72 7.87 7.20 46.13
Water resistivity [m] 0.8 1.6 0.8 0.2
Gradient of surface [V/m] 5.37 12.60 5.76 9.23

Design of Shore Electrodes
The two types of shore electrodes are [7]:
1. Beach electrodes: the active part of the electrode is positioned on the beach, and the
electrode is in direct contact with sand or ground water but not in direct contact with the
seawater.
2. Pond electrodes: the active part of the electrode is positioned on a small part of the beach
protected by breakwater against waves, and the electrode is in direct contact with the
seawater. Figure Error! No text of specified style in document.-23Figure Error! No text of
specified style in document.-23Figure 2-9 shows an aerial picture of a pond electrode of the
SACOI HVDC system at Punta Tramontana, Italy.

Aerial Photo of PuntaTramontanaShore Electrode (Sacoi HVDC Scheme)
Beach electrodes
The Lovens electrode station of the Skagerrak HVDC system in Denmark, the Te Hikowhenua
electrode station of the Benmore-Haywards HVDC system in New Zealand, and the Albuera
(Leyte) and the Calabanga (Luzon) electrode stations of Leyte-Luzon HVDC system in the
Philippines are examples of existing HVDC schemes using beach electrode stations. The Lovens
electrode has number of parallel graphite electrodes arranged in concrete rings with a coke
backfill. The Te Hikowhenua electrode has a linear array of high silicon chromium iron
electrodes arranged in porous concrete cylinders. These cylinders are buried well below the
lowest tide level. The Albuera and Calabanga electrode stations have a number of silicon iron
rods installed in two parallel rows, with the depth of each rod or subelectrode varying from 10 to
13 meters. Current distribution between each subelectrode is balanced by means of resistors
connected in series with the subelectrode cables. A linear array of subelectrodes parallel to the
coast line is the best suitable geometric layout for beach electrodes [7, 14, and 15].
Usually beach electrodes are buried below the groundwater level, and the electrode current is
transferred to the sea through the groundwater between the sea and the electrode. Groundwater
Linear array of sub-
electrodes
Breakwater
Formatted: Font: Not Bold
Bold
in coastal areas typically has two layers: the top layer of freshwater above the sea level and a
deeper layer of saltwater penetrated from the sea [7]. Therefore, these soil layers are saturated
with either freshwater or saltwater depending on the depth of electrode.
If the active part of the electrode is buried within the freshwater level but above the seawater
level, the anodic action of the electrode produces only oxygen, but there will not be any chlorine
evolution. On the other hand, if the goal is to reduce the resistance of the electrode, it may be
buried below the seawater level, resulting in some evolution of chlorine. Because the soil layers
surrounding the electrode are saturated with water, heating of the soil or electro-osmosis is
unlikely.
Pond electrodes
Pond electrodes are installed in a pondlike area, where the breakwater protects the electrode from
tides (see Figure Error! No text of specified style in document.-23Figure Error! No text of
specified style in document.-23Figure 2-9 for more details). Similar to sea electrodes, current
transfer in a pond electrode is directly to the seawater, and therefore, chlorine and oxygen are
produced as a result of the electrolysis process. Because pond electrodes are surrounded by
seawater, heating of soil, or electroosmosis, is not possible [7].
Among the advantages of pond electrodes is the small area required for the electrode station and
lower installation cost compared to sea electrodes. Additionally, visual inspection of
subelectrodes is simple because inspection requires only lifting the electrode rods out of the
water.
The four existing HVDC schemes that use pond type electrode stations are:
- Punta Tramontana electrode station of the Sacoi HVDC scheme
- Sansum Narrows electrode of the Vancouver Island HVDC scheme
- Ekno and the Massange electrode stations of the Gotland HVDC scheme
- Haenam and the Cheju electrode stations of the Haenam-Cheju HVDC scheme
The Punta Tramontana electrode in Italy consists of 30 Platinum coated titanium pipes. The
SansumNarrows electrode in Vancouver Island, Canada consists of 28 graphite rods. The Ekno
and the Massange electrodes in Sweden have 48 Magnetite (Fe3O4) rods in each electrode
station. The Haenam and the Cheju electrodes in South Korea have 20 Duralumin rods in each
electrode station.
Although, different geometrical arrangements are possible, a linear arrangement of rod
electrodes is the most widely used configuration in existing electrode stations.
Conclusions
Recent advancements in HVDC ground electrode technologies includeadvancements in
resistivity imaging techniques, finite element modeling of electromagnetic fields, computer-
aided optimizing of electrode size with respect to the electrical characteristics such as touch and
step voltages, and the development of new materials for the anode. . Additionally, online
monitoring of the current and temperature distribution of electrode segments could also be
considered as a development in ground electrode technology.
grammar

References
[1] HVDC Ground Electrode Design, EL-2020, Research Project 1467-1, International
Engineering Company Inc. for Electric Power Research Institute (EPRI), 1981.
[2] R. K. Shemie and D.S. Simons, Ground Current Return Electrode Design, Manitoba
Power Conference RHC-DC, June 1971.
[3] R. N. Nayak, R.P. Sasmal, R. Suri, A. Manglik, S. K. Verma, and D. Muralidharan,
Selection of Ground Electrode Sites Using Geophysical Techniques for Resistivity
Measurements for First 800 kV HVDC System in India, Cigre paper B4-105, Paris,
2008.
[4] T. G. Magg, H. D. Mutschler, H. D. Mutschler, J. Wasborg, H. Thunehed, and B.
Sandberg, Caprivi Link HVDC Interconnector: Site Selection, Geophysical
Investigations, Interference Impacts and Design of the Earth Electrodes, Cigre paper
B4-302, Paris, 2010.
[5] CDEGS (Current Distribution Electromagnetic Interference Ground and Soil
Structure Analysis) User Manual, Safe Engineering Services &Technologies Ltd.,
Quebec, Canada, 2009.
[6] C. Di Mario, A. Giorgi, R. Guandalini, L. Mattiello, R. Niccolai, G. Pazienza, C.
Pincella, M. Rebolini, R. Rendina, and G. Simioli, Environmental Issues on the
Development of the New HVDC Link Sardinia Island Italian Peninsula (SAPEI
Project),Cigre paper B4-205, Paris, 2006.
[7] General Guidelines for the Design of Ground Electrodes for HVDC Links, CIGRE
technical brochure, Working Group 14.21 TF2, 1998.
[8] Direct Current Transmission Volume 1, E. W. Kimbark, Wiley Interscience, 1981.
[9] HVDC Ground Electrodes Technical Report, S90-003, Swedish Transmission
Research Institute (STRI), 1990
[10] Summary of Existing Ground Electrode Designs, M.R. Nielsen on behalf of
CIGR Working Group 14.21, 1998.
[11] Siemens Technical Brochure, High Voltage Direct Current Transmission
Proven Technology for Power Exchange, High Voltage Division, Germany.
[12] M. T. OBrien, M. Zavahir, M. Clark, D. Boyle, and P. Milosevic, New Zealand
Bipolar Hvdc Earth/Sea Return Operation Environmental Experiences, Cigre paper
B4-206, Paris, 2006.
[13] W. Ruan, J. Ma, J. Liu, F.P. Dawalibi, and R. D. Southey, Performances of
HVDC Ground Electrode in Various Soil Structures, Safe Engineering Services &
Technologies Ltd, 2005.
[14] Compendium of HVDC Schemes throughout the World, CIGRE Advisory Group
B4.04, 2005.
[15] F. S. Correa, L. R. Wilhelmsson, and J. F. Allaire, Supply to Rapidly Growing
Cities and Areas, Cigr Symposium, Kuala Lumpur, Malaysia, September 1999.
[16] F, P. Dawalibi and F. Donoso, Integrated Analysis Software for Grounding,
EMF, and EMI, IEEE Computer Application in Power, Vol. 6, No. 2, pages 19-24,
1993.
[17] Technological Assessment of 800 kV HVDC Applications, CIGRE technical
brochure, working group B4.45, 2010.

Assignments:

- Dale Douglass ACDC Conversion
- Jerry Reding Conductor Selection
- Danna Liebhaber
- Doug Proctor HVDC Design Questions
- Bob Whapham Hardware Testing
- ATC (Bob Kluge)
- John Chan HVDC Basics
- Gary Sibilant HVDC Insulators

IEEE HVDC Guide Outline-July22-2012

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

IEEE HVDC Guide Outline-July22-2012

Hochgeladen von

Copyright:

Verfügbare Formate

Proposed Outline of HVDC Overhead Line Design Guide

Das könnte Ihnen auch gefallen