Changing Network Conditions AND System Requirements: December 2007

335
CHANGING NETWORK CONDITIONS

AND
SYSTEM REQUIREMENTS
Part I
The impact of distributed generation On Equipment rated above 1 kV
Working Group
A3.13
December 2007
WG A3.13
Changing Network Conditions and System Requirements
Part I
The impact of distributed generation

on
Equipment rated above 1 kV
Copyright 2007
Ownership of a CIGRE publication, whether in paper form or on electronic support only
infers right of use for personal purposes. Are prohibited, except if explicitly agreed by CIGRE,
total or partial reproduction of the publication for use other than personal and transfer to a
third party; hence circulation on any intranet or other company network is forbidden.
Disclaimer notice
CIGRE gives no warranty or assurance about the contents of this publication, nor does it
accept any responsibility, as to the accuracy or exhaustiveness of the information. All implied
warranties and conditions are excluded to the maximum extent permitted by law.
ISBN 978-2-85873-023-0
1
Contributors
Members
Anton Janssen, the Netherlands (convenor)
Que Bui-Van, Canada
Francesco Iliceto, Italy
Mark Waldron, UK
Serge de Azevedo Morais, Brazil
Bob Middleton, Canada
Johann Jger, Germany
Franois Gallon, France
Mietek Glinkowski, USA
Other contributors
John Brunke, USA
Carlos Augusto de O. Peixoto, Brazil
Kenji Kamei, Japan
Max-Steffen Claessens, Switzerland
Hiromichi Kawano, Japan
Y. H. Lee, Korea
Johan Enslin, the Netherlands
Mathias Ramold, Germany
Rene Smeets, the Netherlands
Acknowledgement
CIGRE WG A3.13 owes Mr. Dieter Braun of ABB Switzerland Ltd. for his simulation work
with respect to the TRV simulations as caused by out-of-phase switching of a dispersed
generator in a distribution network.
WG A3.13 gives also thanks to Mr. Maarten van Riet (NUON, the Netherlands) for the
photograph shown in figure 6 and Mr. Mathias Ramold (Universitt Erlangen-Nrnberg) for
the photograph shown in figure 21.
2
PREAMBLE
Networks are changing due to business drivers such as environmental concerns (including
reduced dependence on fossil energy sources), competitive power market, further utilisation
of transmission corridors, multi-directional power-flows in distribution networks, increased
capacity needs, increased efficiency requirements, etc., compounded by the request to
improve or at least not to erode the quality of supply. These developments have led to
technology changes (e.g. distributed generation, wind-farms, compensated lines, phase
shifters, filter-banks, non-linear loads, HVDC, FACTS, advanced protection and control
systems) and consequentially to special requirements for HV equipment, for instance with
respect to harmonics, temporary overvoltages (TOV), transient recovery voltages (TRV), out-
of-phase conditions, power quality issues, etc.
CIGRE WG A3.13 Changing Network Conditions and System Requirements has given
special attention to the consequences of the growth in distributed generation (co-generation
plants as well as sustainable power generation) and to the consequences of long distance
transmission (from remote generating stations to load areas and multiple power transfers
between regions and nations), leading to angular and voltage stability problems and the need
for reactive power compensation. In both cases the interaction between protection and control
systems on one hand and the network dynamics on the other hand will play a dominant role in
the severity and probability of the phenomena that have to be withstood by the applied HV-
equipment. These phenomena have to be considered against a background of increased
utilisation of equipment (in terms of age, loading and voltage stresses), of reduction of size, of
increased complexity, of more intelligence incorporation and the application of more over-
voltage protection and smart devices. Methods of condition and utilisation assessment of
power system components with regard to asset management become progressively applicable.
As the two main topics the impact of distributed generation & the impact of long distance
transmission address different expertise in the power industry, WG A3.13 has chosen to
issue two Technical Brochures. This Technical Brochure, Part I, is on the impact of power
plants in distribution networks on the specifications for HV-equipment. In the other Technical
Brochure, Part II, the impact of long distance transmission will be treated.
Other developments have been indicated in a SC A3 internal working document [1], but the
investigations on the impact of these developments on the specifications of HV equipment are
either treated by other working groups or seen as covered in another way. These
developments are:
The generation of harmonics due to the wide introduction of power electronics and the
design of filter banks, including switching of filter banks (will be addressed in both
WG A3.13 Technical Brochures)
Very high frequency phenomena, related to vacuum circuit-breakers (NSDD: non-
sustained disruptive discharges, dealt with by IEC SC 17A) and GIS-disconnectors (in
relation to transformers, dealt with by CIGRE JWG A2/A3/B3.21 and IEEE PC57.142
working groups [2])
Very steep RRRV (rate of rise of recovery voltage) phenomena, such as for medium
voltage: transformer fed faults, reactor limited faults and short-line faults (presently
HV-equipment is equipment with a rated voltage above 1 kV
3
covered by amendment 2 to IEC 62271-100, June 2006, [30]); and for high voltage: 3-
phase short line faults and long line faults (to be studied by WG A3.19)
Special conditions of high frequencies and/or saturation in relation to the behaviour of
instrument transformers (to be investigated by WG A3.15)
Very special phenomena, like the TRV under severe conditions of clearing inrush
currents (are regarded as so unique that for the moment no further investigations by
CIGRE are required).
A summary of the findings, recommendations and conclusions of CIGRE WG A3.13s studies

is given in Appendix A, i.e. a Report presented at the SC A3 Colloquium 2007 in Rio de
Janeiro [54]. Other publications of WG A3.13 are presented in the Appendices B, C and D,
where specific items for distributed generation, general items and items related with out-of-
phase and synchronizing are treated, respectively.
4
CONTENT
Membership, Acknowledgement
Preamble
Content
Acronyms and abbreviations
1. Introduction
2. Technology of distributed generators

2.1 Windmills
2.2 Co-generation
2.3 Other small power generation technologies
3. Regulations and policy with respect to distributed generation

3.1 Fault-ride-through requirements
3.2 Other requirements
4. Network topologies related to distributed generation
5. Transient phenomena
5.1 Out-of-phase conditions
5.2 Longitudinal dielectric withstand capability of switching equipment
5.3 Short-circuit currents
6. Power electronic equipment

6.1 Harmonics
6.2 Filter-banks
6.3 Filter-bank switching
6.4 Other aspects
7. Possible impact on the Standards
8. Conclusions and recommendations
References
Appendices
A. Changing Network Conditions and System requirements
Studies performed by CIGRE WG A3.13
CIGRE SC A3 Colloquium 2007, Rio de Janeiro, Report PS2-01 [54]
B. Changing Network Conditions and System Requirements [22]

Part I: Distributed Generation
CIGRE SC A3&B3 Joint Colloquium 2005, Tokyo, Report 103
5
C. Changing Network Conditions and System Requirements [3]
IEEE/CIGRE Int. Conf. on Future Power Systems 2005, Amsterdam
D. Dielectric, Switching and System Requirements under Out-of-Phase Conditions,

during Synchronisation and under Comparable Stresses [35]
CIGRE SC C4/A1/A2/A3/C1 Symposium 2007, Zagreb, Report 0701
6
ACRONYMS AND ABBREVIATIONS
AG asynchronous generator
AVR automatic voltage regulator
AWAE American Wind Energy Association
BESS battery energy storage system
BFW boiler feed water
CCPP combined cycle power plants (steam and gas turbine)
CCT critical clearing time
CT current transformer
DFIG double fed induction generator
DG dispersed generation or distributed generation
DSO distribution system operator
EHV extra high voltage
FERC Federal Energy Regulatory Commission
FPCF first-pole-to-clear factor
GT gas turbine
HRSG heat recovery steam generator
HV high voltage (to IEC definitions > 1kV, but also used as MV)
MV medium voltage (not an IEC term, but used for systems > 1 kV and < 52 kV)
NERC North American Electric Reliability Council
OHL overhead-line (OH-line)
PV photovoltaic generator
SCADA supervisory control and data acquisition (system)
SG synchronous generator
SMES super-conducting magnetic energy storage
ST steam turbine
STIG steam injected gas turbine
SVC static var compensator
RRRV rate of rise of recovery voltage or steepness of TRV
RV recovery voltage (power frequency)
T30 test duty 30 of IEC 62271-100
TRV transient recovery voltage
TSO transmission system operator
VT voltage transformer
WEC wind energy converter (windmill)
WECC Western Electricity Coordinating Council
7
1. INTRODUCTION
In recent years, in a number of countries, distributed generation has evolved to a substantial or

significant part of the overall electric power generation capacity, especially by wind energy
converters and co-generation plants. And, the prospect is a further huge increase, both
absolutely and relatively. In Europe, for instance, the installed capacity of wind energy
converters increased from 2.5 GW installed capacity in 1995 to 34 GW in 2004 [4]. The
increasing costs of energy and the societal emphasis on environmental issues have also
stimulated the growth of small combined heat and power generation for residential,
commercial and industrial applications as well as for greenhouses, where the concept of total
energy installations (including a further optimization of co-generation through the
exploitation of CO2 from the exhaust-gasses, the trading of electricity and the storage of heat)
took off. Furthermore, the European Community has put forward targets by 2010 for the share
of renewable energy sources and co-generation in the electric power generation. This will lead
to an increase of the share of renewable energy from 14% to 22%, including hydro power, and
of the share of co-generation energy from 9% to 18%. As the output of renewable sources and
combined heat and power plants is very dependent on the weather conditions and heating
demand, at full load the contribution from dispersed energy sources to the peak power balance
in the year 2010 would consequently be 60% [5][6][7]. In Quebec (Canada), nearly 4000 MW
of wind energy are expected to be developed in the next decade. This capacity accounts for
nearly 10% of the annual peak load and 30% of the annual light load. Similar policies to
stimulate a more sustainable power supply have been implemented in other parts of the world,
thus emphasizing that distributed generation will definitely play an important role. At the
CIGRE SC C1 Session 2006, the Special Reporter formulated this development as being on
the threshold of a new revolution in the large scale electric infrastructures, comparable with
the revolution caused by the regional and international interconnections, halfway into the last
century [8]. A further penetration of dispersed generation into distribution-grids as well as
into (sub)transmission-grids, is therefore a reality.
In the past, until today, it was preferred that during each disturbance in the network, the
distributed generators were immediately switched from the grid, to keep the operational
condition of the network simple and clear, safe and suitable for auto-reclosing. The risk that
small generators would feed the fault arc was eliminated and the fault location procedure
remained straight forward.
However, the onset of dispersed and renewable energy sources are forcing system operators to
re-consider the contribution of the distributed power plants to system services as voltage
control, reactive power support, fault-ride-through capacity, post-fault network restoration,
black-start facilities, frequency support, frequency control, etc. [47]. Depending on the
applied technology of the generator and the driving energy converter, the different types of
DGs can facilitate these services to a certain degree. At this moment, the most important
requirements are related to the fault-ride-through characteristics, as short-circuits in the
transmission systems otherwise will lead to the loss of a substantial part of the power
generating capacity. The adverse contribution of DG without an adequate fault-ride-through
capacity to system unbalance, and even the occurrence of large black-outs, has been pointed
out in [9] (D.A. Baeza); see figure 1.
8
Another important feature is the under-frequency tripping of DGs, where usually the
protection level settings are unnecessarily high (for instance 49.5 to 49.7 Hz in a 50 Hz
system), thus adversely influencing power deficit situations and eventually black-outs.
For example, from a technical point of view, modern design of windmills offer the possibility
to employ their inertial energy, freewheeling independent from the power frequency, to over-
come to a large extent the difficulties during voltage and frequency dips. Also many co-
generation plants are capable of contributing to the overall system performance for a longer
period of time rather than to adhere to the old policy of immediate tripping.
When DGs contribute to these and other ancillary services, clearly large advantages for the
power system are offered, but the networks and their equipment are not naturally equipped for
such a behaviour, at least not in distribution-networks [49]. More precisely, the distribution
grids were originally not designed to accommodate power plants at all (opposite to
transmission and some industrial grids). These new developments lead to special requirements
and special phenomena such as higher short-circuit power requirements, higher short-circuit
currents, other patterns of power flows and short-circuit power flows, different voltage
profiles and voltage variations, fast and advanced protection, complicated controls, situations
of potential islanding, synchronisation, out-of-step switching, phase opposition currents,
harmonics, safety issues, fault location, black-start conditions, system restoration, etc. The
impact of these developments on the network conditions and requirements to the equipment
are the topic of this Technical Brochure.
9
2. TECHNOLOGY OF DISTRIBUTED GENERATORS
2.1 Windmills
According to CIGRE Technical Brochure 185 [10] wind energy conversion systems can be
distinguished by 8 different types of WEC, four of which are with an AG (asynchronous
generator) and the others with a SG (synchronous generator).
1. An AG, squirrel-cage rotor type, directly connected to the grid (usually through a
step-up transformer), with a fixed speed of the wind turbine and a reactive power
consumption dependent on the rotor slip. Reactive power compensation by means of a
capacitor-bank may be added, but with a decrease of the network voltage the reactive
power supply from the capacitor-bank decreases quadratic, while, due to a larger slip,
the AG requires even more reactive power. The AG contributes to the short-circuit
current, but with a relatively fast decay (short-circuit current disappears within 5 to 7
cycles).
2. A SG, directly connected to the grid (possibly through a step-up transformer), with a
fixed speed of the wind turbine. Needs to be synchronised. Supports the voltage by a
controllable reactive power output. Contributes substantially to the short-circuit
current.
3. An AG, squirrel-cage rotor type, connected to the grid by means of a full converter,
with a variable speed of the wind turbine. Supports the voltage by a controllable
reactive power output up to the capacity of the converter. Does not contribute to the
short-circuit current, as the converter limits the current to typically 1.1 pu.
4. A SG, connected to the grid by means of a full converter; see 3.
5. An AG with almost fixed speed of wind turbine and a dynamic control of the slip
(thyristor controlled additional resistance to the rotor). Small possibilities to control
the slip (i.e. speed) and therefore the consumption of reactive power. Further see 1.
6. A SG, mechanically direct driven by the rotor of the wind mill (without any gearing)
and connected to the grid through a full converter. See 3.
7. A double fed induction generator (DFIG) with a wound rotor and a converter that
supplies a variable frequency current to the rotor, thus controlling the rotor speed to
an optimum, that depends on the wind speed. The partial converter (supplying the
rotor only) is less expensive than a full converter, but still offers to a large extent the
facilities of a variable speed drive. Under steady state conditions the machine is
behaving like a SG, albeit with different values for the characteristic parameters. Does
contribute moderately to the short-circuit currents.
8. A SG, mechanically direct driven by the rotor of the windmill, with permanent
magnets in the rotor circuit and a full converter as interface with the network; further
see 6.
The generators with a full converter (3, 4, 6 and 8) will behave more or less identical, from
the point of view of the network. Differences are coming from the dimensions of, mainly, the
converter and from the characteristics of the controllers (for speed/power, for voltage/reactive
power, for protection purposes, for start/stop operations, for mastering and communication).
The rotors are freewheeling and their inertial energy can be employed during system
disturbances. Several possibilities of voltage and/or reactive power control are normally
available. The largest advantage is an optimal conversion from wind energy to electric energy
is achieved. This is a large advantage over the less expensive fixed frequency windmills (1, 2,
5), which nowadays are to be considered as obsolete, especially above 750 kW. One popular
10
WEC is the double fed induction generator (DFIG), being an attractive compromise between
the cheaper technologies without frequency-control and the most expensive technologies with
an utmost control range. The technology originates from controlled drive systems, where the
design is popular as well. WECs up to 5 MW are available, and the developments are still
going strong. The other technology employed is the gearless SG with a full converter (6 and
8).
ASYNCHRONOUS
GENERATOR
GEARING CIRCUIT-
BREAKER
ASG
3~
TRANSFORMER
MV-GRID
REACTOR
CONVERTER
DC AC
ROTOR AC DC
CROWBAR
Fig. 2 DFIG
Full converters are also applied in small power plants such as micro-turbines, fuel-cells,
photo-voltaic systems, energy storage systems as SMES (super-conducting magnetic energy
storage), BESS (battery energy storage system) and flywheels. Due to the application of a
converter, these small power plants are able to limit their contribution to fault currents, to
support reactive power balance and voltage control, to restore systems after disturbances, to
adapt to frequency variations, phase angle variations and voltage variations. Thus converters
have an advantageous influence on the system behaviour and therefore on the requirements
for other components in the network. Besides the added costs, one main disadvantage is the
generation of harmonics, although nowadays converter technologies and mitigation
techniques show substantial improvements. The topic of power-electronics will be discussed
in chapter 6.
11
DFIG
In the double fed induction generator (DFIG) the 3-phase power frequency connection to the stator windings is
electrically separated from the 3-phase variable frequency connection to the rotor windings. The control of the
AC/DC and DC/AC converters has been designed and set in such a way that the frequency of the rotor supply is
equal to the power frequency minus the frequency that corresponds to the mechanical speed of the rotor (or through
the gearing to the mechanical speed of the rotor of the windmill). In other words: the magnetic field rotating along
the periphery of the rotor has a rotational velocity that consists of the speed of the magnetic field with respect to the
rotor (and forced by the variable frequency supply of the converters)plus the mechanical rotational speed of the
rotor itself; the total velocity of the rotor magnetic field corresponds to the rotational velocity of the stator magnetic
field and the phase shift or angle between these two magnetic fields determines the torque of the generator. From the
perspective of the magnetic fields, the induction generator is acting as a synchronous machine.
The construction of the rotor windings, the excitation, the magnetic circuit and the magnetic materials, the stray
fluxes are very different and hence the machine characteristic are different: reactances and time constants (much
lower than the time constants of synchronous machines [11]). The behaviour of the DFIG under transient conditions
therefore differs considerably from that of a synchronous and even an asynchronous generator. Its behaviour is to a
large extend depending on the setting of the many control mechanisms as shown in figure 3. Apart from the
complicated actions undertaken by the converters and their controls, the power electronics is protected by a crowbar
that may bypass the rotor-side converter, thus forcing the generator to an induction generator mode of operation.
However, the crowbar will not short-circuit the rotor windings, but add an extra resistance, up to 20 times the rotor
winding resistance [12], which is already relatively high for wound windings. And the crowbar will not act
immediately, but depending on the circumstances and setting sometimes after several power frequency cycles.
Moreover, it can be de-activated again and even very fast after activating. Besides, the rotational speed of the
induction generator can be far from the speed belonging to the power frequency, so that all these factors lead to the
conclusion that the machine will not act as a normal asynchronous generator either.
Another speciality of DFIG (and synchronous generators with a full converter) during transients is that the rotor of
the windmill may freewheel with respect to the system frequency, thus serving as a freely controllable source of
inertial energy.
Fig. 3 General scheme of WEC-controls, incl. DFIG [13]
12
2.2 Co-generation
On the other hand there are numerous power plants with conventional lay-outs and various
generators driven by: reciprocating machines (gas-motors, diesel-engines), gas turbines
(single shaft GTs derived from heavy duty types or multi-shaft GTs derived from the aircraft,
i.e. aero-derivatives), steam turbines (incineration of refuse, of biomass, of bio-gas, and solar
thermal/geo-thermal installations), small hydro plants (including pump storage), or air-
pressure turbines (compressed air storage installations). Furthermore, combined cycle power
plants (CCPP) can have a GT, ST and generator on one shaft or GT and ST on separate shafts
with each its own generator. As most of these technologies are used for co-generation
purposes, practically all are more or less regarded as contributing to environmental
sustainability. The generators can be synchronous or asynchronous machines, and depending
on the size directly connected to the distribution-grid or through a step-up transformer. These
power plants and the DFIG windmills will have the most impact on the (distribution)
networks. The studies of WG A3.13 are therefore focused on such technologies.
Steam
Cooling water
BFW
system
Combustor
Natural gas
HRSG
Exhaust
Air
Gas
Fig. 4 Combined Cycle Power Plant (CCPP)
The exhaust gases flow through Heat Recovery Steam Generators (HRSG), that contain
components as the economiser, the evaporator, a superheater and a drum in the natural flow
HRSG. The HRSG may be without or with supplementary firing (un-fired or fired HRSG) or
even a fully fired conventional boiler can be used instead. Most HRSGs are of the once-through
boiler type (i.e. without drum) [14]. Such alternatives are of course optimized for the processes
where the heat or steam is employed. With supplementary firing more operational flexibility is
possible, while additional cooling or a by-pass to the boiler gives more flexibility for the
generation of electricity. These features are relevant when ancillary services like fault-ride-
through capability, power frequency support and black-start capacity are required.
13
GT and CCPP
Cogeneration plants are relatively simple for the lower power range (some MW up to roughly 10 MW), but more
complicated for units of several tens of MW or even hundreds of MW. The last category are either fuel fired
boilers delivering steam to steam turbines and exhaust heat or steam for industrial processes or combined cycle
power plants (CCPP) with one or more gas turbines, the exhaust gases of which are used to generate steam for a
steam turbine and exhaust heat or steam for industrial processes or district heating. All rotating equipment of a
combined cycle plant may be installed on one axis (air compressor, gas turbine, steam turbine and generator)or
on separate axis with separate generators. When assembled to one axis, the steam turbine often can be
decoupled by a special clutch, in order to facilitate starting up procedures and independent operation of the gas
turbine. Not only the control is depending on the number of axis applied in the CCPP, but also the dynamic
behaviour of the plant, as the gas turbine/generator sets will react different from the steam turbine/generator
sets or the sets with only one axis.
Heavy duty gas turbines consist of a single axis with a large compressor, combustors and a gas turbine, that
delivers mechanical energy to the compressor as well as to the load (i.e. the generator). The inertia is
determined by compressor, gas turbine and generator. The operational speed is relatively low for a gas turbine,
as the generator runs 3000/3600 rpm or even 1500/1800 rpm. Below roughly 40% of the rated speed, the gas
turbine is not capable to deliver enough torque for the compressor and therefore normally the generator is used
as motor to bring the set up to speed.
Aero-derivatives are gas turbines based on the technology used for air craft gas turbines. Where in aircrafts the
power of the gas turbine is used to drive the air prop, aero-derivatives use the supplementary energy to drive a
generator. The size is up to several tens of MW and they normally use two axis, one to drive the compressor and
the other to drive the generator. The compressor and its part of the gas turbine can run at an optimal speed,
independent from the power frequency. The generator with its part of the gas turbine (called the power wheel)
run at 3000/3600 rpm, but with a very low inertia.
The exhaust gases of the gas turbines are used for a steam generator in order to improve the overall efficiency,
that can go up to almost 60% to generate electricity (Brayton cycle plus Rankin cycle) and to far higher
percentages of the energy efficiency, when the remaining exhaust heat can be exploited usefully. To reduce the
NOx-emission a small amount of water (steam) is injected in the combustor chambers, among other measures. A
special design of rather small gas turbines is the STIG (steam injected gas turbine), where the steam turbine is
replaced by the injection of relative large quantities of steam directly into the gas turbine (that acts as a
combined gas and steam turbine). This combined function of the gas turbine is known as the Cheng-cycle.
Fig. 5 Double axis, aero-derivative gas turbine, also available as STIG
14
For instance, in order to contribute to frequency control GTs should be either operated with some
margin from the rated capacity (say at 95% - 97% of its rated power) or be operated with
temporary over-firing by accepting, as consequence, that residual life is consumed at a higher
rate due to the high temperatures at the first row of turbine blades.
The setting of the droop of the primary power-frequency control is another problem with GTs, as
the power output of a GT is proportional to the rotational speed (i.e. power frequency), but also
the pressure build-up of the compressor, thus giving a reduction of the electric power output
proportional to the square of the actual system frequency (P~freq), right at the moment that
more power is required. GTs therefore have to be temporarily overfired or operated with
adequately widened power regulation range to effectively contribute to primary power-frequency
control in particular in isolated systems subject to large frequency decays. In some power plants
the control of emissions and the control of the gas temperatures prohibit overfiring.
Most of STs of CCPP are operated in the sliding pressure mode (inlet valves fully open), thus not
contributing to the primary frequency control because they react to frequency deviations with
some minutes delay. On the other hand, if the ST is operated with inlet valve pressure control, it
can partly compensate for the inherent deficit in power when the frequency drops, so that the
power plant will contribute to the required primary power control.
Of course the percentage of GTs in the whole installed power generation capacity of the system
is an important parameter. In fact there is a clear trend to a higher share. In [14] figures varying
from 25% up to 50% as a percentage of the national installed capacity have been given for CCPP
and simple GT driven plants i.e. percentages that force GTs to substantially contribute to the
droop control. For instance in the power system of Turkey, in 2006 one third of the 40,000 MW
installed capacity is contributed by gas fired CCPPs.
Black-start conditions put forward special measures to bring the GT generator set up to self-
supporting speed, as well as special precautions for the HRSG. Once-through boilers, if applied,
face problems when loaded under, say, 30%, thus making them less suitable for the initial phase
of system restoration. In this respect it should also be stated that DGs of the size of several tens
of MW should be able to withstand low frequency operation, down to 47.5 Hz continuously and
down to 46.5 Hz or 47 Hz for a few seconds in a 50 Hz network.
The inertia of commercial GTs can be learned from table 4.3 of CIGRE Technical Brochure 238
[14]. For a size of hundreds of MW, heavy duty GT generators have an inertia constant: H = 7 to
8 s. Smaller units show smaller inertia constants: H = 5 to 7 s. Small units (less than 15 MW): H
= 2 to 4 s. Aero-derivative GTs: H = 1.3 to 2.2 s. In this respect a first rough indication of the
critical clearing time can be derived from [15]: CCT = (H/50) s (i.e. infinite power at busbar).
Just for reference purposes: modern windmills (DFIGs) of several MW have typically an inertia
constant of 2.5 to 3.5 s.
Table I: Rough indication of Critical Clearing Times

Technology H (s) (H/50) (ms)
Large GT 78 400
Medium size GT 5-7 350
Small size GT 24 250
Aero-derivative GT 1.3 2.2 180
Windmills 2.5 3.5 250
15
2.3 Other small power generation technologies
It has to be remembered that unbalances in electric power can only be survived by the
availability of enough inertial energy! In that respect, some words may be spent to small
plants. The inertial energy of windmills is only profitable, when they remain connected to the
disturbed grid.
Small hydro-plants use either conventional SG, mostly with very simple or even without
speed regulators and a simple AVR, or induction machines (no speed control, no AVR, no
synchronisation device) with or without shunt capacitors to supply the reactive power. Like
with the large hydro power plants, from these generating sets the inertial energy can be
employed for system stability.
High speed micro-turbines need a power-electronic interface to be connected to the LV-grid

and therefore behave as a converter. The same applies to fuel cells, but opposite to micro-
turbines, by a lack of inertia, fuel cells cannot adapt very fast the power output. Also PV-
systems are connected to the grid through converters and like fuel cells they cannot react fast
to power demands, even worse as they are dependant on the intensity of sunshine. SMES,
BESS (battery energy storage system) and flywheels all are connected through converters and
are able to react fast to the active and reactive power demands for a limited amount of time.
Like the micro-turbines they offer some inertial energy for system stability.
Another development could be the large scale implementation of rather simple co-generation
machines for house-hold applications, replacing the conventional gas fired heat supply (high
efficiency boiler). For instance, by applying a Stirling motor that drives a simple single phase
induction generator. Studies and experiments are going on to see the technical, economical
and societal impact of the application of tens of thousands of such sets. One of the phenomena
that may have an impact on LV-grids (< 1 kV) and, possibly, on distribution-grids above 1 kV
as well, is the facility and/or circumstances that all these single phase induction generators
will start at the same moment, thus requiring too much reactive power, leading to voltage
instability.
16
3. REGULATIONS AND POLICY WITH RESPECT TO DISTRIBUTED
GENERATION
In their rules and requirements Regulators have to take into consideration both the reliability
of the power systems and the support of sustainable energy sources. Easy access for massive
installation of wind-mills and of certain types of co-generation plants to the network may
hamper the prevailing security and safety levels of the public electric power supply systems.
With an increase of sustainable power generation as indicated in chapter 1, it is clear that
requirements have to be put forward with respect to power control, power support, voltage
control, reactive power support, voltage and frequency deviations, fault-ride-through,
frequency control, black-start capacity, and other ancillary services.
Fig. 6 Wind farm in the North Sea
So far, the requirements are restricted to medium size power plants and above (larger than 10
MW or some tens of MW) and even then only a few of the ancillary services are addressed by
some TSOs. Where in the past most attention has been given to the grid connection
requirements for medium size co-generation plants, nowadays the major debate is on the
requirements for wind-farms. Nevertheless, from the system point of view, the individual
small power plants may be equally important for the local network stability. Requirements to
prevent a sudden large loss of power generation in the MV-networks should be established as
well (see also chapter 4).
17
At the CIGRE Sessions 2006 a good overview of the present requirements and developments in
some national gridcodes has been given [16][17][18]. The requirements for wind-farms were
discussed with respect to power control, performance in relation to frequency variation range,
reactive power supply and fault-ride-through. Some countries put forward requirements for
wind-farms above a certain size (for instance 30 MW), others in case the wind-farms are directly
connected to transmission grids (UK) or sub-transmission and transmission grids (Germany).
3.1 Fault-ride-through requirements
Most gridcodes nowadays put forward requirements with respect to fault-ride through capacity,
but they are still struggling with the precise descriptions of the requirements. However, there is
worldwide a trend to define the voltage drop at the high-voltage side of the transformer, that
connects the wind-farm to the (sub)transmission grid. Just as an indication, three examples of the
fault-ride through requirements are shown in figure 7. Mostly the wind-farm has to ride through
a voltage dip down to 15% of the rated voltage (Spain and Brazil: 20%), but in England and
Wales (indicated as UK in the figure below) the code requires ride through down to 0% of the
rated voltage. The duration of the dip is 140 ms in UK, up to 625 ms in Germany (500 ms in
Spain and Brazil). Hydro Qubec requires an initial ride through at 0% of the voltage up to 150
ms.
After clearing the fault in the (sub)transmission grid, it generally takes some time before the
voltage is recovered to its original value. The wind-farm has to sustain this post fault period as
well. Moreover, it has to contribute to the recovery of the voltage. In [13], the developments and
debates in the USA, at the end of 2005, are given. In the meantime the FERC (Federal Energy
Regulatory Commission), after hearing the arguments of stakeholders like the AWEA (American
Wind Energy Association), the WECC (Western Electricity Coordinating Council) and the
NERC (North American Electric Reliability Council), decided not to follow requirements as
strict as put forward for instance in Germany, but to require only a fault-ride-through capability
of wind-farms for 150 ms and for a residual voltage > 15 % at the connection point.
Supergrid Voltage Level

(% of Nominal)
100
90
85
80
75
UK HQ GER
25
15
0 1 1.2 2 2.5 3 Time (s)

0.14 0.15
Fig. 7 Examples of Fault-Ride Through requirements
18
There is still world-wide a lot of debate, but, at the same time there are also efforts to harmonise
the requirements. Important to note is the clear trend to require that during voltage dips wind-mill
farms have to stay connected to the network.
These requirements are more or less in line with requirements put forward for conventional
power plants of the same size and/or grid connection (subtransmission or transmission). For
individual wind-mills and smaller power plants no general trend can be seen and utilities still
accept or even apply the old policy of tripping the power plants as soon as a disturbance occurs.
This policy is related to safety issues, which of course is of great concern as well for the
gridcodes.
3.2 Other requirements
Apart from fault ride through requirements for voltage dips, there are also requirements for
fault ride through during temporary overvoltages, asymmetrical faults and frequency
variations. As an example Hydro-Quebecs requirements regarding fault ride through
capability of wind plants connected to its transmission network, as described in [20], and
presented below:
The wind plants must be designed, built, operated and protected so as to remain in service
without tripping any wind plant equipment during the voltage and frequency variations
shown in Table II, Fig. 7 and Table IV as well as during various fault conditions summarized
in Table III. These voltage and frequency variations can occur simultaneously and could be
reached when the system experiences extreme contingencies.
Table II: Minimum time periods during which wind

plant must remain in service without any tripping
during voltage variations
Voltage (pu) note 1 Minimum time
V < 1,0 see Fig. 7

Note 1: Positive sequence voltage on the high-
1,0 V 1,10 Long period note 2 voltage side of the switchyard
1,10 < V 1,15 300 seconds Note 2: According to the time needed to bring back
the network to a normal state, within the allowed
1,15 < V 1,20 30 seconds steady state voltage range (up to a few hours)
Note 3: Temporary blocking is allowed for facilities
1,20 < V 1,25 2 seconds using power electronics when the voltage exceeds
1,25 < V 1,40 note 3
0,10 second 1,25 pu Normal operation must however resume
immediately once the voltage drops below 1,25 pu
V > 1,40 pu note 3 0,033 second
19
Table III: Faults for which wind plants must remain in service
Fault Number of cycles Voltage Comments
9 normally cleared fault
Three-phase
45 0,25 pu (1) remote fault
Double phase-to-ground
Phase-to-phase
Single-phase to ground 15 delayed cleared fault
(1)
Positive sequence voltage at high-voltage side of switchyard during a remote fault
Table IV: Minimum period of time during which wind plant must remain in service
without any tripping during frequency variations
Under Frequency (Hz) Over Frequency (Hz) Minimum time
59,4 F 60,0 60,0 F 60,6 Unlimited
58,5 F < 59,4 60,6 < F 61,5 11 minutes
57,5 F < 58,5 61,5 < F < 61,7 1,5 minutes
57,0 F < 57,5 10 seconds
56,5 F < 57,0 2 seconds
55,5 F < 56,5 0,35 second
F < 55,5 F 61,7 Instantaneous
The requirement regarding the ability to withstand the frequency variations listed in
Table IV is also applicable to any wind plant connected to Hydro-Quebec transmission
system through the distribution system.
In the USA the technical requirements for the interconnection of individual DG are given in
IEEE Standard 1547, a series of standards that is under development. At the moment two
parts have been approved and published [55][56]. In case of system disturbances, the main
standard IEEE Std 1547 emphasizes the fast disconnection of DG rather than putting forward
fault ride-through requirements. For instance: a disconnecting time < 160 ms for a voltage at
the point of common coupling less than 50% or higher than 120% is required or for a
frequency smaller than 57 Hz (59.3 Hz for DG < 30 kW) or higher than 60.5 Hz. Larger
clearing times are allowed for smaller voltage and frequency deviations.
Other requirements are related to voltage and reactive power control, online information
sharing with the TSO or DSO and active power control by the TSO. Especially in Spain REE
is applying a control center to dispatch the power from windfarms larger than 10 MW. Further
requirements are put forward to prevent islanding, for synchronisation, for safety
(disconnectors, earthing, blocking), but these requirements are specially for individual DGs
[57].
There is, certainly in Europe [58], a tendency to harmonise and standardize the requirements
for small power plants, windmill farms and individual wind mills. Standardisation is also in
the interest of manufacturers and operators of small generators.
20
4. NETWORK TOPOLOGIES RELATED TO DISTRIBUTED GENERATION
As illustrated in figure 8, DG in distribution, sub-transmission, and industrial networks leads

to a structural change of the power flows, as the generated energy is mostly independent from
the local energy demand. Changing power flows are of importance also for protection experts,
as protection systems have to cope with the related currents and power flow patterns. Short-
circuit currents will show directions and amplitudes that are dependent on the operational
mode of the DGs. In addition, in the DG systems (especially distributed windmills) the short
circuit profiles along distribution feeders are different from those prevailing in the
conventional systems with single source feeding. These different profiles could be difficult to
understand intuitively and can impact the protection coordination and the short-circuit current
withstand and interruption requirements for the related switchgear.
G G
G
G
G G
Fig. 8 Structural change of power flows in distribution grid
For example, in a MV distribution-network without DG, the fault current is determined by the
impedance of the power-transformer plus the impedance of the distribution-feeder up to the
fault location. The ratio between these two impedances is such that the peak value of the short
circuit current is already reduced considerably at a distance of, say, 1 km from the substation.
At a distance of several km also the AC-value of the short-circuit current is reduced
considerably. But, in the case of small size DG, the impedance of the generator plus step-up
transformer is much larger than the impedance of the HV/MV power-transformer supplying
the MV network: up to some orders of magnitude for each generator. This means that the
impedance of the distribution-feeders is of far less influence, even when a number of small
generators are connected to the same feeder. The impact of the DG on the amplitude of the
short-circuit current may be of importance or not, depending on the electrical distances from
the grid source, on the rated power and the technology and number of DG and the several
impedances.
In industrial plants with self-generation often the power plant is directly connected to the
main MV busbar or even through a separate transformer (i.e. step-up transformer) to the trans-
mission network. Similar scenarios can be found in large power plants with auxiliary
generators (for instance: a pre-inserted GT, a GT driving the feed water pump, a GT for the
21
auxiliary electric plant, a black-start unit). In distribution grids there are many instances where
separate OH lines/cables connect the DG directly to the main busbar in the substation. In
transmission or sub-transmission grids, the DG are connected by separate transformers. For
wind mills one can also find dedicated distribution networks, separate from the distribution
network to supply the load (in some regions of the Netherlands typically new, non-redundant,
20 kV public networks are used to connect wind mills while old, redundant, 10 kV grids are
used to supply power to loads[21], see figure 21). In addition wind farms, which are power
plants consisting of many wind mills, are usually directly connected through step-up
transformers to the transmission network.
At transmission levels less problems are expected with such power plants, as transmission
grids are used to connect large generators. Moreover short-circuit powers are normally far
larger than those of the power plants, and short-circuit currents are cleared much faster than
many of the short circuits in distribution networks. Multi-directional power flows and
complicated short-circuit current patterns are not an issue at the transmission level. Out-of-
phase conditions in case of DG are not so different from out-of-phase conditions of other
plants and can therefore be treated equally (specified duties for switchgear, special protection
and control equipment, synchro-check and synchronising equipment, interlocking, etc.).
Many small power units are protected in such a way that for any disturbance in the network,
they are immediately switched off by a circuit-breaker at the low voltage side or at the high
voltage side of the step-up transformer (after roughly 50 to 100 ms), thus reinstating the
passive feature of the distribution network. In this way they enable the conventional selective
protection, auto-reclosing and fault location in the distribution grids and prevent out-of-phase
conditions, unintended island operation, safety problems and severe damage to the power
plants. However, when distributed generation has expanded to a significant share,
requirements have to be put forward that these generating plants contribute to the stability of
the network and remain connected for a certain period of time so that voltage sags can be
limited in amplitude and duration, overloads are reduced and regional power deficits are
avoided [48]. The ride through time is specified as several hundreds of ms for smaller or
special units up to more than 1 s for large conventional power units (see chapter 3).
In order to profit from DG and to minimize its drawbacks for the network have led to an
innovative optimization of protection settings and control circuits. In this respect, reference is
made to [45] where the following recommendations are outlined:
DG connected to faulty MV or LV distribution-feeders

Disconnect SGs immediately after fault clearing (before the auto-reclosing instant)
Disconnect AGs, if the short-circuit clearing time and dead time before reclosure of
the MV-line are too long and inertia of generators is low
Disconnect all DGs in case of permanent faults
DG connected to healthy MV or LV (radial) distribution-feeders in case of short circuits in an

other MV or LV feeder supplied by the same substation
SGs and AGs should be kept in service if short circuit is cleared in the standard low
time (say, within 200 ms for multi-phase faults)
Protect SGs against loss of synchronism
Protect AGs against sustained overspeed and overcurrent
22
DG connected to healthy MV or LV (radial) lines in case of faults in the supply HV network
SGs and AGs should be kept in service in case of multi phase faults in the HV network
cleared in short time ( 200 ms) and in case of 1-phase-to-ground faults in HV
network even if of long duration (say, 0.5 1.0 s)
Islanding
Disconnect automatically SGs under islanding conditions, unless the network,
protection and control is specifically designed for stable supply of an island from DG
Disconnect AG under self-excitation conditions.
However these measures are only feasible when short protection clearing and auto-reclosing
times (100 ms to 200 ms) can be reached, which may not be the case in distribution-networks
[22]. But, as stated before, DFIG and generators connected through full converters should
inherently be able to fully ride through fault conditions, even for longer clearing times, and
resume power supply within a few cycles after fault disappearance.
Already today, upgrading the protection and voltage control systems and setting-up new
operation and maintenance rules in distribution-grids with a large amount of power generation
is an issue. As local DG increases, the networks will need to be uprated in distribution
capacity or heavily utilized up to their limits, depending on the type of dispatch-ability of DG,
on configuration of the network and on size, number and location of DG. Information on the
operational mode of small power plants is normally not well known by the utility, so it is very
difficult to estimate the load profile. This leads to the situation, where the redundancy in the
power supply system is not fully known. The redundancy is partly covered (to an unknown
extent) by the DGs, which leads to the inherent requirement that not all power generators
should be tripped at the same moment. At the distribution-level, the uncertainties on power
flows and the anticipated short-circuit current flows are a problem, together with the huge
increase in installed capacity in some networks, as indicated also in [23]. Use can be made of
the dynamic loading capabilities of the cables, OH-lines and transformers involved, but the
overloading capacity of switchgear is still an area for research (CIGRE SC B3) and
Standardization.
The need to improve or at least maintain power quality as the complexity of distribution
networks increases, must also be considered. Unpredictable power flows and differentiated
reliability claims may lead to a requirement for sophisticated control and protection in
addition to new switching duties. Network components may need to fulfill new duties such
as controlled switching, very short fault clearing times, dynamic loading capabilities
(utilization of cycling loading capabilities), intelligent switching, advanced communication,
voltage regulation and power flow control. New developments in power electronic
technologies such as active filters, SVCs and flexible series compensation may also influence
basic system parameters such as (short-circuit) current profiles, voltage profiles, TRVs,
overvoltages, harmonics and voltage asymmetry. They may also result in the development of
unusual network topologies compared to historical practice.
As stated previously, operating in an island mode is rather exceptional as it is very difficult to

achieve the proper tuning between DG power and load at the moment of islanding and an
acceptable frequency regulation capacity for a public network. Normally the large
interconnected transmission network serves as an enormous reservoir to deal with power
23
unbalances and the claim is correct that without this large interconnected facility the huge
penetration of DG, and especially of wind mills, would be impossible. The multiple power
transfers between regions and countries (see figure 9), which may be contributed by the
electricity production patterns from DG, is the subject of the CIGRE WG A3.13 Technical
Brochure, Part II.
Fig. 9 European Energy and Network Policy - Priority Interconnection Plan
24
Island operation
To date there are very few examples of public distribution networks that can be operated part-time decoupled in
an island mode [57]. On the other hand the tendency towards achieving islanding capability using distributed
generation is noticeable in the distribution networks, as already applied in industrial plants with auto-
production. This development is also driven by the new regulations of a liberalized power market. Another trend
that can be seen is the use of isolated networks supplied by distributed generators like windmills or solar power
modules combined with generators powered by biomass energy or gas turbines or reciprocating motors thereby
balancing the energy demand.
A radial network with an uni-directional power flow (passive network) can be protected by relatively simple
protection principles, such as undirectional overcurrent/time relays. Radial and meshed networks with a bi-
directional power flow need more complicated protection systems and even new protection designs to prevent
the selectivity of the protection being impaired. Additionally, protection and circuit-breakers have to be designed
for short clearing times to prevent stability problems. Communication between protection systems and between
control systems (reverse interlocking, blocking, trip signal transfer etc.) are needed to ensure selectivity and to
minimize fault durations.
Such systems are equally required to facilitate network decoupling and islanding. An example of a three-stage
decoupling strategy can be given. It is based on the direction of the reactive power flow and the magnitude and
duration of the voltage drop during the fault. For a reasonable islanding only faults taking place outside of the
network which is foreseen to be islanded will be considered. A directional reactive power measurement can
exactly discriminate between internal and external fault. This ensures that decoupling is initiated only in case of
an external fault.
Taking into account the ability to achieve a stable recovery of the distributed generators, the amount of voltage
drop and its duration is decisive. This is why a three-level voltage criterion is introduced. The greater the
voltage drop, the faster the decoupling process will be done. To find out applicable voltage and time delay
thresholds dynamic simulations are necessary. This avoids unnecessary decoupling and islanding.
As traditionally passive distribution networks become active, an issue of importance is safety;

for instance: clearly de-energized and/or earthed parts of the network, reliable status-
information on the generators through SCADA-systems, and interlocking functions where
necessary. Furthermore functions as synchro-check, synchronisation signals, voltage check
(auto-reclosing) and frequency check (islanding) have to be built in at the location of the HV
switchgear. These measures are more or less comparable with those for large power plants
connected to transmission networks and serve to a large extent as protection of the power
plants. Together with the requirements for more advanced and faster protection systems, this
means that more intelligence and communication has to be incorporated in the HV-devices
[24][25][26][27]. The IEC standard 61850 will support this new requirements of
communication in stations.
In the case where islanding of a distribution network or a part of the network is foreseen,
functions are needed such as frequency control, voltage control, adaptive protection, system
restoration, facilities for synchronisation and synchro-check. A few examples are listed:
frequency control
an autonomous frequency control
with settings different from the control during interconnected operation
load shedding facilities
generation dispatching facilities
spinning reserve
facilities to adapt easily the generators output (for instance: by-pass of boiler, over-
firing GTs)
25
voltage control
advanced voltage control (droop control)
reactive power sources
facilities to prevent voltage instability
adaptive protection
different settings due to different power and short-circuit power flows
system restoration
black-start facilities
facilities to tune the active load to the restoration of generation
facilities to tune the reactive load to the restoration of generation
synchronisation
enough facilities for synchronisation within the distribution grid
facilities to re-connect with the surrounding networks
enough possibilities for synchro-check.
Switchgear and other equipment have to be able to support these functions and to fulfil the
related duties [24]. Moreover, the grid, the HV-equipment and the secondary equipment have
to handle system conditions with a deviating power frequency and/or voltage deviations
(amplitude, harmonics, negative sequence and zero sequence voltages). Especially the
combination of transient phenomena under deviating system conditions have to be considered
carefully, when islanding is applied.
26
5 TRANSIENT PHENOMENA
When considering network stability, control/protection systems and power quality, a high
short-circuit power is normally recommended. The consequence however will be the
occurrence of high short-circuit currents. Network equipment is designed to cater for the
prospective short-circuit conditions, however particular aspects, such as high DC-components
and/or out-of-phase performance, need special consideration when network topologies change
and generators are added. Increasing quantities of distributed generation increase the
likelihood of out-of-phase conditions in transmission, sub-transmission, distribution and
industrial grids. A possible consequence of lightweight turbine-generator-sets employed is
that conditions such as full phase opposition and even pole slipping might occur. Under these
circumstances the out-of-phase current is defined primarily by the out-of-phase angle and the
network impedance. With large out-of-phase angles, the out-of-phase current is probably
larger than the contribution of the generator to a busbar short-circuit current. In addition, the
TRV conditions related to full phase opposition can be more severe than those experienced
when interrupting simple short-circuit currents [28].
5.1 Out-of-phase conditions
Power generating units are stressed at the moment of the short-circuit occurrence.
Furthermore, these units are stressed again a few cycles later when they are switched off and
the brake is applied (especially windmills). The generators which remain connected to the
grid will accelerate due to the lack of load. Both frequency and phase angle will start to
deviate from that of the main grid and some low inertia power generators, like aero-derivative
GTs, will accelerate quite fast. Anyway, at the moment that the short-circuit is cleared by the
relevant circuit-breakers, the small generators suddenly have to face the system frequency and
system phase angle again. For SG, a relatively large out-of-phase current will flow between
generator and the system, leading again to high dynamic stresses, maybe even larger than the
stresses due to short-circuit currents; see figure 19. This applies also to AG and even to DFIG
[29]. Users are warned that loss of life and serious damage to shafts, gears, bearings,
(gas/steam/wind)turbine blades, etc. can occur [36]. The stresses have to be withstood by the
HV-equipment as well, but the generators short-circuit or out-of-phase currents are smaller
than the system fed short-circuit currents. Breaking the out-of-phase current is an optional
duty for HV circuit-breakers[30][31][32][33] and its specification probably has to be re-
considered [22][35].
In figure 10 the transient behaviour of a SG, as simulated with NETOMAC, is shown with tf
representing the 3-phase fault duration and toop the out-of-phase condition. Immediately after a
fault clearing time of 233 ms, the out-of-phase angle is 90, but it is still increasing up to
about 150, 150 ms later. A larger fault clearing time will lead to an out-of-phase angle of
180, but the impact of 150 is only a few percent lower than the impact of an angle of 180
[35]; see also figure 18.
For the simulation of the TRV, when clearing the out-of-phase current, it is assumed that the
out-of-phase angle will reach 180, but such an assumption is quite normal as utilities tend to
specify their generator circuit breakers for full phase opposition [19].
A simplified single line diagram of the network simulated is given in figure 11. In the next
figures details of the TRV due to clearing the phase opposition current Ioop are elaborated for
27
several network and service conditions: (i) for the load of the 20 kV network various values
have been chosen, (ii) the short circuit power of the 110 kV-network has been chosen to be
low or high and (iii) a capacitor to reduce the RRRV (rate of rise of recovery voltage) can be
connected to the transformer side of the generator circuit-breaker or not.
100
terminal volt . /%
GEN
0
-100
5
current [ pu]
GEN
0
-5
300
active power /%
GEN
100%
reactive power /% 0
GEN
-300
180 180.
angular
displacement/gra 90.
GEN
0
-180
10
speed variat. /%
GEN
0
tf t oop GEN 25 MVA TA=5.08 sec

-10
0 0.352 0.704 1.056 [s]
Fig. 10 Simulation of 25 MVA synchronous generator (CHP) dynamic behaviour

during/after 233 ms short-circuit in distribution-network
4 || Cables 3x1x150 mm 2 2 || Cables: 25 m 4 || Cables: 25 m

Circuit
total length: 6500 m
Breaker 3x1x300 mm 2 3x1x300 mm 2
'' 25 MVA
Ssc =6000 MVA 10 kV
optional
110 kV 20 kV P, Q switched off

three-phase fault
Fig. 11 Single line diagram of simulated MV-network
In figure 12 the results of the simulation for the case of a high short-circuit power at the 110
kV-side (6000 MVA), no load (i.e. no damping) of the MV network (0% load) without the
application of the optional capacitor can be seen. A high frequency component caused by the
step-up transformer and cables at one side of the generator circuit-breaker and a low
frequency component caused by the 110/20 kV transformers and the cables at the other side of
the circuit-breaker are visible. As a comparison, figure 13 shows the same case but with a
28
rather low short-circuit power at the 110 kV-side (400 MVA). The higher frequency seems
not to be changed, but the lower frequency is about 50% decreased. The amplitude of the
high-frequency component is somewhat lower, whilst that of the low-frequency component
has increased. Despite some difference in the waveshape, the peak-values of both TRVs are
almost identical, but: note the boosting effect of the low frequency component in figure 13.
90 90
*10 3 *10 3
75 75
60 60
45 45
30 30
15 15
0 0
0.0 0.4 0.8 1.2 1.6 *10 -3 2.0 0.0 0.4 0.8 1.2 1.6 *10 -3 2.0
(file TRV_load0_6000_180_4.mat; x-var DEV3_SW1a(1)) DEV3_SW1a(2) (file TRV_load100_6000_180_4.mat; x-var DEV3_SW1a(1)) DEV3_SW1a(2)
Fig. 12 Simulated TRV for 6000 MVA, 0% Fig. 14 Simulated TRV for 6000 MVA,
load, without capacitor 100% load, without capacitor
80 80
*10 3 *10 3
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
0.0 0.4 0.8 1.2 1.6 *10 -3 2.0 0.0 0.4 0.8 1.2 1.6 *10 -3 2.0
(file TRV_load0_399_180_4.mat; x-var DEV3_SW1a(1)) DEV3_SW1a(2) (file TRV_load100_399_180_4.mat; x-var DEV3_SW1a(1)) DEV3_SW1a(2)
load, without capacitor 100% load, without capacitor
The effect of the load on the damping of the TRV is visible in the figures 14 and 15 for a
short-circuit power at the 110 kV-side of 6000 and 400 MVA respectively. The 100% load
has hardly any effect on the high-frequency component but gives overcritical damping to the
low frequency oscillations. For the 400 MVA-case, the boosting effect of the low-frequency
component in the TRV disappears, so that the peak value of the TRV decreases.
In comparison with the requirements in the Standards the steepness of the TRV (RRRV) is
rather high, as will be discussed in a further section . A possibility to reduce the RRRV is to
add a capacitor at one or both terminals of the circuit-breaker. For the 6000 MVA-case, the
impact of a capacitor of 100 nF at the side of the step-up transformer can be seen in the
figures 16 (0% load) and 17 (100% load). The higher frequency is indeed reduced and so is
the RRRV, but the overshoot of the TRV reaches up to a far higher peak value, hence
stressing the circuit-breaker during the dielectric phase of the TRV (in comparison to the high
RRRV-stresses during the thermal phase of the TRV).
29
90 90
*10 3 *10 3
75 75
60 60
45 45
30 30
15 15
0 0
0.0 0.4 0.8 1.2 1.6 *10 -3 2.0 0.0 0.4 0.8 1.2 1.6 *10 -3 2.0
(file TRV_load0_6000_180_C_4.mat; x-var DEV3_SW1a(1)) DEV3_SW1a(2) (file TRV_load100_6000_180_C_4.mat; x-var DEV3_SW1a(1)) DEV3_SW1a(2)
load, with capacitor 100% load, with capacitor
The simulations are based on a system voltage in the MV-network of 20 kV instead of 24 kV,
i.e. the rated voltage given in the IEC-Standards. In order to compare the results of the
simulations with values given in the Standards, the peak values of the TRV will further also
be expressed in pu.
In the next tables a number of typical characteristics of the simulated TRVs are elaborated to
be compared with the specified values in the Standards.
Table V
Angle=180
Ssc_net=6000MVA
Load 0% Load 25% Load 50% Load 75% Load100%
TRV 2 Parameter 2 Parameter 2 Parameter 2 Parameter 2 Parameter
U1(kV)
t1(s)
U1/t1(kV/s)
Uc (kV) 66.84 66.6 66.27 66.16 65.97
Uc (pu) 4.1 4.1 4.1 4.1 4.0
t2(s)
t3 (s) 50.8 50.7 50.6 50.5 50.5
Uc/t3 (kV/s) 1.32 1.31 1.31 1.31 1.31
td (s) 4.08 4.11 4.06 4.04 4.04
As was already clear from the figures, the influence of the load on the first part of the TRV is
nil, but from table V it can be learned that the peak value is rather high: 4.1 pu in comparison
to 3.13 pu as specified in the Standards. Also the RRRV (Uc/t3) is much higher than 0.35
kV/s given in the Standards. In fact the value for the RRRV shown in table V has even to be
raised by 17% in order to be comparable with a value based on 24 kV. On the other hand, as
stated in IEC 62271-100, test duty T30 is considered to cover the higher RRRV-stress that
occurs under out-of-phase conditions. The RRRV for T30 (i.e. a short-circuit test at 30% of
the rated short-circuit current, close to 25% being the rated value for out-of-phase duties) is
2.47 kV/s at 24 kV (class S1), much higher than the values occurring in the simulated cases
30
with a cable network. (For a 24 kV-network with OH-lines, class S2, RRRV is specifies as
0.71 kV/s and 3.01 kV/s for out-of-phase and T30 respectively.)
Note that the peak value of the TRV of T30 is only 2.61 pu, compared to for instance 3.13 pu for
out-of-phase in the Standards, so that the complete rise of the out-of-phase TRV is not covered.
Table VI
Angle=180
Ssc_net=400MVA
U1(kV) 57.54 57.33
t1(s) 50.71 50.61
U1/t1(kV/s) 1.13 1.13
Uc (kV) 67.19 59.69 57.08 56.95 56.78
Uc (pu) 4.1 3.7 3.5 3.5 3.5
t2(s) 426 429
t3 (s) 50.6 50.6 50.5
Uc/t3 (kV/s) 1.13 1.12 1.13
td (s) 4.07 4.06 4.04 4.1 4.03
At a lower short-circuit power in the 110 kV grid, both peak values and RRRVs decrease to a
certain extent (except of no load), but not below the specified values in the Standards; see
table VI.
The effect of a capacitor of 100 nF is given in table VII, where RRRV is less than half the
values shown in table I, but the peak values are far larger.
Table VII
Angle=180
Ssc_net=6000MVA/with Capacitance 100 nF
U1(kV)
t1(s)
U1/t1(kV/s)
Uc (kV) 81.32 80.11 79.04 78.4 77.77
Uc (pu) 5.0 4.9 4.8 4.8 4.8
t2(s)
t3 (s) 142 141 141 140 140
Uc/t3 (kV/s) 0.571 0.566 0.561 0.558 0.555
td (s) 18.4 18.3 18.3 18.2 18.2
Another important parameter is the out-of-phase angle, as can be seen in the next two tables;
table VIII for the case of 6000 MVA and table IX for the case of 400 MVA. The tables show
that the Standards cover, under the circumstances that have been simulated, an out-of-phase
angle up to 99, apart from the RRRV that is considered to be covered by T30 (two-part test).
Peak value and RRRV are proportional to sin() with as the out-of-phase angle [35].
31
Table VIII
Load 0%
Ssc_net=6000MVA
Angle=90 Angle=120 Angle=150 Angle=180
TRV 2 Parameter 2 Parameter 2 Parameter 2 Parameter
U1(kV)
t1(s)
U1/t1(kV/s)
Uc (kV) 47.26 57.88 64.56 66.84
Uc (pu) 2.9 3.5 4.0 4.1
t2(s)
t3 (s) 50.7 50.8 50.7 50.8
Uc/t3 (kV/s) 0.932 1.14 1.27 1.32
td (s) 4.07 4.11 4.07 4.08
Table IX
Load 0%
Ssc_net=400MVA
Angle=90 Angle=120 Angle=150 Angle=180
TRV 4 Parameter 4 Parameter 4 Parameter 4 Parameter
U1(kV) 40.68 49.83 55.58 57.54
t1(s) 50.77 50.75 50.7 50.71
U1/t1(kV/s) 0.801 0.982 1.1 1.13
Uc (kV) 47.51 58.19 64.9 67.19
Uc (pu) 2.9 3.6 4.0 4.1
t2(s) 426 426 426 426
t3 (s)
Uc/t3 (kV/s)
td (s) 4.09 4.09 4.07 4.07
Figures 18 and 19 show that both the RV and the out-of-phase current above 120, do not
differ much from those belonging to an out-of-phase angle of 180 (less than 15%). An angle
of 90 gives stresses that are 30% lower than those belonging to 180. In figure 18 the RV as a
product of the FPCF and the voltage factor is presented; FPCF is usually 1.5 for distribution
grids and the voltage factor is 2*sin(), so the RV is 3*sin(). In figure 19 the current
ratio Ioop/Isc is plotted versus the ratio of reactances X1 and X2, for various out-of-phase
angles. X1 can be regarded as the reactance of the generator and X2 as the reactance at the
other side of the circuit-breaker, i.e. the reactance of the network. Ioop is the out-of-phase
current, that depends on the out-of-phase angle and both reactances, and Isc is the short-circuit
current supplied by the generator in case of a busbar fault and thus depending on X1.
32
fpcf * voltage factor
3,5
3
2,5 fpcf=1.5
2 fpcf=1.3
1,5 fpcf=1.1
1 fpcf=0.9
0,5
0
0 50 100 150 200
out-of-phase angle (degrees)
Fig.18 Recovery voltage RV as product of FPCF and voltage factor
Out-of-phase current for several

angles
2
1,8
1,6 180
1,4
150
Ioop/Isc
1,2
120
1
90
0,8
0,6 60
0,4 30
0,2
0
0 5 10 15
X1/X2
Fig.19 Ioop/Isc as a function of X1/X2 and out-of-phase angle
As references for the simulations [37][38] are mentioned and the authors of [38] assisted
CIGRE WG A3.13 by preparing the simulations with the results shown above.
Further details on out-of-phase conditions are available in the Technical Brochure, Part II,
chapter 6.
33
5.2 Longitudinal dielectric withstand capability of switching equipment
Distribution circuit breakers have to perform new duties, such as system separation (out-of-
phase switching), synchronisation of generators and/or systems, synchro-check before
closing, etc. Such requirements are not so clear from the Standards, where for instance the
specification for a dielectric type test across open contacts is between 3.2 and 4.0 pu (power
frequency/1 min.), while a repetition in service is performed at lower stresses (e.g. 70% i.e.
2.2 to 2.8 pu).
According to IEEE Std. C37.013 [34] higher test voltages have to be applied (table 5 of Std.):
50 kV for rated voltages of 15.8 kV and below (i.e. 5.47 pu @ 15.8 kV), 60 kV for rated
voltages up to 27.5 kV (i.e. 3.78 pu @ 27.5 kV) and 80 kV for rated voltages up to 38 kV (i.e.
3.65 p.u @ 38 kV). The same test voltages are applied for general purpose circuit-breakers
with a rated voltage of 15.8, 27.5 and 38 kV resp. [31][32][33]: i.e. the 1 minute test voltage
across open contacts and to ground under dry conditions.
According to IEC 60694, table 1a, from a rated voltage of 12 kV up to 52 kV the 1 minute
power frequency test across open contacts of circuit-breakers (column (2)) varies from 4.0 to
3.2 pu respectively. Across open contacts of disconnectors (column (3)) it varies from 4.6 to
3.7 pu respectively. Therefore, in clause 4.2 of IEC 62271-100 [30] it is recommended to
require for circuit-breakers applied under conditions of system separation and synchronisation
the longitudinal withstand stresses of column (3).
With the recommendation, as stated in IEC 62271-100 and reduction for tests in service to
70%, the values given remain above 2.5 pu, being the longitudinal stress under conditions of
separated systems with an earth fault at one side. Special attention, however, should be given
to outdoor applications, where the same stresses have to be withstood under rain, and at least
2 pu should be withstood under polluted conditions.
5.3 Short-circuit currents
The impact of the increased number of generators tied to the distribution networks on the
short-circuit currents may include:
high values of dc component resulting from the high rate of decrease of the ac component
in SG
high initial peak value of the short-circuit current,
high peak value of the last current loop before interruption
long arcing time of the circuit-breaker for the major extended loop or very long arcing
time in the case of delayed current zeros.
These conditions have to be related to the short-circuit current supplied through the HV/MV
power transformer. The high X/R ratios of modern, low-loss power transformers also lead to
large DC components. Moreover, these power transformers still form the largest supplier of
the short-circuit current and the contribution of individual small generators can be neglected
in comparison to the contribution from the power transformers. The sum of the contributions
of a large number of small generators, however, may not be neglected. But, to the opinion of
WG A3.13, the increase of the overall DC-time constant is unlikely to exceed values specified
in the Standards [30].
34
For current transformer the situation may be different, as the transient DC-component of
short-circuit currents is considerably influencing the non-linear saturation effect of current
transformers. For pure current measuring purpose this effect is desirable to protect the high
sensitive measuring equipment. For protection purpose saturation is one of the most adverse
phenomena which may lead to malfunction of protection relays.
Figure 20 shows the magnetic flux of an iron core current transformer during a short-circuit
current with a transient DC-component. It can be seen that the magnetic flux reached by the
steady-state component is considerably lower than that of the transient component. If the
magnetic flux is exceeding the saturation flux of the iron core, the current transformer is
going to be saturated and the current transmission behaviour is strongly distorted. That is why
practically the transient component is only responsible for saturation effects.
Fig. 20 Saturation effect of current transformers due to the DC-component
Thus the current transformer dimensioning must be based on the DC-components which are
expected in the network. Up to now in distribution networks a network time N constant of
around 50 ms could be assumed. In DG networks this figure can be expected much higher.
The above equation gives an analytical basis for determining the over-dimensioning factor Ktf
considering DC components according to IEC 60044. S refers to the secondary time constant
of a current transformer which lies in the range of 1 to 5 seconds in case of iron core current
transformers. Ktf gives the factor of overdimensioning above the pure steady-state current
transformer design. This could lead to a current transformer which is several ten times bigger
than the steady-state design.
To keep the currrent transformer dimensioning within tolerable price and space limits, modern
relay technology requires only dedicated saturation free transmission time periods from the
beginning instant of the short-circuit. This strategy ensures a proper protection function, but
allows saturation in principle.
35
A different situation exists in case of a larger generator or set of generators that is directly
connected to the transmission network through a step-up transformer. In that case the
contribution of the single generator will be considerable and the phenomena listed above have
to be taken into account. This applies not only to the generator circuit-breakers, but to all
circuit-breakers in the direct vicinity of these generators. Here the application of the Standard
for generator circuit-breakers has to be considered [34]. An amendment to this Standard is in
the final stage of approval and will cover also cases with generators from 10 to 100 MVA. It
also considers generator circuit-breakers connected by means of shielded cables.
AG deliver a fast decreasing short-circuit current, that disappears within roughly 100 ms.
After clearing the fault, the recovery currents absorbed by AG (and motors) will be severe,
comparable with their contribution to the short-circuit currents. These recovery currents have
a strong influence on the recovery of the voltage and may even lead to voltage instability. The
recovery currents may last quite long, thus stressing the thermal capacity of the generators and
HV-equipment[39]. Another, even more stressful, condition is the sudden open-terminal with
reconnection, where currents larger than the short-circuit current will appear [39].
Fig. 21 Windmills in the Netherlands connected to a separate 20 kV-network
The short-circuit contribution of double-fed induction generators (DFIG) is very special, as it

is influenced by many parameters, such as the operational mode of the generator, the control
and protection setting of the crowbar, the control settings of the rotor-side converter and of
the stator-side converter, the settings of the protection, the pitch control of the vanes, the
36
control of the hydraulic brake, etc. In a number of papers [40][41][42]46] examples have
been given of short-circuit currents as calculated by means of simulations, sometimes
compared with test results [43][44]. The general trend is that the short-circuit currents are
much lower than when supplied by conventional AG or SG. Further the waveshape of the
currents varies very widely and is difficult to understand intuitively. For example, the second
peak can be higher than the first current peak [40] and the DC component can show a pattern
quite different from what will be expected from conventional machines [40][43][44]. This is
caused by the AC-component of the rotor current, whereas the AC-component of the (stator)
short-circuit current is related to the rotor DC-current.
Other transient phenomena related to DG are false switching operations in the vicinity of
dispersed generation; such as accidental system separation or generator disconnection, false
synchronisation of separate systems or generators (plus consequential tripping), network
restoration and auto-reclosing under adverse conditions [28]. Also power system instabilities
(angular, voltage, or frequency instability) could lead to severe transient phenomena and
consequential events. In this respect it should not be overlooked that, due to the trend to work
plant harder, it is expected that systems, including distribution-networks with DG, are
operated closer to the stability limits.
37
6 POWER ELECTRONIC EQUIPMENT
Power electronics normally offer a lot of flexibility, controlability and a smoothing effect in
terms of (over)currents and (over)voltages to the network. Power electronics for instance can
limit short-circuit currents and can supply or consume reactive power quite easily, even under
idle (unloaded) operation. Such properties lead to advantages for voltage control and system
restoration, but lead also to new requirements on the system protection and advanced
protection schemes.
The most notable negative impact is the generation of harmonics, but power electronics are
not the only source of harmonics. Harmonics have a negative effect on power quality.
Conversely, a low power quality has a negative effect on the operation and control of power
electronics, that normally will lead to separation from the network. Depending on the power
source or load behind the power electronics, after such a disturbance a fast restoration is also
possible. Apart from overcurrents and short-circuit currents, power electronics are also
vulnerable to over-voltages and transients in the system.
6.1 Harmonics
Distortion of voltages and currents in power systems is caused by power electronic

equipment, by non-linear load (as arc furnaces, fluorescent and ballst lamps, etc.) and other
non-linear phenomena like inrush-currents, over-excitation of transformers, corona of
transmission lines in bad weather conditions. Power electronic equipment is used at all
voltage levels and even when the harmonic generation of the individual equipment is low (for
instance low voltage applications), their large numbers in combination with a certain level of
resonance in the power system may lead to a high level of voltage distortion. This equipment
mainly consists of rectifiers and converters, which, at the distribution level, are found mostly
in industrial plants and public transportation systems.
A feature of DG, certainly those of a conventional technology, is the ability to absorb

harmonics. Usually this can be regarded as an advantage, unless the machines get thermally
overloaded. A problem faced by utilities that apply ripple control is that the ripple signal will
be absorbed as well by the numerous generators.
6.2 Filter banks
An inductance in series with a capacitor bank can be tuned for a certain harmonic frequency,
thus forming a low impedance path (short-circuit) for the harmonic currents of that frequency.
More advanced combinations of reactors, capacitors and resistors can be tuned to form a filter
with a certain bandwidth for more harmonic frequencies [50] or a low pass filter [51]. It is
necessary to study the harmonic impedance of the network involved under a variety of system
conditions in order to optimise the filter characteristics and the influence of shunt capacitor
banks in general.
The effect of connecting a filter-bank to a substation is that the harmonic distortion in the
voltage will to a large extent disappear, but at the same time, the filter-bank will force
harmonic currents to flow towards and through the filter-bank. Conversely, switching off
38
means that the distortion in the currents will be reduced but that the voltages will be disturbed
again.
Filter-banks and filter-bank switching are not covered by the Standards, as the circumstances
for their application vary too widely for harmonisation. Phenomena like filtering
characteristics, load currents, transient over-voltages, TRVs, etc. seem to be too varied for
effective standardization. Normally, dedicated system studies are performed to define the
system and component stresses under normal and abnormal conditions.
All equipment within a filter-bank are subjected to a high content of harmonic currents,
especially since filter-banks are intended to absorb the harmonics. All equipment has to be
designed for the thermal, electrical and dynamic stresses caused by the additional harmonic
currents. For instance, in ANSI C37.012 (2005), cl.4.7.1, it is recommended that circuit-
breakers be rated for a single or back-to-back capacitor-bank switching current that is 1.25 to
1.35 times the nominal capacitor-bank current. Also, for capacitor-banks it is recommended to
choose a nominal voltage that is greater than the operating voltage: for normal shunt
capacitor-banks (not filter-bank capacitors) a nominal voltage that is roughly 10% higher than
the operating voltage is recommended to cover the extra dielectric stresses due to harmonics.
6.3 Filter bank switching
Equipment used to switch the filter-bank currents are circuit-breakers, load-interrupters and
sometimes disconnectors. Circuit-breakers are able to make and break all currents up to their
rated short-circuit current, while switches are designed to break load and capacitive currents
only whilst having full fault making capability. Disconnectors have no specific rating for
switching load or capacitive currents and are only rated to withstand the short-circuit current.
However, since the Standards for switching equipment do not cover filter banks the
phenomena related to switching filter-banks will be compared with those related to switching
capacitor-banks.
With a high percentage of harmonic distortion, especially at the lower harmonics, the currents
may show more than two current zeros per power frequency cycle. In this way the breaking of
the current might take place earlier than at the regular power frequency current zero, but since
a filter-bank behaves mainly as a capacitor-bank, current interruption at an instant not
corresponding exactly to voltage maximum will lead to lower trapped charges (DC-voltages)
on the filter-bank (capacitor-bank), thus reducing the dielectric stresses at current interruption.
As stated before, energised filter-banks give distorted currents with almost sinusoidal
voltages, while disconnected filter-banks give low harmonic currents but distorted voltages.
Compared with switching off a capacitor-bank, the TRV for filter-bank switching will show
super-harmonic voltages on the normal TRV resulting in a higher peak value of the TRV.
Another phenomenon to be considered is that at switching off of a capacitor-bank the short-

circuit impedance causes a negative voltage jump (lower RV), but switching off a reactor in
series with the capacitor-bank (as in a filter-bank) causes a positive voltage jump and thus a
higher peak value of the RV.
To verify the breaking capacity of a circuit-breaker, the following statements can be accepted.
The harmonic content of the filter-bank current is not relevant for the breaking performance,
especially not for modern technologies, due to the short physical time constants in the arc. It
39
is expected that the initial part of the TRV at filter-bank switching is not critical for vacuum
technology circuit-breakers, as applied in distribution networks.
With respect to the peak-value of the TRV, the harmonic component in the RV, plus the other
phenomena (positive voltage jump, TOV, possible resonance) can be ascertained after proper
system studies and the applicable TRV envelope can be defined. The calculated peak value
can then be compared with the performance of the proposed circuit-breaker, the circuit-
breaker can be tested for the higher peak-value or a circuit-breaker with a higher rated voltage
can be chosen.
To summarise; filter-banks have to be switched by switchgear with a very low probability of

re-strike. Modern circuit-breakers with a very low probability of re-strike should face no
problems with the slightly higher RRRV for filter-banks in comparison to capacitor banks,
while the distortion of the current will be no problem at all. It is important to carefully select
circuit-breakers able to withstand the peak-value of the TRV, which is higher than the RV for
capacitor-banks. Furthermore circuit-breakers with a high level of mechanical endurance are
necessary. It is also recommended to apply controlled switching in order to limit inrush
transients from the filter-banks as much as possible. Controlled switching can also be used to
further reduce the probability of re-strikes at switching off.
6.4 Other aspects
All equipment involved has to withstand the thermal, dielectric and mechanical effects of the
distorted currents and related TOVs, especially the capacitor banks and reactors. In dealing
with the harmonic stresses for current transformers (CT) and voltage transformers (VT), the
application for measuring and protection purposes must be separately considered. The design
of CTs and VTs for protection purposes is not affected by harmonics, but the accuracy is. An
overloading of capacitive VTs with harmonics is possible in principle, but not realistic
because their fields of application are HV and EHV transmission networks where a large
harmonic content is not observed.
For measuring purposes, in connection with the control system of thyristor-controlled devices,
the accuracy and the frequency response has to be mentioned. In practice the frequency
response of conventional CTs and VTs is quite poor (figure 22 for inductive type CTs
andVTs) and special attention must be paid to their accuracy in respect of the absolute value
and phase shift [52].
50,00 20,00
0,00
0,00 -20,0010,00 100,00 1.000,00 10.000,00
10,00 100,00 1.000,00 10.000,00 -40,00
%
-50,00 -60,00
-80,00
-100,00 -100,00
-120,00
-150,00 -140,00
Hz Hz
1S 1S
Figure 22: Sample frequency responses of VT (left graph) an CT (right graph)
40
7 POSSIBLE IMPACT ON THE STANDARDS
Distribution circuit breakers have to perform new duties, such as system separation
(out-of-phase switching), synchronisation of generators and/or systems, synchro-check
before closing, etc. Taking into consideration the recommendation given in clause 4.2
of IEC 62271-100 [30], the longitudinal stresses in service are generally covered for
the condition with separated systems on both sides of the circuit-breaker. Special
attention however should be given to outdoor applications, where the same stresses
have to be withstood under wet, polluted and aged conditions.
Increase of power carrying capacity by taking advantage of dynamic (cycling) loading
has an increasing trend to be used for cables, OH-lines and transformers. The
switchgear, connected to these components, that has inherently a shorter thermal time
constant, should not limit the capacity of the whole branch in the distribution network.
Switchgear therefore has to be up-rated or a dynamic loading capacity has to be
defined. The topic of dynamic loading is addressed with CIGRE SC B3 and must be
considered in the Standards, either by overloading capacities or by certain
recommendations. Note that the present revision of IEC 60694 is taking dynamic
loading into account (subclause 8.2 of future IEC 62271-1).
Frequency and voltage deviations larger than what is covered by the Standards are
possible in case of islanding operation, putting the discussion whether type test results
for 50 Hz can be applied in a 60 Hz network and visa versa in another light. Note that
present Standards require T100 and T60 to be performed with both frequencies.
Out-of-phase switching is addressed in the Standards, but while the specified recovery
voltage covers an out-of-phase angle of 115, the TRV peak value seems to cover an
angle up to about 99. Light drives for conventional generators and/or a lack of inertial
energy in separated systems may cause out-of-phase angles to reach higher values than
90 in short times, thus forcing specifications to cover a wider range of out-of-phase
angles. In case of restrikes and failing circuit-breakers, the damage to equipment,
especially power generating plants, can be enormous. Control of separation instant in
order to avoid out-of-phase interruption with too large angles, as may be performed in
the HV-EHV transmission grids by out-of-step tripping relays, is not easily applicable
in the distribution networks [59].
RRRV as specified for the out-of-phase TRV is not steep enough, but IEC 62271-100
[30] mentions that this is covered by the RRRV of T30. This, however, is true up to an
out-of-phase angle of roughly 99, as for larger angles the first line of T30s TRV-
envelope has to reach higher.
41
8 CONCLUSIONS AND RECOMMENDATIONS
Distribution networks are changing rapidly, both in a technical sense and in a

organisational setting, because of developments with large implementation of
dispersed generation, more strict requirements for power quality and liberalisation (the
unbundling of transmission and distribution functions, the unbundling of production
and network functions, the unbundling of commercial activities and monopolistic
activities, the bundling of infrastructures for different commodities).
The consequences of these developments are quite clear in the secondary equipment:
more metering, tele-metering, more signalling, tele-control, fast and advanced
protection, extended telecommunication, advanced information available on-line to
system operators and off-line to asset managers, data mining techniques, data sharing
techniques, automatic system restoration functions.
The consequences of these developments for primary plant include: more redundancy
(even meshed systems), more surge arresters at vital points, some single phase systems
become three-phase, instrument transformers and protection become three-phase
systems, more advanced neutral treatment, switchgear and fuses become upgraded to
circuit-breakers, booster-transformers and other voltage control equipment are applied,
smart operating voltages are introduced, together with higher loading capabilities and
higher short-circuit powers.
Higher short-circuit currents, higher load currents, more complicated network
topologies lead to distribution grids that show a large similarity with transmission
grids. An adequate specification of the equipment involved is required. However, the
wider application of more advanced, upgraded and/or complicated equipment does not
mean intrinsically that the Standards for this equipment have to be adapted.
Conventional generators contribute to the short-circuit currents and power quality in a
well-known way. Care should be taken of the fact that DGs are small and dispersed,
meaning that the feeder itself will give less damping to the AC- and DC-components
in comparison to the experience with short-circuit currents supplied from the main
busbar (transformers).
Modern generators, equipped with full converters or DFIG (as applied in modern
windmills) show a limited contribution to the short-circuit current in comparison to the
conventional generators.
It is not expected that the behaviour of the DC-components of the short-circuit current
will lead to DC-time constants larger than those stated in the Standards. But, special
attention should be given to current transformers, as in some cases the DC-time
constant may be larger than specified.
Nowadays, in case of system disturbances, small generators are, to a large extent,
forced to separate from the distribution network, whilst medium size and large size
generators and windmill-parks, connected to the (sub)transmission grids are generally
required to ride-through faults and to contribute to system stability and system
restoration. Due to the massive increase in dispersed generation, there is a trend to
require small generators connected to distribution networks for fault-ride-though
capability as well as system support. Modern technologies are very suited to deliver
some of these ancillary functions, when properly specified [53].
The examples of the regulations as given in chapter 3, in combination with the
possibilities of modern technologies, justify the conclusions that international
harmonisation of the requirements of ancillary services for the todays popular
dispersed generators is possible. This harmonisation is also good for other involved
stakeholders such as operators and manufactures of windmills.
42
An innovative optimization for fault-ride-through rules, especially for conventional
dispersed generators is given in chapter 4 and presented as follows:
1. DG connected to faulty MV or LV distribution-feeders:
Disconnect SGs immediately after fault clearing (before the auto-reclosing
instant)
Disconnect AGs, if the short-circuit clearing time and dead time before
reclosure of the MV-line are too long and inertia of generators is low
Disconnect all DGs in case of permanent faults
2. DG connected to healthy MV or LV (radial) distribution-feeders in case of short
circuits in an other MV or LV feeder supplied by the same substation:
SGs and AGs should be kept in service if short circuit is cleared in the standard
low time (say, within 200 ms for multi-phase faults)
3. DG connected to healthy MV or LV (radial) lines in case of faults in the supply
HV network:
SGs and AGs should be kept in service in case of multi phase faults in the HV
network cleared in short time ( 200 ms) and in case of 1-phase-to-ground
faults in HV network even if of long duration (say, 0.5 1.0 s)
4. Islanding:
Disconnect automatically SGs under islanding conditions, unless the network,
protection and control is specifically designed for stable supply of an island
from DG
Disconnect AG generators under self-excitation conditions.
System conditions, like fault-ride-through, system restoration and/or islanding require
primary and secondary equipment to be able to cope with phenomena like detection of
system separation, frequency control, spinning reserve, voltage control, load rejection,
out-of-phase condition, synchronisation, synchro-check, longitudinal dielectric
stresses, but also safety issues like disconnected parts, earthed parts, automatic re-
closing, etc.
These system conditions lead to requirements, especially for switchgear, that are
related with smart network operation: advanced circuit-breakers provided with control
circuits, able to communicate on-line, to perform diagnostics, to take decisions, to
perform functions as auto-restoration and synchronisation, to act within the advanced
protection systems of the distribution grids, to perform controlled switching, etc.
Special requirements that may have an impact on the Standards have been addressed
in the former chapter, and are, for the moment, restricted to dynamic loading of
switchgear and out-of-phase switching by circuit-breakers.
The problems related to an increased content of harmonics was discussed in chapter 6,
where generally speaking the quality of electronic power converters reduces the
impact of individual apparatus, but this effect is compensated by the increase of the
application of power electronics. So far, no need for special requirements for the
conventional equipment is foreseen. When filter-banks are applied, its components of
course require special considerations as well as the switchgear that has to switch on
and off the filter-banks (see chapter 6).
43
REFERENCES
[1] Scoping Document by CIGRE WG A3.13

Changing Network Conditions and System Requirements
A3-04(SC)09 IWD, A3-06(WG13)59 IWD
[2] Electrical Environment of Transformers - Impact of Fast Transients
M. Glinkowski e.a.
CIGRE JWG A2/A3/B3.21
lectra No.218, Feb05, pp.24-37
[3] Changing Network Conditions and System Requirements
A.L.J. Janssen, e.a.
IEEE/CIGRE International Conference on Future Power Systems 2005, Amsterdam
[4] Contribution of Wind Farms to Ancillary Services
P. Bousseau, e.a.
CIGRE SC C6 Session 2006, Report C6-103
[5] IEEE Panel Session on the Impact of Dispersed and Renewable Generation on System
Structure..
IEEE 2006 General Meeting, Montreal
Contribution 5: Electricity Networks of the Future
Chr. Sasse
[6] Greenhouse gas emiision trends and projections in Europe 2004
European Environmental Agency
EEA Report No. 5/2004
[7] Communication as the Key for a Sustainable Network Integrating Dispersed and Renewable
Generation
B.M. Buchholz, e.a.
[8] CIGRE SC C1 Session 2006 General Report
CIGRE Sessions 2006 Proceedings
[9] CIGRE SC C1 Session 2004, answer to Question 2.5 by D.A. Baeza
CIGRE Sessions 2004 Proceedings
[10] Modeling New Forms of Generation and Storage
CIGRE Technical Brochure 185, April 2001
[11] Advanced Controls for Mitigation of Flicker Using Doubly-Fed Asynchronous Wind
Turbine-Generators
R.A. Walling, K. Clark, N.W. Miller, J.J. Sanchez-Gasca
CIRED 2005, Turin, No.672
[12] DFIG Control Performance under Fault Conditions for Offshore Wind Applications
G. Panell, D. Atkinson, R. Kemsley, L. Holdsworth, P. Taylor, O. Moja
[13] Making Connections
R. Zavadil, N. Miller, A. Ellis, E. Mujjadi
IEEE Power & Energy Magazine, Vol.3, Nr.6, Nov/Dec. 05, pp. 26-37
[14] Modeling of Gas Turbines and Steam Turbines in Combined Cycle Power Plants
CIGRE TF C4.02.25
CIGRE Technical Brochure 238, Dec. 2003
[15] Interaction between the Distributed Generation and the Distribution Network:
Operation, Control and Stability Aspects
V. Knyazkin, Th. Ackermann
CIRED, Barcelona 2003
[16] International Grid Code Developments for Wind Turbines and DC-connectors
G. di Marzio, O.B. Fosso
44
[17] Large Wind Power Integration in the Brazilian Power System and Revision of the National
Grid Code
F. C. Medeiros, e.a.
CIGRE SC C6 Session2006, Report C6-102
[18] Experience Integrating and Operating Wind Power in the Peninsular Spanish Power System
G. Gonzlez, e.a.
[19] CIGRE SC A3 Session 2006, answer to Question 3-15 of Special Report
L. Zehnder, M. Kriegel
CIGRE Session 2006 Proceedings
[20] Hydro-Quebec Grid Code for Wind Plant Interconnection
S. Bernard, M. Fecteau, D. Beaulieu, M. Lamothe
CIGRE Canada Conference on Power Systems 2006, Montreal, Paper no. 3
[21] Upgrade of Distribution-Grids to facilitate Wind-Energy Conversion
A.L.J. Janssen
EWEA/ETSO Conference 2006, Brussels, on Large Scale Integration of Wind energy in the
European Power Supply, B3 Infrastructures investments and financing
[22] Changing Network Conditions and System Requirements Part I: Distributed Generation
A.L.J. Janssen, e.a.
[23] Impact of Embedded generation on the French Power System
E. Bue, e.a.
[24] New Control Strategies to Prevent Blackouts:
Intentional Islanding Operation in Distribution Networks
B. Enacheanu, M. Fontela, Ch. Andrieu, Ha Pham, A. Martin, Y. Besanger
[25] CIGRE SC A3 Session 2006, General Report
CIGRE Session 2006 Proceedings
[26] Multi-agent application of substation protection coordination with distributed generators
H. Wan, K.K. Li, K.P. Wong
European Transactions on Electrical Power 2006; 16, pp. 495-506
[27] Integration of protection, control, monitoring, and remote communication into one intelligent
electronic device for small hydro-power plants pilot installations
G. Stranne, U. Nygren, S. Bod, M. Callvik, L. Wiktorsson, M. Lindberg
CIGRE SC C1 Symposium, Athens, April 2005, Rep. 110
[28]Distributed Generation in Relation to Phase Opposition and Short-circuits
A.L.J. Janssen e.a.
10th SCC Symposium 2002 in Poland
[29] Dynamic Modeling of Doubly Fed Induction generator Wind Turbines
J.B. Ekanayake, L. Holdsworth, XG. Wu, N. Jenkins
IEEE PS, Vol.18, No.2, May03, pp.803-809
[30] IEC Standard 62271-100, 2001 (Am.2, 2006)
Part 100: High-voltage Alternating-current Circuit-breakers
[31] IEEE Standard C37.04 (1999)
IEEE Standard Rating Structure for AC High-Voltage Circuit Breakers
Rated on a Symmetrical Current Basis
[32] ANSI Standard C37-06 (1997)
AC High-Voltage Circuit Breakers Rated on a Symmetrical Current Basis
Preferred Ratings and Related Required Capabilities
[33] IEEE Standard C37-09 (1999)
IEEE Standard Test Procedures for AC High-Voltage Circuit Breakers
45
[34] IEEE Standard C37.013 (1997)
Standard for AC High-Voltage Generator Circuit Breakers
[35] Dielectric, Switching and System Requirements under Out-of-Phase Conditions,
during Synchronisation and under Comparable Stresses
CIGRE SC 4 Symposium Zagreb, 2007, Report 0701
[36] Adjoint wind turbine modeling with ADAMS, Simulink and PSCAD/EMTDC
S. Uski, B. Lemstrm, J. Kiviluoma, S. Rissanen, P. Antikainen
Nordic Wind Power Conference, March 2004, Chalmers University of Technology
[37] Simulation and Analysis of Power System Transients Using EMTP-RV, Volume I:
Users Manual, 2003-2004.
[38] Transient Recovery Voltages During the Switching Under Out-of-Phase Conditions.
D. Braun, G. Koeppl
International Conference on Power Systems Transients, New Orleans, 2003
[39] Behavior of Induction Motor due to Voltage Sags and Short Interruptions
Juan C. Gomez, e.a.
IEEE-PD, Vol.17, No.2, Apr02, pp.434-440
[40] Fault Current Contribution of DFIG Wind Turbines
O. Anaya-Lara, N. Jenkins
3rd IEE Intern. Conf. RTDN, London 2005
[41] Transient Analysis of Grid-Connected Wind Turbines with DFIG
after an External Short-Circuit Fault
Tao Sun, Z. Chen, F. Blaabjerg
Nordic Wind Power Conference, 1-2 March 2004, Chalmers University of Technology
[42] Doubly Fed Induction Generator Model for Power System Simulation Tools
I. Nordheim, K. Uhlen, J.O. Tande, T. Toftevaag, M.P. Plsson
[43] Voltage dip ride through of a doubly-fed generator equipped with an active crowbar
J. Niiranen
[44] Simulation of Doubly Fed Induction Generator Wind Turbine with an Active Crowbar
J.K. Niiranen
EPE-PEMC, Riga 2004
[45] Modelling and Computer Simulation of Dispersed generation in Distribution Networks.
Measures to prevent disconnection during system disturbances
F.M. Gatta, F. Iliceto, S. Lauria, P. Masato
IEEE Power Tech Conference Proceedings, Bologna 2003, Vol. 3
[46] Dynamic Modeling of Doubly Fed Induction generator Wind Turbines
J.B. Ekanayake, L. Holdsworth, XG.Wu, N. Jenkins
IEEE PS, Vol.18, No.2, May 03, pp.803-809
[47] Distribution Systems and Dispersed Generation
A. Invernizzi, SC C6
lectra 213, Apr.04, pp. 17-21
[48] Impact of Wind Energy generation on the Safety of the Electrical Transmission Network
J.M. Rodriquez, e.a.
[49] Development of dispersed generation and consequences for power systems
CIGRE WG C6-01
Electra 215, Aug.2004, pp. 39-47
[50] Switching of Harmonic Loaded Capacitor Banks and Filter Circuits
K.H. Krger, W. Schultz, N. Trapp, G. Morav
CIGRE SC13 Session 1990, 13-104
46
[51] Impact of Reactive Power Compensation Equipment on the Harmonic Impedance of High
Voltage Networks
J.H.R. Enslin, e.a.
IEEE PES 2003 Power Tech Conf. Bologna, Jun. 23-26
[52] A New Method for the CCVT Performance Analysis using Field Measurements,
Signal Processing and EMTP Modeling
Lj. Kojovic, e.a.
IEEE PD, Vol.9, No.4, Oct.94, pp. 1907-1915
[53] GE Perspective on Grid Connection Requirements
S. Achilles
EWEA/ETSO Conference 2006, Brussels, on Large Scale Integration of Wind energy in the
European Power Supply, B2 - Integrating wind power into a single European electricity grid
[54] Changing Network Conditions and System Requirements.
Studies performed by CIGRE WG A3.13.
A.L.J. Janssen, F. Iliceto, Q. Bui-Van, S. Morais, B. Middleton, J. Jger
CIGRE SC A3 Colloquium 2007, Rio de Janeiro, Report PS2-01
[55] IEEE Standard 1547 (2003)
IEEE Standard for Interconnecting Distributed Resources with Eletric Power Systems
[56] IEEE Standard 1547.1 (2005)
IEEE Standard Conformance Test Procedures for Equipment Interconnecting Distributed
Resources with Electric Power Systems
[57] Connection Criteria at the Distribution Network for Distributed Generation
CIGRE TF C6.04.01
CIGRE Technical Brochure 313 (2007)
[58] Wind Power in the UCTE Interconnected System
UCTE NetWork of Experts on Wind Power, November 2004
[59] Application of Out-of-step Relaying for Small Generators in Distributed Generation
K. Malmedal, P.K. Sen, J.P. Nelson
IEEE Trans. Industry Applications, Vol. 41, No.6, Nov-Dec. 2005, pp. 1506-1514
47
Appendix A
CHANGING NETWORK CONDITIONS AND SYSTEM REQUIREMENTS

STUDIES PERFORMED BY CIGR WG A3.13
Report presented at the CIGR Sc A3 Colloquium 2007, Rio de Janeiro

Report PS2-01
48
Studies performed by CIGRE WG A3.13
A.L.J. Janssen F. Iliceto Q. Bui-Van S. Morais B. Middleton J. Jger
SUMMARY
This report is a summary of the two CIGRE Technical Brochures that, on behalf of CIGRE SC A3, will be
published in 2007 on Changing Network Conditions and System Requirements, with a focus on the
consequences for HV AC Equipment.
KEYWORDS
Distributed generators, shunt/series compensated lines, out-of-phase switching, synchronizing, TOV,
TRV, Ferranti-effect, transient overvoltages, SPAR, TPHSR, HWLL, MOSA, MOV, SMOSA
1. INTRODUCTION
As society has developed, the role of utilities has changed in response to commercial, regulatory and
environmental pressures. Many utilities with responsibilities for transmission and distribution networks
now have limited influence on the planning and operation of large power plants, which, historically, were
an integral part of network operation and, particularly, of voltage control. Large generators are typically
sited on the basis of primary energy source availability (fuel, hydro capability, etc), or of environmental
restrictions, and are consequently often far removed from load centers. The transmission of large
quantities of energy over long distances poses certain problems of its own and creates a need for the
greater use of local reactive power generation to control the network voltage profile.
It is also clear that, over time, installed capacity of distributed generation increase. Despite its local
nature it might also contribute to the difficulty of managing the networks. Shunt capacitor banks, Static
Voltage Regulators (SVRs) and FACTS are now being widely installed to supply reactive power for
voltage control in place of more conventional means (large generators, rotating compensators). Many of
these new devices generate harmonics and create the possibility of harmonic resonance due to capacitor
banks,of distorted voltages/currents from power electronics and other non-linear phenomena (magnetizing
currents), which must all be controlled. Often this leads to a need for additional harmonic filtering.
anton.janssen@continuon.nl
49
Interconnections between major networks are becoming increasingly important but also more difficult to
achieve due to the difficulties associated with obtaining new rights-of-way for the erection of overhead
lines. Consequently, existing overhead lines have to be up-rated in loading capacity whilst more shunt or
series compensated lines, and even shunt-compensated cables or HVDC connections, are applied.
Future power systems can be characterized by a mixture of distributed generation connected at medium
voltage and low voltage, and long distance transmission from remote large-scale power plants. Problems
such as reactive power control (shunt capacitor banks, shunt reactors, FACTS), long overhead lines
(compensation requirements), large short-circuit currents (high dc time constant, delayed current zeros,
fault current limiting) and severe TRVs (generator circuit-breakers, transformer fed faults, faults behind
reactors, out-of-phase switching) can all be foreseen and are considered in this document.
In addition to changes in the high voltage nature of networks, more advanced protection schemes,
intelligent interlocking and tele-tripping, controlled switching and the collection & exchange of
information are all influencing network development.
In summary, network conditions are changing due to the introduction of new components, new designs,
new applications and new ways of operation. The introduction of modern surge arresters and varistors,
non-conventional applications of switchgear, series and shunt capacitor banks, filter-banks, FACTS,
phase-shifters, advanced protection and control, windmill-farms, converter connected power plants are
creating more complicated transmission and distribution networks. These changes bring with them
changing stresses such as transient and temporary overvoltages, high harmonic levels, asymmetrical
voltages, and changes in the nature of short-circuit currents.
In order to structure these versatile changes, the main directions of the changes have to be found, taking
into consideration the possible impact on the specifications for HV equipment. From the point of view of
CIGRE SC A3 High Voltage Equipment, the most relevant network changes are on one side more
complex grids due to distributed generation and the introduction of advanced protection and control
systems, and on the other side the greater utilization of long distance interconnections with its reactive
power and voltage control. These are considered against a background of increased utilization of
equipment in terms of age, loading & voltage stresses, complexity and reduction in equipment size,
incorporation of more intelligence and the application of more overvoltage protection and smart devices.
In 2002 CIGRE SC A3 has established a new WG A3.13 Changing Network Conditions and System
Requirements to investigate these developments and its consequences for HV equipment. In 2003 WG
A3.13 started its studies and delivered in 2004 to SC A3 an internal working document, called the Scoping
Document [1]. In the same year it has been proposed and accepted by SC A3 to split the work into two
WGs and WG A3.19 has been established to study the effects of three-phase SLF and LLF on the TRV
requirements. (LLF is a line fault at a long distance thus introducing high TRV peak values at the line side
[6]). In 2007 WG A3.13 has finalized its studies and two CIGRE Technical Brochures will be published;
one on the impact of distributed generation and one on the impact of long distance transmission.
A summary of the findings and conclusions of WG A3.13s studies will be presented here. [2][3][4][5]
50
ASYNCHRONOUS
GENERATOR
GEARING CIRCUIT-
BREAKER
ASG
3~
TRANSFORMER
MV-GRID
REACTOR
CONVERTER
DC AC
ROTOR AC DC
CROWBAR
Fig. 1 DFIG: double fed induction generator with variable speed
2. DISTRIBUTED GENERATION
The large increase in dispersed generation and the trend to an even further growth give enough evidence
that the effects and impact of distributed generators on the systems behavior cannot longer be neglected.
The developments lead to special requirements and special phenomena such as higher short-circuit power
requirements, higher short-circuit currents, other patterns of power flows and short-circuit power flows,
different voltage profiles and voltage variations, fast and advanced protection, complicated controls,
situations of potential islanding, synchronisation, out-of-step switching, phase opposition currents,
harmonics, safety issues, fault location, black-start conditions, system restoration, etc.
Many small power units are protected in such a way that for any disturbance in the network, they are
immediately switched off by a circuit-breaker at the low voltage side or at the high voltage side of the
step-up transformer (after roughly 50 to 100 ms), thus reinstating the passive feature of the distribution
network. In this way they enable the conventional selective protection, auto-reclosing and fault location in
the distribution grids and prevent out-of-phase conditions, unintended island operation, safety problems
and severe damage to the power plants. However, when distributed generation has expanded to a
significant share, requirements have to be put forward that these generating plants contribute to the
stability of the network and remain connected for at least a certain period of time so that voltage sags can
be limited in amplitude and duration, overloads are reduced and regional power deficits are avoided [7].
The ride through time is specified as several hundreds of ms for smaller or special units up to more than
1 s for large conventional power units.
Measures to profit from DG and to minimize its drawbacks for the network are with a clever optimization
of protection settings. In this respect, reference is made to [8] where the following recommendations are
given:
DG connected to faulty MV or LV distribution-feeders
Disconnect synchronous generators immediately after fault clearing (before the auto-reclosing instant)
Disconnect asynchronous generators, if the short-circuit clearing time and dead time before reclsoure of
the MV-line are too long and the inertia of the generators is low
Disconnect DG in case of permanent faults
DG connected to healthy MV or LV (radial) distribution-feeders in case of short-circuit in another MV or LV

feeder supplied by the same substation
Synchronous and asynchronous generators should be kept in service if short-circuit is cleared in the
standard low time (say, within 200 ms for multiphase faults)
Protect synchronous generators against loss of synchronism
Protect asynchronous generators against sustained overspeed and overcurrent
51
DG connected to healthy MV or LV (radial) lines in case of faults in supply HV network
Synchronous and asynchronous generators should be kept in service in case of multiphase faults in the HV
network cleared in short time ( 200 ms) and in case of single phase to ground faults in HV network
even when of long duration (say, 0.5 1.0 s)
Protection synchronous generators against loss of synchronism
Protection asynchronous generators against sustained overspeed and overcurrent
Islanding
Disconnect automatically synchronous generators under islanding conditions, unless the network, protection
and control is specifically designed for stable supply of an island from DG
Disconnect asynchronous generators under self-excitation conditions.
However these measures are only feasible when short protection clearing (100 ms to 200 ms) and auto-
reclosing times can be reached, which may not be the case in distribution-networks [2]. But, modern
technology wind mills, as DFIGs and generators connected through full converters, should inherently be
able to fully ride through fault conditions, even for longer clearing times, and resume power supply within
a few cycles after fault clearing. From a technical point of view, modern design of windmills offer the
possibility to employ their inertial energy, freewheeling independent from the power frequency (figure 1),
to over-come to a large extent the difficulties during voltage and frequency dips. But also many co-
generation plants are capable to contribute to the overall system performance for a longer period of time
than corresponding to the old policy of immediate tripping.
Some examples of fault ride through requirements as a function of duration of the voltage dip and the
residual voltage are given in figure 2: England and Wales (UK), Hydro Quebc (Canada) and Germany.
Supergrid Voltage Level

(% of Nominal)
100
90
85
80
75
UK HQ GER
25
15
0 1 1.2 2 2.5 3 Time (s)

0.14 0.15
Fig. 2 Examples of fault-ride through requirements
The examples of the regulations, in combination with the possibilities of modern technologies, justify the
conclusion that international harmonisation of the requirements of ancillary services for most of the
nowadays popular dispersed generators is possible, but will also be a good development for involved
stakeholders.
52
Because of developments as the large penetration of dispersed generation, more strict requirements for
power quality and liberalisation, distribution networks are changing rapidly, both in a technical sense as
in a organisational setting. The consequences of these developments are quite clear in the secondary
equipment: more metering, tele-metering, more signalling, tele-control, fast and advanced protection,
extended telecommunication, advanced information available on-line to system operators and off-line to
asset managers, data mining techniques, data sharing techniques, automatic system restoration functions.
Looking at the primary plant, more redundancy is applied, even meshed distribution systems, more surge
arresters at vital points, some single phase systems become three-phase, some instrument transformers and
protection become three-phase systems, more advanced network neutral treatment, switchgear and fuses
become upgraded to circuit-breakers, booster-transformers and other voltage control equipment are
applied, smart operating voltages are introduced, together with higher loading capabilities and higher
short-circuit powers.
Modern generators, equipped with full converters or DFIG, show a limited contribution to the short-circuit
current in comparison to the conventional generators. It is not expected that the behaviour of the DC-
components of the short-circuit current will lead to DC-time constants larger than those stated in the
Standards [10][11][12][13].
System conditions, like fault-ride-through, system restoration and/or islanding require primary and
secondary equipment to be able to cope with phenomena like detection of system separation, frequency
control, spinning reserve, voltage control, load rejection, out-of-phase condition, synchronisation,
synchro-check, longitudinal dielectric stresses, but also safety issues like disconnected parts, earthed parts,
automatic re-closing, blocking, etc. These system conditions lead to requirements, especially for
switchgear, that are related with smart network operation: advanced circuit-breakers, able to communicate
on-line, to perform diagnostics, to take decisions, to perform functions as auto-restoration and
synchronisation, to act within the advanced protection systems of the distribution grids, to perform
controlled switching, to assist fault location, etc.
Special requirements that may have an impact on the Standards are, for the moment, restricted to dynamic
(cyclic) loading of switchgear and out-of-phase switching by circuit-breakers. Out-of-phase switching is
addressed in the Standards, but is covering an out-of-phase angle up to about 90. Light drives for
conventional generators and/or a lack of inertial energy in separated systems may cause out-of-phase
angles to reach higher values than 90 in short times, thus enforcing specifications to cover a wider range
of out-of-phase angles. In case of re-strikes and failing circuit-breakers, the damage to equipment,
especially power generating plants, can be enormous. RRRV as specified for the out-of-phase TRV is by
far not steep enough, but IEC 62271-100 [10] mentions that this is covered by the RRRV of T30. This,
however, is true up to an out-of-phase angle of roughly 90, as for larger angles the first line of T30s
TRV-envelope is not long enough. Larger out-of-phase angles (180) are specified by many users and
manufacturers are willing to design, test and supply adequate generator circuit-breakers.
Recent development in the Standards show more attention to steep and high TRVs for MV-networks
(amendment 2 to IEC 62271-100) and to generator circuit-breakers for small power plants, 10 to 100 MW
(IEEE C37.013 [14]). Users should be aware of these improvements, taking in mind the aforementioned
short-comings.
3. LONG DISTANCE TRANSMISSION

The privatisation and liberalization of the electric energy sector has considerably increased interregional
and international energy transactions and created a greater need for large power transfers over long
distances. More use and reliance on transmission has also been caused by the growing environmental
53
restrictions for the construction of new power plants and transmission lines. In several cases the
independent power producers (IPPs) have located their new thermoelectric power plants, to be fired with
imported fossil fuels (gas, coal, fuel oil), in places remote from the loads to be served, or far from optimal
locations as regards transmission. In a few countries, the construction of large windmill farms has also
required transmission along relatively long distance and the need to provide new transmission and
generation reserve capacities in order to face the volatile patterns of generation inherent to operation of
large windmill farms. Congestion of infrastructures and strong demand of environment conservation may
justify the use of long stretches of EHVAC cross-linked polyethylene insulated underground cables, even
in non-urban areas, requiring a large shunt compensation. These cable stretches may be solidly connected
to overhead lines, to form mixed cable-overhead lines, whose design and operation have special technical
features.
Several methods to increase the transmission capacity are used: utilize OH-lines, cables and transformers
up to their dynamic (cyclic) loading capacity, increase the operating voltage, apply shunt and/or series
compensation, apply phase shifting transformers, apply FACTS devices and HVDC systems, apply HSIL-
lines (high surge impedance loading lines) and EXB-lines (expanded bundle lines). Not yet applied, but an
economically promising technology is very long distance transmission by HWLL (half wave length lines).
The continuous operation voltage should remain within the maximum values stipulated by the IEC
Standards, but in case these values are exceeded in order to utilize the benefits of higher operating
voltages [9], precautions should be adopted in the specification of equipment and, in existing systems,
some equipment may have to be replaced/up-rated in order to avoid certain inconveniences or an increase
of risk of failure has to be accepted.
Temporary overvoltages (TOVs) in large transmission systems are mainly caused by the Ferranti-effect
(load rejection), ferro-resonance (single pole auto-reclosure in shunt compensated lines), low-order
harmonic resonance (long EHV cable length in combination with OH-lines). Countermeasures are: tele-
tripping, automatic connection of shunt reactors, the application of (switch-able) surge arresters and
circuit-breakers specified to handle the long unloaded line switching and out-of-phase switching
conditions.
The transient overvoltages specific for large transmission systems are the slow front transient overvoltages
(switching overvoltages), occurring at energizing OH-line and single pole (SPAR) or three-pole re-closure
(TPHSR) of OH-lines. Depending on the switching impulse withstand level (SIWL) of the OH-lines
(typically 3.0 pu in Europe with withstand probability of 99.86% (V50% - 3) and 2.5 pu in North-
America) countermeasures (closing resistors, controlled switching, surge arresters, staggered poles, pre-
connected shunt reactors) may be necessary to withstand the switching surges, generally smaller than 3.0
pu. But TPHSR requires special attention, ad the switching surges may be larger, dependent on the system
lay-out.
Secondary arc current during SPAR is mainly a problem of arc extinction on the line, and not a problem
for the line circuit-breakers involved. Shunt reactors can only effectively reduce the arc current to a value
below 40 A, if a properly tuned neutral reactor is added between the neutral of the shunt reactors and
earth. But as, certainly at the highest operating voltages, it may be difficult to find a value for the neutral
reactor, suitable for all prevailing situations, it is not always possible to limit the arc current to acceptably
low values. On the other hand, shunt reactors and associated neutral reactors may have to be out of service
when the lines are heavily loaded, i.e. when the SPAR is more useful. The re-closing time may also be
critical and therefore an alternative solution to limit the arcing current is the use of High Speed Grounding
Switches (HSGSs).
Electrostatic and electromagnetic induced currents are larger for long OH-lines than for medium or small
lengths. This affects the specification of earthing switches in particular of multiple-circuit lines, as well as
secondary arc phenomena. Measures to reduce the secondary arc current to values that force self-
54
extinction are mentioned in the former paragraph, but for some long compact lines even these
countermeasures cannot prevent the necessity to resort to TPHSR.
Switching off of unloaded long lines differs from clearing the capacitive currents on small or medium
length lines, as the capacitive currents will be higher (certainly under circumstances of disturbances), the
induced voltages will be higher, the Ferranti effect more dominant and switching under high TOV-
conditions more probable. On the other hand, usually shunt compensation may be applied and possibly
also series compensation, both leading to less severe recovery stresses when switching off the long OH-
line. Nevertheless utilities with long OH-lines specify voltage factors higher than those given in the IEC-
Standards for capacitive current switching.
Clearing of fault currents flowing through series capacitor banks result in high TRVs, due to the electric
charge on the series capacitors. TRV peak values can be limited by switched opening resistors, by fast by-
passing the capacitor bank, by MOSAs connected phase-to-earth or MOV in parallel to the arcing
chambers of the CBs. The first two solutions bring along problems (reliability, retardation, selectivity)
discouraging application in modern transmission systems. Use of special MOSAs with very low SIPL
can limit TRV-peak value within 3.2pu, whereas MOVs in parallel with arcing chambers of CBs can limit
TRV-peak value within 3pu or 2.5pu, depending on whether there is or not the requirement of system
synchronization with the involved CBs. [15][16]
Fig. 3 Example of a 735 kV Series Capacitor Bank
Full phase-opposition switching of a generating plant at the HV-side leads to TRV peak values of 2.9 to
3.2 pu; and higher values for systems where an earth fault occurs or a temporary overvoltage. For full
phase opposition switching of OH-lines, calculations show peak values of the TRV from 2.7 pu (420
kV/63 kA, 100 km length, Ioop = 15%) to 3.1 pu (420 kV/40 kA, 100 km length, Ioop = 25%) and even
beyond for longer line lengths). Opposite to the reality in the systems, the RRRV specified in the
Standards for out-of-phase switching is lower than the RRRV specified for test duty T100. But, to the
Standards, the RRRV for out-of-phase switching is considered to be covered by test duty T30 (multipart
testing), what is not completely true for large out-of-phase angles. Despite the fact that in many cases the
out-of-phase angle will be random, the Standards are based on an angle that is limited to about 90 [17].
55
Special attention is asked for synchronizing, where the power frequency withstand voltage across open
contacts (under rain in type test) of circuit-breakers should have the values specified in the Standards for
the disconnectors. In that case, the external insulation across open contacts of live tank circuit-breakers
should be specified to withstand two times the phase-to-earth maximum power-frequency operation
voltage with pollution representative of the expected local environmental conditions. Line circuit-breakers
used for synchronizing of separated sub-systems (or used for the automatic line re-closure) are exposed in
open position to lightning overvoltages which may exceed their withstand capacity. Protection can be
warranted by MOSAs connected phase-to-earth at the line end, or by specially shaped air spark-gaps.
In an appendix of the Technical Brochure on Long Distance Transmission the service experience of
utilities from several parts of the world with respect to series compensation is given.
4. CONCLUSION
The work of WG A3.13 has revealed several system changes due to new technical, economical and
political circumstances influencing the power systems. New amount of stresses and requirements on the
power system devices have been found consequently. This phenomena are widely know and not new by
themselves, but up to now only treated as theoretical, because the network conditions which lead to such
extreme were not realistic. Now these effects can not be neglected anymore because these are more and
more becoming common in future power systems.
The content of both Technical Brochures is given in Annex A. Annex B shows an overview of the
temporary and slow front overvoltages (TOVs and SFOs) as discussed in Technical Brochure on Long
Distance Transmission.
REFERENCES
[1] Scoping Document by CIGRE WG A3.13: Changing Network Conditions and System Requirements
A3-04(SC)09 IWD, A3-06(WG13)59 IWD
[2] Changing Network Conditions and System Requirements Part I: Distributed Generation
[3] Changing Network Conditions and System Requirements
IEEE/CIGRE International Conference on Future Power Systems 2005, Amsterdam
[4] Long Distance AC Power Transmission and Shunt/Series Compensation Overview and Experiences
CIGRE SC A3 Session 2006, Rep. A3-206
[5] Dielectric, Switching and System Requirements under Out-of-Phase Conditions, Synchronisation and under
Comparable Stresses
CIGRE SC 4 Symposium Zagreb, 2007, Report 0701
[6] Severe Duties on High-Voltage Circuit-Breakers Observed in Recent Power Systems
H. Hamada, Y. Kasahara, T. Shimato, K. Hirasawa, K. Suzuki, T. Yoshizumi
CIGRE SC 13 Session 2002, Report 13-103
[7] Impact of Wind Energy generation on the Safety of the Electrical Transmission Network
J.M. Rodriquez, e.a.
[8] Modelling and Computer Simulation of Dispersed generation in Distribution Networks.
Measures to prevent disconnection during system disturbances
F.M. Gatta, F. Iliceto, S. Lauria, P. Masato
IEEE Power Tech Conference Proceedings, Bologna 2003, Vol. 3
[9] Severe stresses on Switching Equipment of 500 kV Transmission System in Japan
T. Shirato, K. Yokotsu, H. Yonezawa, J. Kida, T. Yokota, T. Sugiyama
CIGRE Session 2006, Rep. A3-303
[10] IEC Standard 62271-100, 2001 (Am.2, 2006)
Part 100: High-voltage Alternating-current Circuit-breakers
56
[11] IEEE Standard C37.04 (1999)
IEEE Standard Rating Structure for AC High-Voltage Circuit Breakers
[12] ANSI Standard C37-06 (1997)
AC High-Voltage Circuit Breakers Rated on a Symmetrical Current Basis
Preferred Ratings and Related Required Capabilities
[13] IEEE Standard C37-09 (1999)
IEEE Standard Test Procedures for AC High-Voltage Circuit Breakers - Rated on a Symmetrical Current
Basis
[14] IEEE Standard C37.013 (1997)
Standard for AC High-Voltage Generator Circuit Breakers - Rated on a Symmetrical Current Basis
[15] TRVs across circuit breakers of series compensated lines. Analysis and operation experience in the 420kV
Turkish grid
F.M. Gatta, F. Iliceto, S. Lauria, B. Dilli
[16] Application of 500kV circuit breakers on transmission lines with MOV protected series capacitor bank
B.L. Avent, D.F. Peelo, J.H. Sawada
[17] CIGRE Technical Brochure 305 (2006)
Guide for the application of IEC 62271-100 and IEC 60694, Part 2
ANNEX A: content of the Technical Brochures on Changing Network Conditions and System Requirements
Apart from general chapters (Membership, Acknowledgement, Preamble, Content, Acronyms and Abbreviations,
Introduction, Conclusions and recommendations, References, Appendices) the two Technical Brochures contain the
following chapters:
Technical Brochure Part I on the impact of distributed generation on equipment rated above 1 kV
Technology of distributed generators
(windmills, co-generation)
Regulations and policy with respect to distributed generation
Network topologies related to distributed generation
Transient phenomena
(out-of-phase conditions, longitudinal dielectric withstand capability)
Power electronic equipment
(harmonics, filter-banks)
Possible impact on the Standards
Technical Brochure Part II on the impact of long distance transmission on HV equipment

Continuous operation at voltages exceeding standard recommended values
Temporary overvoltages (TOV)
(load rejection, resonance, ferroresonance, HWLL)
Transient overvoltages
(OH-line energization, reclosure, fault clearing, MOSA)
Transient recovery voltages
(clearing faults on series compensated OH-lines, capacitive currents, under high TOV)
Out-of-phase switching
Behaviour of CBs during synchronisation and reclosure
Transient currents
(induced currents, secondary arc currents, DC offset currents)
Series capacitor bank technologies and operating experiences
Flexible AC Transmission Systems (FACTS)
(power electronic equipment, harmonics, filter banks, phase-shifting transformers, variable shunt reactors)
Possible impact on the Standards
As appendices the papers presented by CIGR WG A3.13 are attached to the Technical Brochures.
57
Annex B: overview of temporary and slow front overvoltages (TOVs and SFOs) as discussed in
the Technical Brochure Part II on Long Distance Transmission
TOV
Section phenomenon pu measure remark
3.1 Load rejection + TOV 1.9-2.0 MOSA Hydro Quebec
1.6 SMOSA Hydro Quebec
2.0 MOSA Vietnam
1.7 MOSA Chili
1.4 Unloaded lines tele- Turkey
tripping, automatic
insertion of shunt reactors
3.2 Ferroresonance 1.4-3.5 mitigation possible
3.3 Inrush/magnetizing current 2.0 Build-up resonance
2.5 + TOV
3.4 Cable/OH-line 2.0 resonance
3.5 HWLL 2.7-3.0 3 ph. short-circuit at
close to line ends
2.0-2.2 1 ph. short-circuit at
close to line ends
3.6 SPAR + ferroresonance 1.6 Low onset corona
voltage
2.2 High onset corona
voltage
SFO
Section phenomenon pu measure remark
4.1 Line energization < 3.0
2.0-2.2 PIR
1.5-1.6 Two-step PIR
< 2.0 Controlled sw.
1.7-2.2 MOSAs with low SIPL
< 2.4 Staggered poles
Re-closing < 2.4 SPAR
2.7-3.0 SPAR Unusual
configurations
3.8-4.0 TPHSR
< 3.0 TPHSR + SR or inductive
PTs
1.7-2.2 TPHSR + MOSAs
< 2.0 TPHSR + controlled sw. Without trapped
charge on line
< 3.0 TPHSR + controlled sw. With trapped charge
on line
< 2.5 TPHSR + PIR
4.2 Shunt cap.bank closing 1.6
Transformer closing 2.0-2.5
4.3 Fault clearing 1.8-2.1 1 ph. fault
3.1-3.3 2/3 ph. fault
58
Appendix B

PART I: DISTRIBUTED GENERATION
Report presented at the CIGR SC A3&B3 Joint Colloquium 2005, Tokyo

Report 103
59
PART I: DISTRIBUTED GENERATION
A.L.J.Janssen M.Waldron J.Jger Q.Bui-Van

M.Glinkowski M.Claessens F.Gallon H.Kawano
CIGRE WG A3.13
Keywords
Distributed generation, DFIG, out-of-step, generator circuit-breakers, out-of-phase TRV
Abstract
Attention is given to the impact of increased distributed generation on system requirements as put
forward to HV-equipment: bi-directional power flows, unusual patterns of short-circuit currents,
probability of out-of-step conditions, consequences of out-of-phase (especially TRVs), false
synchronisation, islanding, requirements for synchronisation, harmonics. The way of connection and
the behaviour of distributed generation plants is divided in several categories, like with or without
converters, synchronous, induction or doubly fed induction generators (DFIG). A comparison is made
with DG connected to transmission grids.
1. Introduction
Networks are changing due to business drivers such as environmental concerns (including concerns as
running out of fossil energy sources), competitive power market, further utilisation of transmission
corridors, multi-directional power-flows in distribution networks, increased capacity, increased
efficiency, etc. These developments lead to technology changes (e.g. distributed generation, wind-
farms, compensated lines, phase shifters, filter-banks, non-linear loads, HVDC, FACTS, advanced
protection and control systems) and consequentially to special requirements, for instance with respect
to harmonics, temporary overvoltages (TOV), transient recovery voltages (TRV), out-of-phase
conditions, power quality, etc.
CIGRE WG A3.13 Changing Network Conditions and System Requirements is giving special
attention to the consequences of the growth in distributed generation (co-generation plants as well as
sustainable power generation) and to the consequences of long distance transmission (a larger distance
between power generation and power distribution, leading to voltage problems and the need for
reactive power compensation). In both cases the interaction between protection and control systems on
one hand and the network dynamics on the other hand will play a dominant role in the severity and
probability of the phenomena that have to be withstood by the applied HV-equipment. These
phenomena have to be considered against a background of increased utilisation of equipment (in terms
of age, loading and voltage stresses), of reduction of size and complexity, of the incorporation of more
anton.janssen@nuon.com
60
intelligence and the application of more over-voltage protection and smart devices. Methods of
condition and utilisation assessment of power system components with regard to asset management
become progressively applicable. [1]
As most utilities face the increase of distributed generation, WG A3.13 has chosen to devote this paper
to the impact of power plants in distribution networks on the specifications for HV-equipment. In
future publications the impact of long distance transmission and other developments will be treated.
2. Description of power plants connected to distribution networks

In line with the scope of CIGRE SC A3 High-Voltage Equipment, WG A3.13 is giving special
attention to the consequences of distributed generators on the equipment in the networks. With that in
mind the distributed generators can be divided into two groups: (a) generators directly coupled to the
grid, possibly through a step-up transformer, and (b) generators that are coupled by means of a full
converter, probably through a step-up transformer.
To the last group (b) belong power plants such as micro-turbines, fuel-cells, photo-voltaic systems,
energy storage systems as SMES (super-conducting magnetic energy storage), BESS (battery energy
storage system) and flywheels, certain types of windmills. Due to the application of a converter, these
small power plants are able to limit their contribution to fault currents, to support reactive power
balance and voltage control, to restore systems after disturbances, to adapt to frequency variations,
phase angle variations and voltage variations. Thus converters have an advantageous influence on the
system behaviour and therefore on the requirements for other components. One main disadvantage is
the generation of harmonics, although nowadays converter technologies and mitigation technologies
show substantial improvements [2]. Nevertheless one has to be aware of the possibility of harmonic
resonance problems and that low and high voltage equipment may have to be specified for the higher
content of harmonics in applied voltages and/or currents. Also the protection systems have to work
more and more on frequency discriminating basis. But, it seems, to the opinion of CIGRE WG A3.13,
not to be necessary to investigate such technologies more in detail due to their softening impact on
transients in the system.
The studies of WG A3.13 are therefore focused on (a): power plants with a conventional lay-out and
flexible generators driven by reciprocating machines (gas-motors, diesel-engines), by gas turbines
(single shaft gas turbines derived from heavy duty types or multi-shaft gas turbines derived from the
aircraft, i.e. aero-derivatives), by steam turbines (incineration of refuse, of biomass, of bio-gas, and
solar thermal/geo-thermal installations), by small hydro plants (including pump storage), by air-
pressure turbines (compressed air storage installations) and by windmills. Furthermore, combined
cycle plants with gas turbine, steam turbine and generator on one shaft or gas turbine and steam
turbine on separate shafts with each its own generator. As most of these technologies are used for co-
generation purposes, practically all are more or less regarded as contributing to environmental
sustainability. The generators can be synchronous or asynchronous machines, sometimes equipped
with power electronic devices (the so-called double-fed generators) to control the rotor current
(voltage control, slip, frequency for variable speed, reactive power output).
For the transient phenomena in networks (occurrence of a short-circuit, clearing of a short-circuit, mis-
synchronisation, system separation, islanding) the fuel control systems of the conventional plants are
so slow that they can be regarded as fixed. The dynamics of the prime mover are relevant during the
first second for the very fast control systems and during the first minutes for the slower systems (for
instance: simple hydro plants, simple steam turbines, steam turbines in combined cycle plants). [3][4]
The inertia of small machines used in those systems can vary and the rotor speed can change relatively
fast (within a few hundreds of ms) to reach out-of-phase conditions in case of short circuits. A rough
indication of the critical clearing time is (H/50) s (derived from [5]). Some examples of inertia
constants H of gas-turbine driven power plants are: 7 to 9 s for heavy duty gas-turbines (hundreds of
MW), 5 to 7 s for smaller units, down to 2 to 4 s for small units (tens of MW), and 1.3 to 2.2 s for
aero-derivative gas-turbines [4]. Windmills with induction generators give an H of 2.5 [6] to 3.5s [7].
61
From the excitation point of view (magnetic fields of the machines and the magnetic field control)
there are a few time windows of interest. In the very short time frame (within the first half cycle) no
effect of the excitation is noted. In the second time frame (from half cycle to few hundreds of ms) the
dynamics of the excitation systems have to be considered for modern generators as used in DG. In
many instances the excitation systems can be set to behave completely differently thus providing high
degree of behavioural variability. In the third (long) time frame of more than a few hundreds of ms
other parts of the DG controls play a role. [7][8][9][10]
3. Power flow and short-circuit currents

Distributed generation in distribution, sub-transmission, and industrial networks leads to a structural
change of the power flows, as the generated energy is normally independent from the local energy
demand. The unidirectional, mainly radial, power flow from the transmission grid through the
distribution network to the load will change to a bi-directional, mainly meshed, complicated power
exchange system (see Fig. 1).
G G
G
G
G G
Fig. 1: From an unidirectional power flow to a bi-directional power flow distribution system
Such developments may trigger problems such as overloading of components and connections, voltage
profile, voltage control and voltage instability. Moreover, short circuit currents will flow in a different
pattern, non-selective tripping may occur, transient phenomena behave in a different way.
Depending on the operational mode of the generators power flows vary widely. Under certain
circumstances the power flow may change quite suddenly, for instance due to wind shifts or due to
tripping of distributed generators during system disturbances. It is a task of the system planner to
foresee such changes in power flow patterns and to put forward loading requirements for the
components involved. Equipment overload capabilities may be required. Changing power flows are of
importance also for protection experts, as protection systems have to cope with the related currents and
patterns. Short-circuit currents will show directions and amplitudes that are dependent on the
operational mode of the distributed generators. In addition, in the DG systems (especially distributed
windmills) the short circuit profiles along distribution feeders are different then in the conventional
systems with single source feeding. These different profiles could be difficult to understand intuitively
and can impact the protection coordination.
4. Transient effects of a short-circuit current

Short-circuit currents cause mechanical and thermal stresses on equipment. Protection systems are
designed to be fast enough to prevent any thermal overstressing of the plant or network. The problem
of the dynamic forces due to the crest value of the short-circuit current is normally not covered by
62
protection systems (with the exception of current limiting fuses). In distribution networks the full
asymmetrical short-circuit peak current (i.e. with a complete DC-offset) is up to 2.5/2.6 times the rms
value for 50/60 Hz, respectively. However, in systems with a share of (a)synchronous machines like
DG contribution the SC peak current ratio can deviate as shown in the figure 2.
100
50
-50
-100
-150
-200
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Fig. 2: Short-circuit current in the vicinity of a generator with varying AC- and DC-component
(X-axis: time t in s; Y-axis: currents as percentage, 100% = AC-component at t=0)
Since the DGs are located in distribution systems they will increase the DC component in their
vicinity. However, to the opinion of WG A3.13, this increase of overall DC-time constant is unlikely
to exceed values specified in the Standards [11] apart from the breakers connecting the generators to
the network where the generator circuit breaker standard has to be considered [12]. On the other hand
in DG networks DC-components have to be considered more in detail for transient current transformer
dimensioning, despite the fact that the DC-time constants are still be expected to be lower than in
transmission networks.
If a DG system is located in the vicinity of an existing installation where the equipment is already
close to its SC limits, it might be necessary to utilize means of current limiting like Is-limiters, CLIPS
and other current limiting technologies. Conventional fuses could not be used to trip DG because they
would not be fast enough (for ratios of SC current to continuous current typical to the DG systems) to
act within the first half-cycle where the limiting is needed.
A closer look into modern (> 750 kW) windmills reveals that their behaviour during short-circuits is
much more complicated. As reported in [6][13] directly coupled synchronous generators are nowadays
less popular and the manufacturers trend is towards doubly fed induction generators. The reason for
the popularity of these variable speed DFIG is that the rated power of the expensive converter is about
one third of the rated power of a full converter, with rather comparable control possibilities. Their
transient behaviour during and immediately after short circuits can vary considerably depending on the
setting of the converter controls, including the setting of protection systems such as the crowbar across
the rotor-side converter [7][9][14]. As the settings are normally tuned for an optimal conversion of
wind energy [15], realistic settings of DFIG controls should be considered.
Power plants are designed to cope with three-phase terminal short-circuits, but are generally not tested
for the full short-circuit current. To our knowledge manufacturers of generators and manufacturers of
generating sets (windmills, co-generators, gas-turbine packages) are not fond of such experiments
despite the fact that short-circuits in MV-networks are very realistic and that the damping by the
63
network itself is very limited. Much effort, however, is given to establish simulation tools, that are
verified by small scale tests [14].
5. Transient phenomena at clearing the short-circuit
Many small power units are operated in such a way that for any disturbance in the network, they are
immediately switched off by a circuit-breaker (so, after roughly 50 to 100 ms). The protection of the
small units (less than a few MW up to 10 MW) is conventional: instantaneous overcurrent, time
overcurrent, instantaneous undervoltage, stator/rotor earth fault detection, time under/overvoltage,
time under/overfrequency, unbalance, high temperature, differential, under/over excitation, reverse
power, rotor overload. For generators connected to the MV-network without a step-up transformer, the
connecting circuit-breaker should be specified as a generator circuit-breaker to IEEE [12]. For
generators connected through a step-up transformer, the circuit-breaker at the MV-side may be
specified to IEC [11], but it is recommended to consider a design tested for test duty T30S (circuit-
breakers directly connected to transformers) [16].
For large power plants, and in many cases also for medium size power plants, utilities put forward
requirements that generating plants contribute to the stability of the network and remain connected for
a certain period of time. Especially after the disturbance these power units are needed to restore the
active and reactive power balance. The required ride through time varies from several hundreds of
ms for smaller or special units up to more than 1 s for large conventional power units.
Due to the enormous increase in distributed generation in recent years, utilities tend to put forward
similar requirements for the small power generating units. In a number of countries ride through-
requirements have been imposed on windmill-farms, groups of small power plants that together form
an equivalent of a medium size or even a large size power plant. Apart from the behaviour during
system disturbances, requirements are put forward with respect to voltage control and even for
100
terminal volt . /%
GEN
0
-100
5
current [ pu]
GEN
0
-5
300
active power /%
GEN
100%
reactive power /% 0
GEN
-300
180 180.
angular
displacement/gra 90.
GEN
0
-180
10
speed variat. /%
GEN
0
tf t oop GEN 25 MVA TA=5.08 sec

-10
0 0.352 0.704 1.056 [s]
Fig 3: Simulation of 25 MVA synchronous generator (CHP) dynamic behaviour

during/after 233 ms short-circuit in MV-network
64
frequency droop control. However, there is still a large dispute between those responsible for the
networks and those responsible for the small power generating units, with the main question of who is
to be blamed for the disturbances and who has to pay for the countermeasures necessary to warrant
system stability (see [17][18] and the related discussions).
As discussed in the former chapter, smaller units are stressed at the moment of the short-circuit
occurrence. Furthermore, these units are stressed a few cycles later when they are switched off and the
brake jammed on. The generators which remain connected to the grid will accelerate due to the lack of
load. Both frequency and phase angle will start to deviate from that of the main grid and some low
inertia power generating technologies, like for example aero-derivative gas turbines [8] or modern
windmills [6] will accelerate quite fast. Anyway, at the moment that the short-circuit is cleared by the
relevant circuit-breakers, the small generators suddenly have to face the system frequency and system
phase angle again. Figure 3 shows corresponding behaviour of a synchronous generator simulated with
NETOMAC with tf representing the 3-phase fault duration and toop the out-of-phase condition. For
synchronous machines, a relatively large out-of-phase current will flow between generator and the
system, leading again to high dynamic stresses, maybe even larger than the stresses due to short-circuit
currents [8], even with DFIG [9]. Users are warned that loss of life and serious damage to shafts,
gears, bearings, (gas/steam/wind) turbine-blades, etc. can occur [19].
Asynchronous machines are known to show a fast decreasing contribution to the short-circuit current
and after, say, 100 ms no back-emf will be experienced. At the disappearance of the short-circuit,
however, the large amount of reactive power required by the asynchronous generator will lead to a
sustained voltage dip and long start-up currents, but in amplitude lower than the short-circuit current
[20]. A more stressful condition is the short open-terminal situation where, at reconnection, currents
larger than the short-circuit current will appear [20].
In comparison with the above mentioned DGs, the

behaviour of the Double Fed Induction Generators
(DFIGs) is complicated. It is difficult to draw
general conclusions on the impact of these units on
the network conditions. For instance, DFIGs use
crowbars across the rotor circuits disconnecting the
converter from the rotor winding. After the crowbar
operates the DFIG becomes a conventional
induction machine, while before, it operates more or
less as a synchronous machine, as can be learned
from the equivalent circuits shown in [15]; see fig.4.
Therefore the total DFIG behaviour will also depend
on how fast and how long the crowbar will operate
Fig.4: Equivalent circuit of DFIG (sometimes only a few cycles, or even less [14])[21].
On the other hand, due to its ability to control in a
fast way the rotor torque and because of rotor speed decoupling, the dynamic behaviour of DFIG
shows to be generally more stable than that of synchronous generators [13].
6. Transient phenomena at clearing out-of-phase

It is clear that out-of-phase conditions under ride through-requirements are very realistic. Depending
on the ratio between the system short-circuit power and the generator short-circuit power, out-of-phase
currents may become more severe than short-circuit currents. Small synchronous units have to be
designed for out-of-phase conditions, otherwise the ride through-requirement will lead to
considerable damage.
Interrupting out-of-phase current may cause problems as well. Breaking the out-of-phase current is a
duty of the generator circuit-breaker; i.e. the circuit-breaker that connects the generator to the
distribution or industrial grid. This circuit-breaker may be specified according to IEEE [12] or it may
65
be a regular breaker specified to IEC [11]. Where in [12] an optional out-of-phase duty is required in a
straight forward approach, to [11] out-of-phase is more treated as a multi-part test: the initial part of
the TRV is covered by short-circuit tests and the crest of the TRV by a non-mandatory phase-
opposition test.
The wave-shape of the TRV under out-of-phase

conditions can be split into several time domain parts.
The power frequency recovery voltage part will
depend on the phase angle between generator and
system, and on the difference in frequency. This will
lead to the recovery voltage factor (theoretically up to
3.0 for the first clearing pole). The natural frequencies
and damping of the generator-system at one hand and
the network at the other hand will determine the TRV
in the ms range. Together or separate they determine
the amplitude factor and, in combination with the
recovery voltage factor, the crest value of the TRV.
The initial part of the TRV (i.e. the first hundreds of
s) forms the steep rise of the TRV (RRRV) which is
determined by the local capacitances and surge
impedances at both sides of the circuit-breaker. In
combination with dI/dt, the sum of equivalent surge
impedances and capacitances gives the RRRV and the
td (time delay). See figure 5.
IEC [11] gives a recovery voltage factor of 2.5 and an

amplitude factor of 1.25 for rated voltages below 100
kV. The crest value of the TRV under out-of-phase is
therefore 3.13 pu, where in IEEE [12] 3.18 pu is
specified. For smaller power units IEEE requires a
RRRV of 3.3 kV/s; IEC requires approximately ten
times lower RRRV, but the amendment 2 of [11]
requires that circuit-breakers directly connected to
transformers have to fulfil a RRRV larger than 3.3
kV/s, at least for test duty T30S [16].
Fig 5: Components of out-of-phase-TRV
CIGRE WG A3.13 has also conducted some system studies to simulate the phenomena described. To
that purpose a distribution network, typical for German cities, has been modelled, representing a 20
kV cable grid with municipal load and industrial load, including a co-generation plant of 25 MVA, fed
by two 110/20 kV transformers. A 3-phase short-circuit occurs somewhere in the 20 kV-network, not
far from the main substation, thus reducing the voltage from 20 kV to almost nihil. The generator will
accelerate up to full phase-opposition, when the short-circuit current is tripped and the generator, at the
20 kV-side of its step-up transformer, is disconnected from the grid. The TRV across this last circuit-
breakers first clearing pole has been simulated by EMTP, under the assumption of a pre-fault voltage
of 20 kV. The results are shown in figure 6. The simulated generator is a synchronous machine and its
excitation is assumed not to deviate too much from normal, due to the relatively low short-circuit and
out-of-phase currents (see also fig. 3).
Several cases have been studied, varying the load of the network (0, 10, 20 MW) leading to different
TRV-peak values (5.1 pu, 4.4 pu, 4.1 pu resp.), equivalent to amplitude factors as high as 1.7, 1.5, 1.4
resp. Also the connection between circuit-breaker and step-up transformer has been varied, leading to
an initial RRRV of 2.6 kV/s and a td of 0.3 s for a bare connection resp. a RRRV of 1.6 kV/s and a
td of 1.1 s for a cable connection.
66
100
[kV]
80
60
40
20
0
0.0 0.2 0.4 0.6 0.8 [ms] 1.0
(file 10MW .pl4; x-var t) v:1A -THA
Fig.6: TRV simulated for out-of-phase clearing, 20 kV, 10 MW load, cable-connection
In figure 6 can be recognised a high-frequency part generated by the step-up transformer (with the
generator and the cable connection to the circuit-breaker) and a low-frequency part at the other side
of the circuit-breaker, generated by the cable network. As in this case the short-circuit power of the
network is very low (the network impedance is relatively high), a very large part of the recovery
voltage ( totally: 3 pu due to out-of-phase) is experienced at the network side of the circuit-breaker and
the low damping of the cable network is dominant (a high amplitude factor). Opposite, the steep initial
part is completely dominated by the high eigen-frequency of mainly transformer and its connection to
the circuit-breaker. To amendment 2 of IEC 62271-100 [11] in such cases the circuit-breaker has to be
specified as class CSS (intended to be used in cable systems and in small capacitance connections to
transformers). For test duty T30S a RRRV of 4.95 kV/s (@ 24 kV) is required, double the RRRV for
normal capacitance connections, class CS [16].
But where the RRRV is covered by IEC, the peak-value of the out-of-phase TRV is only 61 kV, 3.1 pu
(@ 24 kV). The discrepancy between 3.1 pu and the simulated crest values is caused by the fact that
within IEC reduction factors have been introduced to take into consideration the low probability of full
phase opposition, and the low probability that the eigen-frequencies at both sides of the circuit-breaker
will match in such a way that both amplitudes may be added (i.e. a common high amplitude factor).
The examples given, however, show that full phase opposition is far from academic for DG, certainly
in case of ride-through requirements, and that high overall amplitude factors can be reached. In
addition it can be stated that, despite the fact that the out-of-phase RV will show a beat-pattern (fig. 3),
an out-of-phase angle above 90 gives already considerable TRV-peak values [8]. Furthermore, cases,
where the short-circuit power from the network are much higher than in the example of fig.5, will
show a proportionally larger amplitude for the high-frequency part and a much lower amplitude for the
network part.
It should be mentioned that the use of MV surge arresters and surge capacitances could in certain cases
reduce the magnitude as well as the steep rise of TRV.
67
7. Mis-operation and out-of-phase synchronisation
Other transient phenomena that have to be mentioned are false switching operations in the vicinity of
dispersed generation. In chapter 5, above fig. 4, an example of a short duration open terminal has been
given, leading to stressful currents at reconnection of asynchronous or synchronous generators [8][20].
Similar conditions will appear when in the vicinity of small generators a large power unit is connected
to the network without adequate synchronizing (mistake in operation, defect synchronisation check,
defect synchronizing equipment, pre-strike or re-strike in generator circuit-breaker, flash-over across
open poles of breaker or disconnector, etc.); but also somewhere in the network with re-closing or
automatic system restoration. Under such conditions the distributed generators will be disconnected
from the grid and switching duties comparable with those in the former chapter will be required.
Although it is outside the scope of this WG, it is recommended that the DG units utilizing synchronous
machines or equivalent machines be specified to withstand full terminal short circuit, out of phase
synchronising, loss of synchronism in the power network, and out-of-phase synchronising of
neighbouring units. It may be desirable that full terminal short circuit and out of phase synchronising
be verified with type testing.
8. System separation and islanding

With the growing use of the DG in distribution networks these DG systems should preferably support
the same requirements as large power units, i.e. ride-through most of power system disturbances
without being disconnected from the network. At the same time, since traditionally passive
distribution networks become active, distribution circuit breakers may have to perform new duties,
such as system separation, out-of-phase switching (chapter 6), synchronisation of generators and/or
systems, etc. Such requirements are not so clear from the Standards, where for instance the
specification for a dielectric type test across open contacts is roughly 2.9 pu/1 min., while a repetition
of a type test in service is performed at even lower stresses (e.g. 70% i.e. 2.0 pu).
Another issue of importance is safety: clearance, earthing, status-information, SCADA, interlocking

meaning that in the HV-devices more intelligence and communication has to be incorporated.
9. Transmission- and Subtransmission-connected DG

In some countries distributed generation is connected to a (sub)transmission grid in stead of a
distribution or industrial network. Such connections are for instance windfarms (in stead of singular
windmills), larger power plants, locations without distribution networks or too weak distribution
networks, power plants far from load centres and industrial plants with generating facilities directly
connected through separate transformers.
At transmission levels less problems are expected with such power plants, as transmission grids are
used to connect large generators. Moreover short-circuit powers are normally far larger than those of
the power plants, and short-circuit currents are cleared much faster than in distribution networks.
Multi-directional power flows and complicated short-circuit current patterns are not an issue at the
transmission level. Out-of-phase conditions in case of DG are not so different from out-of-phase
conditions of other plants and can therefore be treated equally (specified duties for switchgear, special
protection and control equipment, synchro-check and synchronising equipment, interlocking, etc.).
To put it the other way around: the problems described in former chapters show that distribution
networks are changing to more complicated grids, comparable with transmission networks. Therefore
deeper technical analyses, more comparable with the approach of transmission experts, are required.
10. Conclusions
DG, connected to the distribution grid through full converters are expected not to trigger any
special requirements for HV-components, apart from, infrequently, special requirements due to
harmonic resonance or to avoid harmonic resonance.
DG, based on rotating generators directly connected to a distribution or industrial network (maybe
through a step-up transformer) should be treated carefully with respect to its contribution to short-
68
circuit currents (complicated patterns), the appearance of out-of-step conditions (certainly when
ride through is required), reconnection and re-closing and more severe TRV-requirements.
Out-of-phase becomes more probable due to the low inertia constants and cannot longer be
ignored. The necessity for synchronising and islanding put forwards extra requirements for
primary and secondary equipments. Safety aspects will add additional requirements.
DFIG shows a complicated transient and dynamic behaviour, but will generally lead to softer
transients and a better dynamic performance (short-circuit currents, ride through stability, out-of-
step). Its behaviour is strongly influenced by the settings of the converter and crowbar controls.
There is a need to carefully specify and test both HV-equipment and power plant equipment for
short-circuit, out-of-phase conditions, overloading capabilities and system separation.
11. References
[1] Scoping Document, CIGRE WG A3.13 (Changing Network Conditions and System Requirements)
A3-04(SC)09 IWD / A3-04(WG13)59 IWD
[2] Issues Regarding the Integration of Induction Wind Turbines in Weak Electrical Networks
O.A. Monzn Alejandro
[3] Modeling New Forms of Generation and Storage, CIGRE TF 38.01.10
CIGRE Technical Brochure 185, April 2001
[4] Modeling of Gas Turbines and Steam Turbines in Combined Cycle Power Plants, CIGRE TF C4.02.25
CIGRE Technical Brochure 238, December 2003
[5] Interaction between the Distributed Generation and the Distribution Network:
Operation, Control and Stability Aspects, V. Knyazkin, Th. Ackermann
17th CIRED, Barcelona, May 2003
[6] Wind Power Modelling and Impact on Power System Dynamics, J.G. Slootweg
Dissertation at Technical University Delft (the Netherlands), 2003
[7] Fault Current Contribution of DFIG Wind Turbines, O. Anaya-Lara, N. Jenkins
IEE 3rd RTDN, London, 2005
[8] Distributed Generation in Relation to Phase Opposition and Short-Circuits
A.L.J. Janssen, e.a. 10th SCC 2002 in Poland
[9] Dynamic Modeling of Doubly Fed Induction Generator Wind Turbines
J.B. Ekanayake, L. Holdsworth, XueGuang Wu, N. Jenkins
IEEE PS, Vol.18, No.2, May03, pp. 803-809
[10] Dynamic simulation of gas-turbine generating unit, W.W. Hung
IEE Proceedings C, Vol. 138, No.4, Jul91, pp.342-350
[11] IEC Standard 62271-100, 2001
High-voltage alternating-current circuit-breakers
[12] IEEE Standard C37.013, 1997
Standard for AC High-Voltage Generator Circuit Breakers Rated on a Symmetrical Current Basis
[13] Transient Stability Studies Comparing Synchronous and Doubly Fed Induction Generators
J. Eek, K. Uhlen, T. Gjengedal CIGRE SC C1 Symposium, Athens, April 2005, Rep. 111
[14] Voltage dip ride through of a doubly-fed generator equipped with an active crowbar, J.K. Niiranen
[15] Incidence on Power System Dynamics of High Penetration of Fixed Speed and Doubly Fed
Wind Energy Systems: Study of the Spanish Case
J.M. Rodrguez, J.L. Fernndez, D. Beato, R. Iturbe, J. Usaola, P. Ledesma, J.R. Wilhelmi
IEEE PS, Vol.17, No.4, Nov02, pp.1089-1095
[16] Harmonization of TRVs in ANSI/IEEE and IEC Standards for High-voltage Circuit Breakers <100 kV
D. Dufournet, G.F. Montillet IEEE-PD, to be published
[17] Impact of Wind Energy Generation on the Safety of the Electrical Transmission Network
Juan M. Rodriguez, e.a. CIGRE SC C1 Session 2004, C1-204
[18] Large Scale Integration of Wind Power and the Impact on Power Systems
T. Gjengedal, e.a. CIGRE SC C1 Session 2004, C1-206
[19] Adjoint wind turbine modelling with ADAMS, Simulink and PSCAD/EMTDC, S. Uski, e.a.
[20] Behavior of Induction Motor Due to Voltage Sags and Short Interruptions, Juan C. Gomez, e.a.
IEEE PD, Vol.17, No.2, Apr02, pp.434-440
[21] Modelling, Control and Stability of Wind Turbines with Doubly Fed Induction Generator
G. Tsourakis, C. Vournas CIGRE SC C1 Symposium, Athens, April 2005, Rep. 117
69
Appendix C
Report presented at the IEEE/CIGR International Conference on Future Power

Systems 2005, Amsterdam
Report O 06-09
70
Changing Network Conditions and
System Requirements
A.L.J.Janssen M.Waldron J.Jger F. Iliceto
M.Glinkowski B. Middleton F.Gallon Q. Bui-Van
CIGRE WG A3.13
Abstract
CIGRE WG A3.13 is studying the impact of changing net- CIGRE WG A3.13 Changing Network Conditions and
work conditions and system requirements on the specifica- System Requirements is giving special attention to the
tion of conventional HV equipment within the scope of SC consequences of the growth in distributed generation (co-
A3. They see three main developments leading to future generation plants as well as sustainable power generation)
networks: the strong increase in distributed generation, the and to the consequences of long distance transmission (a
increase in the transmission of bulk power over long larger distance between power generation and power
distance and the increase in the application of power distribution, leading to voltage problems and the need for
electronics. The impact of each development is described, reactive power compensation). In both cases the interaction
conclusions are given or studies going on are highlighted. between protection and control systems on one hand and the
network dynamics on the other hand will play a dominant
Keywords role in the severity and probability of the phenomena that
high voltage equipment, distributed generation, long have to be withstood by the applied HV-equipment. These
distance transmission, reactive power compensation, phenomena have to be considered against a background of
transients, TOV, TRV, out-of-phase, filter banks increased utilisation of equipment (in terms of age, loading
and voltage stresses), reduction of size and complexity,
1. Introduction incorporation of more intelligence and the application of
more over-voltage protection and smart devices. [1]
Networks are changing due to business drivers such as
environmental concerns (including concerns as running out
of fossil energy sources), competitive power market, further 2. Distributed Generation
utilisation of transmission corridors, multi-directional power-
flows in distribution networks, increased capacity, increased In line with the scope of CIGRE SC A3 High-Voltage
efficiency, etc. These developments lead to technology Equipment, WG A3.13 is studying the consequences of
changes (e.g. distributed generation, wind-farms, shunt distributed generators on the equipment in the networks.
compensation, shunt/series-compensated lines, phase With that in mind the distributed generators can be divided
shifters, filter-banks, non-linear loads, HVDC, FACTS, into two groups: (a) generators directly coupled to the grid,
advanced protection and control systems) and possibly through a step-up transformer, and (b) generators
consequentially to special requirements, for instance with that are coupled by means of a full converter, probably
respect to harmonics, temporary overvoltages (TOV), through a step-up transformer. The directly coupled
transient recovery voltages (TRV), out-of-phase conditions, generators (a) can be synchronous or asynchronous
power quality, etc. machines, sometimes equipped with power electronic
_______________________________ devices (the so-called double-fed induction generators DFIG)
A.L.J. Janssen is with NUON Assetmanagement, the Netherlands, and to control the rotor current (voltage control, slip, frequency
Member of IEEE and CIGRE (anton,Janssen@nuon.com) for variable speed, reactive power output).
M. Waldron is with National Grid Transco, UK, and member of
CIGRE (mark.waldron@uk.ngrid.com) Due to the application of full converters, the latter group (b)
Prof. J. Jger is with the University Erlangen-Nrnberg, Germany, and is able to limit its contribution to fault currents, to support
a Member of IEEE and CIGRE (jaeger@eev.eei.uni-erlangen.de) reactive power balance and voltage control, to better restore
Prof. F.Iliceto is with the university of Rome La Sapienza, Italy, and systems after disturbances, to better adapt to frequency
a Life Fellow of IEEE (francesco.iliceto@uniroma1.it)
M. Glinkowski is with ABB Inc., Raleigh/USA, Bergamo/Italy, and a variations, phase angle variations and voltage variations.
Fellow of IEEE (mietek.glinkowski@us.abb.com) Thus converters have an advantageous influence on the
B. Middleton is with BC Hydro, Canada, system behaviour and therefore on the requirements for other
(bob.middelton@bchydro.bc.ca) components. Besides the added costs, one main disadvantage
F. Gallon is with Areva T&D, France, and Member of CIGRE
(francois.gallon@areva-td.com) is the generation of harmonics, although nowadays converter
Q. Bui-Van is with Hydro Quebec-TransEnergie, Canada technologies and mitigation technologies show substantial
(bui_van.que@hydro.qc.ca) improvements. The topic of power electronics will be dealt
All are members of CIGRE WG A3.13. with later on.
71
As illustrated in Fig. 1, distributed generation in distribution, leading again to high dynamic stresses, maybe even larger
sub-transmission, and industrial networks leads to a than the stresses due to short-circuit currents. This applies
structural change of the power flows, as the generated energy also to asynchronous generators and even to DFIG. The
is mostly independent from the local energy demand. stresses have to be withstood by the HV-equipment as well,
Changing power flows are of importance also for protection but the generators short-circuit or out-of-phase currents are
experts, as protection systems have to cope with the related normally smaller than system fed short-circuit currents. But,
currents and flow patterns. Short-circuit currents will show breaking the out-of-phase current is an optional duty for MV
directions and amplitudes that are dependent on the circuit-breakers[13][14] and its specification has probably to
operational mode of the distributed generators. In addition, in be re-considered [2].
the DG systems (especially distributed windmills) the short
circuit profiles along distribution feeders are different from A closer look into modern, larger (> 750 kW) windmills with
those prevailing in the conventional systems with single doubly fed induction generators (DFIG) reveals that their
source feeding. These different profiles could be difficult to behaviour during short-circuits is much more complicated.
understand intuitively and can impact the protection Their transient behaviour during and immediately after short
coordination and the short-circuit current withstand and circuits can vary considerably depending on the setting of the
interruption requirements for the related switchgear. converter controls, including the setting of protection
systems such as the crowbar switch across the rotor-side
converter. On the other hand they show inherently a much
G
better stability performance and fault ride through capability.
G
G
The developments in these modern drive technologies is still
rapid and general conclusions cannot yet be drawn although
G
there is an inkling that their contribution to transient currents

G
G
and voltages may in fact be favourable in comparison to that
of conventional generators (less severe, but detectable, so
G
that protection systems can still be selective). The main drive
for these characteristics is the fact that the WEC (wind
G
G
energy converter) itself will suffer less from system
G G
transients and can continue its operation.
Fig. 1: Structural change of power flows in distribution grid
Other transient phenomena related to DG are false switching
Many small power units use to operate in such a way that for operations in the vicinity of dispersed generation; such as
any disturbance in the network, they are immediately accidental system separation or generator disconnection,
switched off by a LV or MV circuit-breaker (after roughly 50 false synchronisation of separate systems or generators (plus
to 100 ms), thus enabling selective protection, auto-reclosing consequential tripping), network restoration and auto-
and fault location in the distribution grids and preventing reclosing under adverse conditions. Also power system
out-of-phase conditions, unintended island operation, safety instabilities (angular, voltage, or frequency instability) could
problems and severe damage to the power plants. However, lead to severe transient phenomena and consequential events.
more and more utilities put forward requirements that In this respect it should be stated that, due to the trend to
generating plants contribute to the stability of the network work plant harder, it is expected that systems, including MV-
and remain connected for a certain period of time so that networks with DG, are operated closer to the stability limits.
voltage sags can be limited in amplitude and duration,
overloads are reduced and regional power deficits are Measures to gain the profits from DG and to minimize its
avoided. The ride through time is specified as several drawbacks for the network are with a clever optimization of
hundreds of ms for smaller or special units up to more than 1 protection settings, i.e.: no disconnection within the first
s for large conventional power units. protection zone in transmission networks (for instance: 200
ms), disconnect synchronous generators from faulty MV-
Power generating units are stressed at the moment of the lines before the auto-reclosing time, disconnect
short-circuit occurrence. Furthermore, these units are asynchronous generators with too low inertia constants from
stressed again a few cycles later when they are switched off faulty MV-lines within auto-reclosing time, disconnect
and the brake is applied (especially windmills). The synchronous generators automatically under islanding
generators which remain connected to the grid will accelerate conditions, disconnect asynchronous generators under self
due to the lack of load. Both frequency and phase angle will excitation conditions, etc [3]. However these measures are
start to deviate from that of the main grid and some low only feasible when short protection clearing and auto-
inertia power generators, like for example aero-derivative reclosing times (100 ms to 200 ms) can be reached, which is
gas turbines or modern windmills, will accelerate quite fast. normally not the case in MV-networks [2]. As stated before
Anyway, at the moment that the short-circuit is cleared by DFIG and generators connected through full converters
the relevant circuit-breakers, the small generators suddenly should inherently be able to fully ride through fault
have to face the system frequency and system phase angle conditions and supply power within a few cycles after fault
again. For synchronous machines, a relatively large out-of- disappearance.
phase current will flow between generator and the system,
72
Another issue of importance is safety concerns, such as: Shunt compensation is applied in meshed networks as well as
clearly de-energized and/or earthed parts of the network, in long radial transmission systems. Due to the unbundling
reliable status-information on the generators through between power generation and power transmission, the
SCADA-systems, and interlocking functions where power plant locations are in many cases not optimal from
necessary. Furthermore functions as synchro-check, the transmission network point of view. Consequentially a
synchronisation, voltage check (auto-reclosing) and structural mismatch of reactive power and need for reactive
frequency check (islanding) have to be built in at the location power control becomes evident in almost all HV-grids. With
of the HV/MV switchgear. These measures are more or less the increased application of other reactive power compensat-
comparable with those for large power plants connected to ion equipment, there is a demand for highly reliable circuit-
transmission networks and serve to a large extend to protect breakers to switch shunt capacitor banks or shunt reactors on
the power plants. Together with the mitigation measures a daily basis or even more frequently. The test procedures
mentioned in the former paragraph and the requirements for prescribed in the IEC 62271-100 for capacitive current
more advanced and faster protection systems, this means that switching and in the IEC 62271-110 for small inductive
more intelligence and communication has to be incorporated current switching [21] should be rigorously applied to reflect
in the HV-devices the most frequent operational duties . Additionally there will
be a significant increase in the application of controlled
Since traditionally passive (no power source) distribution switching devices (point-on-wave) in order to avoid high
networks become active, distribution circuit breakers may switching overvoltages or inrush transients during shunt-
have to perform new duties, such as system separation, out- reactor or shunt-capacitor switching. In recent years CIGRE
of-phase switching, synchronisation of generators and/or WG A3.07 recommended special test requirements for
systems, etc. Such requirements are not so clear in the circuit-breakers with point-of-wave controllers [20].
existing Standards for general purpose circuit-breakers, but
in paragraph 4.2 of IEC 62271-100 an insulation level has The reactive power compensation in series compensated
been suggested for specific circuit-breakers with systems is created by series-capacitor banks that are located
occasionally separate systems at both terminals [13]. Also along transmission lines, preferably in the substations at the
out-of-phase switching is an optional duty in IEC 62271-100, end of long line sections. In most applications fast re-
seldom applied, but in the near future a larger number of insertion of SCs is required for the effective up-rating of
MV-circuit-breakers might be subjected to this duty transmission angular stability. Instantaneous re-insertion is
(depending on fault ride through requirements and the achieved by protecting SCs against transient overvoltages by
implementation of mitigation measures). Out-of-phase angles means of adequately rated metal-oxide varistors (MOV),
above 90 lead already to considerable peak values of the supplemented by a forced triggered spark-gap and a by-pass
TRV[4]. Moreover the related RRRV (Rate of Rise of circuit-breaker; Fig. 2. Adverse effects of series compensat-
Recovery Voltage), as specified in the IEC-Standard, has to ion can be mentioned: special requirements for line protect-
be taken from test duty T30S [19], meaning that a class S ion relays, sub-synchronous resonances (SSR), the increase
circuit-breaker (directly connected to a step-up transformer) of switching overvoltages (SOV) along transmission lines
has to be specified [2]. IEC TC 17A/C has accepted the IEEE and the increase of transient recovery voltages (TRV)
Standard on generator circuit-breakers C37.13 [16], where imposed on the line circuit-breakers. Special requirements,
for special applications the above requirements are covered. beyond the Standards, are put forward to cover the stresses
In both the Standards and the Application Guides users have anticipated in series compensated systems. The requirements
to be attended at the above requirements for HV switchgear are specific per application of series compensation and
that may face out-of-phase and system separation events. within CIGRE WG A3.13 experts try to compile an overview
Furthermore utilities are looking for guidance to support of the international practices, in order to look for possibilities
them in specifying appropriate short-circuit current withstand for harmonisation and the exchange of experience [18].
and clearing requirements [5][6], more comparable to those
of industrial and off-shore plants.
3. Long Distance Transmission

When active power is transmitted over very long distances,
reactive power compensation is required in order to keep
voltage profiles along transmission lines within acceptable
operating voltage range for various load-flow conditions.
The following equipment can generate or absorb reactive
power: shunt-capacitor banks, shunt-reactor banks,
generators in leading/lagging modes, synchronous
condensers, series-capacitor banks, series reactors,
transformers, cables, lightly/heavily loaded lines, Static VAr
Compensators (SVC), HVDC converters, FACTS, and
variable Mvar reactors [7].
Fig. 2: basic components of a series capacitor bank
73
In the case of long distance transmission, extreme circuit-breaker trip sequence on the QB circuit as the currents
disturbances can cause a loss of synchronism with distant in each of the QB phases are interrupted.
generating centres, leading to system separation and to full
load rejection by simultaneous opening of several line During three phase interruption of a single-phase fault to
circuit-breakers under loss-of-synchronism or out-of-phase earth on the series terminals of the QB, and depending on the
conditions. Following full load rejection, TOV due to the characteristics of the circuit-breaker, it is probable that the
Ferranti effect appear on long unloaded lines that are still faulted phase will be the last to clear. This imposes a voltage
connected to generators. Special TOV protection schemes of 1pu across the QB series winding, causing the voltage on
consist of the fast removal, after full load rejection, of the two un-faulted phases to rise (approx. 2.5pu), as a result
unloaded lines from generators. Therefore, line circuit- of the coupling between the series and shunt transformer.
breakers should have adequate TRV withstand capability for The remaining energised phase of the shunt unit can further
switching unloaded lines under high TOV conditions. Again contribute to the total voltage seen on the un-faulted
utility experts come with specific requirements for their HV terminals, resulting in a transient over-voltage of up to 4 or 5
equipment (circuit-breakers, arresters, special switching pu [1].
equipment, switchable MOA [10], controlled switching,
protection and interlocking systems [17], instrument The effect is non-linear, meaning that even tap positions
transformers, OH-line insulation, etc.). adjacent to the centre tap can contribute to very large
voltages on the QB during single phase to earth faults.
Additional topics such as higher than rated operating Unless the QB is at centre tap the voltage applied at the QB
voltages, transient overvoltages, out-of-phase conditions, terminals is excessive (beyond the insulation level) and can
switching duties on long lines and mitigation measures are only be controlled by the use of surge arresters. Because the
under study within CIGRE WG A3.13 and results will be duty is severe, high-energy surge arresters must be fitted to
released at the CIGRE SC A3 Session 2006. Within the both sides of the QB. Another solution may be found in
scope of another CIGRE WG A3.19 is the specification of special protection schemes with single pole tripping.
the TRV caused by faults on long lines, where triangular
wave shapes similar to those of the short line fault (SLF) are QBs have special requirements as regards dielectric tests and
generated, but with a much higher peak value and lower withstand capacity to the short circuit electro-dynamic
frequency [12]. Conclusions and recommendations for both stresses, which make their design complicated [22].
single phase and three-phase faults are in preparation.
4. Power Electronics
Another special topic to be addressed under long distance
transmission is that of long compensated AC-cables. Power electronics normally offer a lot of flexibility,
EHV/UHV cables with a length of several tens of km will controllability and a smoothing effect in terms of
normally need shunt reactors, preferably with variable Mvar overcurrents and overvoltages in the network. Unless pulse
absorption capacity, to compensate the reactive power width modulation (PWM) or an equivalent technology is
generated by the cables. Several articles have been published applied, a negative impact is the distortion of voltages and
on the interaction of compensated EHV-cables with the currents due to the generation of harmonics, giving an
series connected OH-lines and/or the surrounding network. adverse effect on power quality. Power electronic devices
Phenomena like harmonic resonance and sustained themselves are also vulnerable to power quality, over-
secondary arcs have been reported. Third harmonic currents, over-voltages, short-circuit currents and transients.
resonance for example could build up due to the strong
reduction of the systems Eigen frequency, as caused by the The level of distortion in the system voltages can be reduced
large capacitive cables. Saturation effects of transformers or by the application of harmonic filters. However, filter banks
shunt reactors can lead to severe temporary overvoltages and filter-bank switching are not covered by the Standards
(TOV), for instance when energizing an unloaded mixed [13][14]. All equipment within a filter-bank is subjected to a
EHV-line [8]. Harmonic overvoltages, large secondary arc high content of harmonic currents and is to be specified
currents and ground return currents with or without accordingly.
countermeasures put forward requirements for additional
and/or special HV-equipment (for instance: heavy duty With a high percentage of harmonic distortion, especially at
arresters, high-speed grounding switches, additional circuit- the lower harmonics, the currents may show more than two
breakers, neutral reactors, etc.). current zeros per power frequency cycle. In this way the
breaking of the current might take place earlier than at the
A component also related to long distance transmission is a regular power frequency current zero but as a filter-bank
phase shifter or quadrature booster (QB). The main reported behaves mainly as a capacitor-bank, current interruption at
transient phenomenon is the behaviour of QBs under an instant not corresponding exactly to voltage maximum
unbalanced fault conditions. In either buck or boost mode will lead to lower trapped charges (DC-voltages) on the
the QB relies on a balanced three-phase system to function filter-bank (capacitor-bank), thus reducing the dielectric
properly since the method of operation utilises two phases to stresses at current interruption. On the other hand, compared
energise the third phase. Consequently, for a single-phase with switching off a regular capacitor-bank, for filter-banks
fault at the terminals, very large voltages can appear on un- the circuit-breaker TRV will show harmonic voltages
faulted phases of the QB. These appear transiently during the superimposed on the normal TRV resulting in a higher
74
overall TRVpeak. In addition it is also possible that voltage overcurrents and overvoltages exceeding the design
resonance occurs at the busbar, controlled by the filter-bank capabilities will quickly destroy the power electronics.
to be switched. Another phenomenon to be considered is that Although for the primary equipment the limitation of fault
of switching off of a reactor in series with the capacitor-bank currents can be regarded as an advantage, for system
(as in a filter-bank) causing a positive voltage jump and thus protection it normally is a disadvantage. To that effect,
a higher peak value of the recovery voltage. A third issue to CIGRE WG A3.16 (Fault Current Limiters) is studying the
be considered is that, especially with HVDC applications, interaction between protection systems and current limiting
load shedding or blocking of the converter may occur. devices.
Through the Ferranti effect TOVs might occur lasting up to
1s, depending on the automatic voltage regulation (AVR) of Another remark to be made is that the transient behaviour of
nearby generators, or up to tens or hundreds of seconds, converters is to a certain extend also determined by the
depending on transformer tap-changer actions [9]. During process behind the converter, certainly when active power is
TOV condition the reactive power compensation, including involved. Sometimes the electric power generation
the filter-banks, should be switched off, again leading to a technology behind the converter reacts far slower than the
higher than normal TRV. reaction time of the power electronic devices, thus limiting
the inherent good performance of the converters. Fuel cells,
To verify the breaking capacity of a circuit-breaker, the for instance, are reported to give the effect of very slow
following statements can be accepted. The harmonic content responses to transients in the network, thus eliminating
of the filter-bank current does not impact the breaking certain advantages of the converters. This however, is more a
performance of a circuit-breaker, especially of modern problem for system studies than for the interaction between
technology, due to the short physical time constants of the systems and HV-equipment.
arc (the same phenomenon allows for the current injection as
applied at synthetic tests in power labs). However, the The increased injection of harmonics from LV-grids into MV
superimposed harmonic components in the recovery voltage and HV-networks is a problem of power quality and is more
might be relevant during the first milliseconds after related to the vulnerability of the same sort of LV-equipment
interruption. Nevertheless, as many circuit-breakers are as the equipment that generates harmonics. For the HV-
designed to withstand the RRRV of short-line fault tests, it is equipment in the distribution and transmission networks it is
expected that the initial part of the TRV at filter-bank not (yet) regarded to form a problem, apart from the
switching is not critical (at least at somewhat longer arcing equipment used in filter banks and for filter bank switching
times). as described before.
With respect to the peak-value of the TRV, the harmonic

component in the recovery voltage, plus the other 5. Conclusions
phenomena (positive voltage jump, TOV, possible
resonance) can be ascertained after proper system studies and Networks are going to be operated at their limits with regard
the applicable TRV envelope can be defined. Then, (a) the to loading, means of operation, geographical extensions and
calculated peak value can be compared with the performance stability criteria. The physical effects as e.g. dynamic
of the circuit-breaker under purchase or (b) the circuit- behaviour of generators and TRV due to long line switching,
breaker can be tested for the higher peak-value or (c) a are well-known for a long time, but the consequences are
circuit-breaker with a higher rated voltage can be chosen or now becoming realistic because of the increased DG as well
(d) special measures can be taken e.g. MOV parallel to the as long distance transmission projects. Measures to mitigate
arcing chambers. Test circuits are available to verify circuit- or reduce these effects and adaptation of the Standards and
breaker performance under filter-bank switching conditions Application Guides will be necessary. Utilities are looking
and the phenomena described are most pronounced in the for guidance and harmonisation of their specific
first pole to clear [11]. requirements.
Filter-banks should be switched by switchgear with a very Specific topics mentioned are:
low probability of re-strike. The distortion of the filter-bank
current will not be a problem, but the recovery voltage will Load current and short-circuit current profiles in net-
be (slightly) higher than that at switching regular shunt works with DG change with respect to flow, amplitude
capacitor banks, as specified in the IEC Standard [13]. and variation. Consequences for involved HV-
Furthermore, circuit-breakers with a high class of mechanical equipment are evident.
endurance (M2) are necessary and it is recommended to In case of transients, immediate disconnecting of DG
apply controlled switching, if available [1]. will limit the special requirements for HV-equipment
only to the withstand capability of short-circuit peak
Note that power electronic converters are expensive so that values.
they are not oversized and their ability to withstand Fault ride-through requirements will have an impact on
conditions beyond their rating is very limited. Moreover, an the specifications of circuit-breakers, that have to be
important reason for a converter to limit its short-circuit able to withstand and switch under out-of-phase
current contribution is to protect itself. Moderate conditions, synchronisation, and islanding. IEEE C37.13
or IEC 62271-100, class S, are recommended for such
75
applications; the dielectric withstand capability should G. St-Jean e.a. CIGR 1986, 33-01
be adapted and more guidance to users is needed. [11] Switching of Harmonic Loaded Capacitor Banks and
The impact of transients is, relatively speaking, larger on Filter Circuits
dispersed power generating units than on the network K.H. Krger, W. Schultz, N. Trapp, G. Morav
and its HV-components. Countermeasures such as a CIGRE SC13 Session 1990, 13-104
clever application of advanced protection schemes, as [12] Severe Duties on High-Voltage Circuit Breakers
illustrated, are therefore preferred. Observed in Recent Power Systems
It is expected that new technologies like DFIG will relax H.Hamada, Y.Kasahara, T.Shimato, K.Hirasawa,
the phenomena described. K.Suzuki, T.Yoshizumi
Filterbanks require specific requirements for the peak CIGRE Session 2002 Rep. 13-103
value of the TRV. Controlled switching is recommended [13] IEC Standard 62271-100, High-voltage switchgear and
for frequent switching duties of filter- and shunt controlgear
compensation banks and adequate test specifications Part100 High-voltage alternating-current circuit-breakers
are needed. [14]ANSI Standard C37-06, AC High-Voltage Circuit
Series compensation, loss-of-synchronism, switching Breakers Rated on a Symmetrical Current Basis
under TOV or high operating voltage conditions, long Preferred Ratings and Related Required Capabilities
compensated EHV-cables still require special attention [15]IEEE Standard C37-09, IEEE Standard Test Procedures
for each individual case. Attempts are made to come to for AC High-Voltage Circuit Breakers Rated on a
an overview of practices and to some harmonisation of Symmetrical Current Basis
solutions and specifications for HV-equipment. [16]IEEE Standard C37-013, Standard for AC High-Voltage
Generator Circuit Breakers Rated on a
Literature Symmetrical Current Basis
[1] Scoping Document, CIGRE WG A3.13 (Changing [17] Control of Over-voltages on Hydro-Qubec 735-kV
Network Conditions and System Requirements) Series-Compensated System during a Major Electro-
CIGRE A3-04(SC)09 IWD / A3-04(WG13)59 IWD mechanical Transient Disturbance
[2] Changing Network Conditions and System Q. Bui-Van e.a. IPST 2001, Brazil
Requirements. Part I: Distributed generation [18] TRVs across Circuit Breakers of Series Compensated
A.L.J. Janssen, e.a. Lines. Analysis, design and Operation Experience in the
CIGRE SC A3&B3 Colloquium 2005, Tokyo, Rep.103 420 kV Turkish Grid
[3] Modelling and Computer Simulation of Dispersed F. Iliceto
Generation in Distribution Networks. Measures to CIGRE Session 2002 Rep. 13-109
prevent disconnection during system disturbances. [19] Harmonization of TRVs in ANSI/IEEE and IEC
F. Iliceto, e.a. Standards for High-voltage Circuit Breakers
IEEE Bologna Power Tech Conference 2003, Vol.3 Rated less than 100 kV
[4] Distributed Generation in Relation to Phase Opposition D. Dufournet, G.F. Montillet
and Short-Circuits IEEE-PES, to be published
A.L.J. Janssen, e.a. 10th SCC 2002 in Poland [20] Controlled Switching of HVAC Circuit Breakers
[5] Short-circuit Currents with High DC Time Constants: Planning, Specification and Testing of controlled
Calculation Methodology and Impact on Switchgear switching systems
Specification & Rating Electra 197 (Aug. 2001), pp. 23-33
T. Fairey, M. Waldron CIGRE Technical Brochure 264
CIGRE SC A3&B3 Colloquium 2005, Tokyo, Rep.104 [21] IEC Standard 62271-110, High-voltage switchgear and
[6] General Report of CIGRE SC A3&B3 Colloquium 2005, controlgear (2005) Part 110: Inductive load switching
Tokyo, CIGRE A3-05(SC)28 (available on WEB-site) [22] IEEE Standard C57.135 (2001), IEEE Guide for the
A.L.J. Janssen, e.a. Application, Specification and Testing of Phase-Shifting
[7] Design and Application of Variable Mvar Output Shunt Transformers
Reactors with On-load Tap-changer.
Operation Experience in Africa
F. Iliceto, G. Bertagnolli, e.a. In 1977, as M.Sc. in Electr.
CIGRE SC 12 Session 1998, Rep. 12-308 Engineering, Anton Janssen
[8] Design and operation of EHV transmission lines joined PNEM, a Dutch utility,
and became manager
including long insulated cable and overhead sections
Transmission. In 1993 he joined
F. Iliceto, L. Colla, F.M. Gatta, S. Lauria KEMA as manager High-Power
7th Int. Power Engineering Conf. Singapore, 2005 Laboratory. Since September
[9] Overvoltages in HVDC Transmission System Caused 2002 he is with NUON, another
by Transformer Magnetizing Inrush Currents Dutch utility, as Assetmanager.
J.P. Bowles IEEE PES SM73 He has been active within
[10] Solution to the Problem of Separation of the Hydro- CIGRE SC 13/A3, IEEE-PES
Quebec 735 kV System Switchable Metal-oxide and IEC SC 17A/C working
Surge Arresters groups. He is convenor of
CIGRE WG A3.13.
76
Appendix D
DIELECTRIC, SWITCHING AND SYSTEM REQUIREMENTS

UNDER OUT-OF-PHASE CONDITIONS, DURING SYNCHRONISATION AND
UNDER COMPARABLE STRESSES
Report presented at the CIGR Sc C4/A1/A2/A3/C1 Symposium on

Transient Phenomena in Large Electric Power Systems
in Zagreb, April 2007-03-18
Report 0701
77
0701 ZAGREB
Dielectric, Switching and System Requirements under Out-of-Phase

Conditions, during Synchronisation and under Comparable Stresses.
A.L.J. Janssen F. Iliceto Q. Bui-Van S. Morais M. Waldron B. Middelton

The Netherlands Italy Canada Brazil U.K. Canada
On behalf of CIGRE WG A3.13
SUMMARY
Recent developments in electrical networks can increase the probability of out-of-phase switching and
dielectric stresses being applied to open circuit-breakers, due to asynchronous systems at both sides.
This report presents a systematic study of TRV-stresses associated with generator separation and
system separation. TRV peak values are higher than required in the Standards, even for relatively
small out-of-phase angles (75 to 90), and the dielectric stresses are high with respect to the short-
duration power frequency withstand voltages across a circuit-breaker open contacts, especially taking
into consideration the external insulation under pollution and ageing processes. To the opinion of the
authors, the Standards should be revised to give users clear and adequate guidance on the assessment
and specification of TRV-values and dielectric withstand requirements under out-of-phase conditions.
KEYWORDS
Out-of-phase, synchronisation, TRV, RRRV, First-Pole-to-Clear Factor (fpcf), longitudinal dielectric
stress
1. INTRODUCTION
Within CIGRE SC A3 High-voltage Equipment, WG A3.13 Changing Network Conditions and
System Requirements has investigated the impact of developments in electrical networks upon
conventional high voltage apparatus. The major relevant trends identified are:
1) increasing implementation of distributed generation
2) increasing distances of bulk power transmission
3) increasing application of power electronics (generation, transmission, distribution and load).
One of the phenomena studied is the increased probability of out-of-phase conditions. Operating of
systems closer to their limits may lead to steady-state, transient and dynamic stability problems and
the problems are exacerbated by the increasing complexity of the power systems: large distances
between load and power generation centres, regional concentrations of wind farms and associated
power transmission and reserve problems, the changed nature of distribution grids and a trend to
consider island operation of parts of the (distribution) system.
anton.janssen@continuon.nl
78
Fig. 1 Longitudinal stresses across the first-pole Fig. 2 Characteristics of the RV and TRV
to clear out-of-phase
The stresses on HV equipment, especially circuit-breakers (figure 1), under out-of-phase conditions
and during synchronisation of generators and networks have been investigated and are presented in the
following sections. Present Standards [1][2][3][4] define out-of-phase TRV (transient recovery
voltage) and RRRV (rate of rise of recovery voltage) conditions on the basis of parameters including
out-of-phase angle (), out-of-phase current (Ioop), recovery voltage (RV), natural frequency, damping
and amplitude factors (AF) and travelling wave behaviour. Figure 2 shows schematically the different
time domains which are relevant for the TRV-studies and reference [5] presents the relation between
the different parameters in 3-phase systems. The out-of-phase test duty leads to the highest TRV-peak
requirements for circuit-breakers.
WG A3.13 will publish more detailed information in two CIGRE Technical Brochures during 2007.
2. SYSTEM CONSIDERATIONS
The circumstances that may lead to system separation, either singly or in combination, include:
transient instability (slow fault clearing, false synchronisation of large network elements or large
power plants)
voltage instability (inadequate reactive power and/or voltage regulation, poor or adverse
tapchanger control)
small signal instability (amplification of power swings due to negative damping)
frequency instability (system inability to react to sudden load/generation unbalances)
cascade trippings (multiple lightning, weather conditions, overloading, vegetation growth,
temporary overvoltages)
protection mal-operation
false synchronisation of a single generator.
Large increases in distributed generation, including many windmills and windmill parks, and multiple
power transfers across longer distances, increase the probability of occurrence of many of these events
as detailed in the following examples:
79
medium voltage networks typically have fault clearing times which exceed the maximum clearing
time for continued stability of small generating plants equipped with synchronous generators
the optimal control of reactive power supply and voltage regulation by small generators has not
been established yet
windmills are very sensitive for wind variations, especially under high wind conditions, which
may result in co-incident tripping of many units
small cogeneration plants (e.g. for greenhouses) are operated in large groups without consideration
of wider network requirements
systems are more commonly operated up to, or even beyond, their loading capabilities
(small) generators are tripped and synchronised more regularly than ever before
certain distributed power generation technologies cannot provide inertial energy required for the
immediate dynamic response to sudden load/generation unbalances. This reduces the average
inertia constant of the whole system and hence reduces the margin to the dynamic stability
it is important that dispersed generators remain connected to the network during voltage and/or
frequency deviations caused by faults and other disturbances as specified for the large
conventional synchronous generators, thus contributing to ride through system disturbances with
their active and reactive outputs and their inertia
on the other hand, the growing use of dispersed generation increases the probability of out-of-
phase conditions.
All these trends lead to the conclusion that out-of-phase conditions have to be studied more carefully
than in the past. A better understanding of the effects and consequences of out-of-phase conditions and
of the present and future probabilities of occurrence is necessary.
3. OUT-OF-PHASE PHENOMENA
In common with the recently published Guide for

Application of IEC 62271-100 and IEC 62271-1
[15], two out-of-phase cases are considered here:
(i) generating units that separate from the network
(ii) major systems that separate.
Whilst the focus of the above mentioned Guide is to
explain the values given in the Standards, a more
fundamental approach is taken here with emphasis
on the behaviour of system topologies not directly
considered in the Standard.
3.1 Case (i)

Out-of-phase switching may be applicable to a
generator circuit-breaker at the MV-terminals of a
generator, as specified in IEEE Standard C37.013
(1997) [10], or to a generator circuit-breaker at the
HV side of the step-up transformer, normally
specified as a general purpose circuit-breaker to IEC
62271-100 or ANSI/IEEE C37.04/06/09. In both
situations, as shown in figure 4, the total RV is
caused by the disappearance of the voltage drop
across the reactances of the generator, the step-up
transformer and the system and the overall fpcf:
RV = Ioop * fpcf * (Xd + Xtr + Xs). The overall fpcf
is a combination of the fpcf (depending on the
neutral treatment Zn) of the systems at both sides of
Fig. 3 Out-of-phase RV with f = total fpcf the circuit- breaker and can be deduced from the
double Neptune-scheme as shown in figure 3.
80
The largest voltage drop will generally be across the generator sub-transient reactance. The transformer
reactance is in the range from 0.1 to 0.15 pu whilst many modern generators have a sub-transient
reactance in the range 0.18 0.27 pu; lower values (0.12 0.15 pu) were typical in old 2-pole turbine
generators. The system reactance is typically five (or more) times smaller than sub-transient generator
plus transformer reactance. Further, the natural frequency of the generator windings is 2 to 3 times lower
than the natural frequency of the transformer windings. System frequencies usually have the lowest values
defined primarily by the travelling waves of the shortest OH-lines. In terms of surge impedances and local
capacitances, the generator will offer the lowest surge impedance (in the range of several tens to less than
100 Ohms) with the highest capacitance (typically 0.1 F) and the transformer the highest surge
impedance (thousands of Ohm) with the smaller local capacitances. The systems surge impedance does
not exceed 300 - 400 Ohm with local capacitances comparable with the capacitance of a transformer
(thousands of pF).
Fig. 4 Generator circuit- breaker RV
Seen from the HV-side of a step-up transformer (winding configuration: YN-D), it is assumed that the
earth fault factor and therefore the first-pole-to clear factor (fpcf) are very low: k=Z0/Z1 is 0.7 to 0.9.
With k being about 0.8 the fpcf becomes 0.92, and the second and last pole clearing factors are larger
than the fpcf [5]. On the other hand, as shown in figure 3, the total fpcf is to be considered and not the
individual fpcf at each side of the circuit-breaker. For k=0.8 at the generator-transformer side and a
fpcf of 1.1 at the net-side, the total fpcf is 0.94 to 0.96, depending on the ratio of normal sequence
reactance at the generator/transformer side versus the normal sequence reactance at the network side.
With a fpcf of 1.3 at the net-side, the total fpcf becomes 1.06 for a sub-transient generator/transformer
reactance which is five times the reactance at the net side. To derive the total RV, this total fpcf has to
be multiplied with the out-of-phase voltage which depends on the out-of-phase angle. In this example,
the total RV for full phase opposition will reach a value of 2.12 pu, with 5/6 of the voltage appearing
at the step-up transformer side of the circuit-breaker and the remaining 1/6 appearing at the network
side, in addition to the pre-clearing voltage of 4/6 pu. So, at the step-up transformer side the terminal
voltage of the first clearing pole jumps from 0.67 pu to 1.10 pu ( = 1.77 pu) and at the other
terminal from 0.67 pu to 1.02 pu ( = 0.35 pu). For smaller out-of-phase angles , the total RV, the
two parts of the RV and the voltage jumps are smaller in proportion to sin().
At the generator-transformer side the amplitude factor (over-swing) will be quite large (for instance
80%: amplitude factor 1.8), as the losses will be relatively low (X/R ratio of 50 or more) and the
generator side capacitance large. A significant depression of the voltage at the generator terminals and
therefore of the recovery voltage at the HV-side of the transformer can be expected [16], figure F.1 of
[1]. This phenomenon leads to a considerable reduction of the voltage at the HV circuit-breaker,
typically resulting in a residual voltage of 80% to 90%; i.e. a sub-subtransient source voltage of 0.8 to
0.9 pu, in the first few hundred s after clearing the out-of-phase current. The effect is larger at larger
currents but is not observed for generators with fully laminated poles and a damper winding [16].
The amplitude factor of the RV is determined by the natural frequencies of each side of the circuit-
breaker and normally the natural frequencies differ substantially such that the components of the
transient recovery voltage at both sides of the circuit-breaker swing independently and their crests do
not coincide.
81
Fig. 5 Out-of-phase Recovery Voltage. Case (i)
Shortly before clearing the voltage at both terminals of the circuit-breaker pole is defined by:
Vcb = Es + (Eg-Es) * Xs/(Xd+Xt+Xs), where Vcb is the circuit-breaker terminal voltage, Eg is the
source voltage at the generator side and, Es is the source voltage at the net side. If Eg = -1.0 pu, Es =
+1.0 pu (full phase opposition) and Xd+Xt = 5*Xs then Vcb = 0.67 pu. The net side RV will swing
from 0.67 pu to about 1.0 pu (see former page and figure 5). With an over-swing of the voltage jump
corresponding to an amplitude factor of 1.5, a peak value of 1.17 pu is reached.
The transformer side will swing from 0.67 pu to -0.92 pu (assuming net side fpcf of 1.3 and 10%
depression) with an amplitude factor of 1.8, thus giving a peak value of 2.19 pu. For a net-side fpcf
of 1.1 (and 10% depression), the voltage will jump at the transformer side from 0.67 pu to - 0.77 pu;
with an amplitude factor of 1.8, a peak value of -1.92 pu is reached.
In order to estimate the crest value of the total TRV, the assumption is made that the peak at one side
coincides with the power frequency recovery voltage at the other side. In this case,
1. the peak at the net side 1.17 pu coincides with 0.92 pu (resp. 0.77 pu) at the step-up
transformer side, summing up to a TRV peak value of 2.1 (resp. 1.9 pu)
2. the peak at the step-up transformers side -2.19 pu (resp. -1.92 pu) coincides with 1.0 pu at the
net side, summing up to 3.2 pu (resp. 2.9 pu).
These peak values are higher than 2.5 pu, as specified in IEC 62271-100 for systems with fpcf = 1.3.
In figure 5, the wave-shapes on both sides of the first clearing pole are schematically given assuming
full phase opposition. Reducing the out-of-phase angle will shift Vcb from 0.67 pu towards 1.0 pu,
thus decreasing the over-swing at the net side but increasing the over-swing at the step-up transformer
side. Moreover, due to the lower out-of-phase current the generator will show less depression and this
leads to a higher residual voltage.
82
At an out-of-phase angle of 90 the out-of-phase voltage is 1.41 pu. Assuming no depression, a
reactance ratio of 5, k = 0.80 at the step-up transformer-side and fpcf = 1.3 at the net side, it can be
calculated that the peak-value of the TRV is 2.5 pu: a value which is recognised in the Standards. In
other words, for these specific assumptions, the Standards do not address out-of-phase angles in excess
of 90.
In a system with a floating neutral, or equipped with Peterson coils, the fpcf = 1.5 and the maximum
RV will be 3.0 pu. The total TRV for the same example case can reach 4.55 pu at full phase
opposition. Without depression, an out-of-phase angle of 75 gives a RV of 1.83 pu and a peak value
of the TRV of 3.1 pu which is close to 3.13 pu as given in the Standards (for systems with fpcf of 1.5).
For a circuit-breaker at the MV-side of the step-up transformer, the IEEE/ANSI Standard C37.013
[10] is applicable. In this Standard an out-of-phase angle of 90 has been taken as the basic assumption
to specify the TRV requirements. It has to be mentioned however that many utilities specify an angle
of 180; see for instance [8].
3.2 Case (ii)

When, during out-of-phase conditions, the equilibrium point (virtual short-circuit point) is somewhere
on the OH-line that connects the two systems going out of synchronism, protection systems will trip
the circuit-breaker. Whilst it is possible to install advanced and complicated out-of-phase blocking
systems to delay the tripping command until the beating out-of-phase angle is small, this is uncommon
and switching can normally occur over a wide range of out-of-phase angles. The TRV across the first
clearing pole is determined by the system parameters on the busbar side of the circuit-breaker and by
the line parameters at the line side. As the largest impedance will be on the line side, the largest
voltage excursion will also appear at the line side.
The out-of-phase current is, to a large extent, dependent on the out-of-phase angle and the length of
the OH-line. Due to the traveling wave effects, the TRV at the line side will exhibit a triangular shape
and its peak value can be calculated as twice the wave traveling time along the OH-line multiplied by
the RRRV (rate of rise of the recovery voltage). The traveling time is proportional to the line length,
but the RRRV shows a decreasing trend with increasing line length due to the decrease in out-of-phase
current (Ioop). Specifically RRRV = fpcf * Zeq * dIoop/dt where Zeq is the equivalent surge
impedance. Due to the influence of the source impedances of both systems, the amplitude of Ioop is not
inversely proportional to the line length. Therefore, the peak value of the line side TRV will still
increase with an increasing line length. This effect, however, becomes smaller for OH-lines with
longer lengths.
1/(Xs1+XL+Xs2)
0,06
0,05
40 kA
0,04
31.5 kA
0,03
20 kA
0,02
infinite
0,01
0
0 200 400 600 800 1000 1200
line length (km)
Fig. 6 Total admittance as function of line length
Figure 6 shows the total admittance of both systems and the interconnecting line as a function of the
line length, for different (but arbitrarily chosen to be equal at both sides) source impedances of the
systems; i.e. for 420 kV systems with a short-circuit power equivalent to short-circuit currents of 40
kA, 31.5 kA and 20 kA in comparison to infinite short-circuit powers.
83
In addition to the line side TRV, the busbar side TRV should be added. As Ioop is defined to be 25% of
rating in the Standards and is often less than this in reality (15%), the system side TRV can be
estimated to be 25% (15%) of the TRV associated with, for instance, T100. The peak value is then less
than 0.37 pu (0.22 pu). For an OH-line with a length of 100 km, the return traveling time will be
roughly 650 s, close to T2, as defined for T100. For a 420 kV/40 kA circuit-breaker, the peak value
of the total TRV will be close to 4.1 pu for Ioop = 25% and 2.5 pu for Ioop = 15% of the rated short-
circuit current.
In figure 7, the TRV peak values (line side) as a function of line length are shown for the example
above (figure 6). The out-of-phase currents are based on full phase opposition. As the TRV peak value
at the line side is proportional to Ioop, it is also proportional to sin ().
Line side TRV peak
2000
1500
40 kA
kV
1000 31.5 kA
20 kA
500
0
0 500 1000 1500
line length (km)
Fig. 7 Line side TRV peak value as function of length in a 420 kV-network
It can be concluded that the TRV peak values can be considerably higher than specified in the
Standards (857 kV @ 1335 s for a rated voltage of 420 kV), even when taking into account smaller
out-of-phase angles. For instance with a line length of 200 km and source impedances corresponding
to a short-circuit current of 20 kA, an out-of-phase angle of 120 will still give a line side peak value
of 1075 kV. Combined with a system side RV of roughly 100 kV this results in a total TRV peak of
1175 kV. A line length of 100 km under the same conditions will give 875 kV at the line side and 975
kV in total. The IEC peak value of 857 kV is reached at out-of-phase angles as small as 75 and 90 for
200 km and 100 km lines respectively.
Calculations and simulations for real networks show out-of-phase TRV peak values as high as 3.3 to
3.5 pu [9] or even 3.9 pu [19] for very extended networks (hundreds of km, low currents) and 3.0 to
3.5 pu for meshed networks (hundred or less km, relatively high currents).
4. DIELECTRIC STRESSES ACROSS OPEN CONTACTS

During synchronization the longitudinal voltages applied to open contacts vary from zero to 2 pu in a
periodic, beating pattern for periods of seconds or minutes. In case of frequent synchronisation, clause
2.3.2.4 of IEC 60071, part2 [13] recommends consideration of the occurrence of an earth fault during
synchronisation (at one side!), thus leading to higher longitudinal voltages: up to 2.5 pu for a short
time.
Clause 2.3.2.5 of [13] recommends careful examination of the probability of simultaneous occurrence
of circumstances that lead to temporary overvoltages. Examples include an earth fault with
consequential line tripping firstly at load side, load rejection with high overvoltage causing an earth
fault, load disconnection under heavy pollution conditions, or a failure of a circuit-breaker to trip a line
fault with a generator still feeding the earth fault. In such cases a careful system study is required.
84
Clause 2.3.2.2 of [13] indicates that full load rejection will lead to temporary overvoltages, which are
normally less than 1.2 pu for moderately extended systems but which could reach values up to 1.5 pu
for large extended networks and even more in case of (ferro)resonance. (Ferro)resonance, however,
should be avoided and mitigation measures are suggested (cl. 2.3.2.3 and 2.3.2.6). The longitudinal
overvoltages across the circuit-breaker open terminals are equal to the temporary overvoltages when
the rejected load was of a static nature. But, in case of generators the longitudinal overvoltage can
reach values up to 2.5 pu and in very extended systems even more. A power frequency longitudinal
overvoltage as high as 2.5 pu is also given in clause D.1.3.2. of IEC 60694 [11].
With regard to the dielectric requirements under synchronising operations simultaneously with a
substantial transient or temporary overvoltage, clause 4.2 of IEC 62271-100 [1] indicates that the
standard requirements may be insufficient and the application of the requirements as specified for
disconnectors across open contacts is recommended. In clause 4.2 of IEC 60694 [11] different
requirements for the longitudinal withstand voltage across open contacts for the safety function (eg.
disconnectors) and for the working function (eg. circuit-breakers) are specified for rated voltages
245 kV. The values given in column (2) of the tables 1a and 1b [11], applicable for rated voltages
245 kV, are used for the specification of the longitudinal requirements of circuit-breakers, while the
values given in column (3) are used for the longitudinal requirements for disconnectors. For rated
voltages 300 kV, the values of column (3) of the tables 2a and 2b are specified for the 1min power
frequency type test across open contacts of both circuit-breakers and disconnectors, however the
values of column (2) are accepted for routine tests. In the following table the power frequency short-
duration withstand voltages are reported for some rated voltages for comparison, including the
withstand voltages in pu. For rated voltages 245 kV, the highest class of insulation has been taken
from table 1a, and for 300 kV the values given in table 2a:
Rated voltage (kV) (2) 1min withstand (2) 1min withstand (3) 1min withstand (3) 1min withstand
(kV) +) (pu) +) (kV) ) (pu) )
24 50 3.61 60 4.33
72,5 140 3.34 160 3.82
145 275 3.28 315 3.76
245 460 3.25 530 3.75
420 520 2.14 610 2.52
550 620 1.95 800 2.52
550 710 2.24 890 2.80
800 830 1.80 1150 2.49
from table 2b: additional rated insulation levels in North America.
+)
Specified for longitudinal insulation of circuit breakers with rated voltage 245kV
)
Specified for longitudinal insulation of disconnectors (all rated voltages) and of circuit breakers with rated voltage 300kV
IEC-Standard 62271-203 Gas-insulated metal-enclosed switchgear for rated voltages above 52 kV

[17] (the previous Standard 60517), makes reference to these tables in IEC 60694 but for the highest
rated voltages different short-duration power frequency withstand voltages are specified:
Rated voltage (kV) (2) 1min withstand (2) 1min withstand (3) 1min withstand (3) 1min withstand
(kV) (pu) (kV) (pu)
420 650 2.68 815 3.36
550 710 2.24 925 2.91
800 960 2.08 1270 2.75
External and internal flashovers across the open contacts of EHV circuit-breakers have occurred in
operation during synchronizing of generating units (due to contamined wet insulators in live tank
circuit-breakers, due to failure of grading capacitor in dead tank breakers, etc.) or during the dead time
before line automatic re-closure. These events generally cause a busbar fault, and also explosions of
circuit-breaker poles. It is therefore necessary to specify the circuit-breakers to withstand with a
reasonable margin the over-voltages liable to occur during these manoeuvres and to preserve this
capacity in operation.
85
Some reported cases of circuit-breaker failures during synchronizing of generating units have been
caused by flashovers on contaminated and wet external insulation of the interrupting chambers of live
tank circuit-breakers, by failure of the grading capacitor in parallel with one of the contacts, or by
inadequately specified power-frequency withstand voltage of circuit-breakers across open contacts
eroded by aging or by other reasons. Rare flashovers across the open contacts of line circuit-breakers
during the dead time before the automatic re-closure have been reported to be caused by multiple
lightning strokes in absence of surge arresters or of special protective air gaps at the open line terminal
[18]. Figure 8 shows a special protective gap shaped such as to minimize the influence of polarity and
wave shape of LIs and SIs on flashover voltage and to provide a time to flashover shorter in the gap
than in the protected open circuit-breaker. For decades, in Italy, there is very good service experience
with the application of these special protective gaps.
Fig.8 Special protective spark gap fitted in

the line anchor insulator strings to
substation gantry; D = 1700 mm for 380
kV lines; D = 800 mm for 150 kV lines.
SI 50% flashover voltage=1040 kV(3 pu)
In live tank circuit-breakers the external insulation between terminals is not energized when the
circuit-breaker is closed. It is recommended that for the external insulation across open contacts of live
tank circuit-breakers used for synchronizing, the withstand voltages as specified to column (3) of the
tables 1a, 1b, 2a and 2b of IEC 60694 [11], should be withstood in a type test under wet test conditions
and also under representative artificial pollution conditions.
All the dispersed statements in the Standards support the view point that with respect to out-of-phase
conditions and synchronisation, circuit-breakers longitudinal dielectric withstand should be specified
to column (3) rather than column (2) for all rated voltages.
5. OTHER CONDITIONS LEADING TO HIGH TRV PEAK VALUES

When clearing single or multi-phase faults distant from the substation on an OH-line (instead of at a
short distance, as with short-line faults), the well-known triangular wave-shape of the TRV at the line
side will rise to high values, depending on the wave travelling time from the circuit-breaker terminal
up to the location of the fault and back. This phenomenon is known as long line fault (LLF) and has
been discussed in [14]. As the time to the peak value of the TRV is rather long, it is comparable with
the TRV for out-of-phase switching. Peak values of 2.4 pu have been reported [14] and LLF is a
subject of study for CIGRE WG A3.19.
Clearing faults in series compensated OH-lines leads to TRV values in excess of the values specified
in the Standards, due to the charging voltage on the series capacitor banks. Peak values of the TRV as
high as 4.6 pu (420 kV-system in Turkey) and 4.8 pu (800 kV system in Canada) could be expected
without certain countermeasures. By means of special MOSA with a low SSPL (switching surge
protective level) of 1.57 pu, Hydro Qubec manages to reduce the TRV peak value to 3.2 pu. In
Turkey, MOV parallel to the arcing chambers of circuit-breakers have been applied successfully.
Depending on the requirement of re-synchronisation by the circuit-breaker, the TRV peak can be
reduced to 2.5 pu or 3.0 pu. These solutions lead nevertheless to TRV peak values comparable with or
beyond those given before for out-of-phase conditions.
Although there is no real application of half-wave length lines (HWLL, 3000 km at 50 Hz; 2500 km at
60 Hz), a number of studies on over-voltages and TRVs have been performed for this interesting
technology for long distance bulk power transmission. Simulations show that clearing faults in HWLL
will lead to TRV peak values as high as 3.2 pu, again comparable with the TRV peak values
mentioned before for out-of-phase clearing [9].
86
Another switching phenomenon giving high TRV values is the de-energization of unloaded OH-lines
under high TOV (temporary overvoltages) conditions [15]. For the 800 kV system of Hydro Qubec
TRV peak values of 3.3 pu to 3.5 pu have been reported under such conditions; see figure 9b. [9]
Out-of-phase switching on series compensated OH-lines has not been addressed yet, but it is evident
that the electrical charge on the series capacitors will add to the peak value of the TRV.
Unfortunately, right at the moment of current clearing the voltage across the series capacitors is at
maximum value, unless the capacitors have been by-passed by the self-triggered or forced triggered
spark gaps. The situation is similar to clearing short-circuit currents. In modern series capacitors
metal-oxide varistors are installed in parallel to the capacitor bank. Such varistors limit the voltage
across the capacitor banks. Moreover special surge arresters connected phase-to-ground or varistors
across the arcing chambers of the circuit-breakers are applied, thus limiting the total TRV peak value
at clearing short-circuit currents and out-of-phase currents as well. The countermeasures for limiting
the peak value of the TRV at clearing short-circuit currents are also effective at clearing out-of-phase
currents; see figure 9a. [9]
a) Max. TRV stress due to out-of-phase clearing b) Max. TRV stress due to de-energization of unloaded line
under high TOV (p.u./600 kV)
Fig. 9: Maximum TRV stresses due to out-of-phase clearing and de-energization of unloaded line under high TOV
6. CONCLUSIONS
The Standards showed to be based on an out-of-phase angle substantially less than 180, despite the
fact that in many cases the angle will be random, ranging up to 180. For generator circuit-breakers
and special applications users already ask for out-of-phase angles of 180.
The RV in the Standards [1][4] is 2.0 or 2.5 pu respectively for systems with an effectively earthed
neutral or a non-effectively earthed neutral with a TRV of 2.5 and 3.13 pu respectively.
Both IEEE and IEC specify the RRRV of the TRV for out-of-phase switching to be lower than the
RRRV specified for T100, whereas higher values occur in the systems. The RRRV for out-of-phase
switching is considered to be covered by T30 (multi-part testing).
Standardised TRV is based on system conditions in the absence of an earth fault. For situations with
frequent out-of-phase switching and synchronisation, the Standards recommend to specify actual
TRVs in the system (taking into account tripping and blocking relays for out-of-phase conditions
when applied) and to adapt the requirements for the longitudinal dielectric strength accordingly.
Under rather normal system conditions (no earth fault, no temporary overvoltages), full phase-
opposition switching of a generating plant at the HV-side leads to TRV peak values in the range of
2.9 to 3.2 pu in systems with effectively earthed neutral and higher values for systems with un-
earthed neutral. The peak values of the TRV as specified in the Standards, cover out-of-phase angles
up to 90 or, in case of systems with unearthed neutral, even less (75).
For out-of-phase switching of OH-lines, calculations show peak values of the TRV from 2.5 pu (100
km length, Ioop = 15%) to 4.1 pu (100 km length, Ioop = 25%) and even beyond for longer line
lengths, under full phase opposition. The Standards cover out-of-phase angles up to 90 (line length
< 100 km) or up to 75 (line length < 200 km).
During synchronisation, a longitudinal power frequency withstand test voltage larger than 2.0 pu,
preferably 2.5 pu or even 3.0 pu, is a reasonable requirement for circuit-breakers used for that
purpose. The related auxiliary components, such as grading capacitors, MOVs, insulating materials,
external insulation, should be equally specified and tested.
87
Under out-of-phase switching conditions the first-pole-to-clear factor is determined by the neutral
status of the systems at both sides of the circuit-breaker, as shown by the double Neptune-scheme.
Depression of the generator source voltage has to be taken into account, unless the rotor is equipped
with fully laminated poles and a damper winding.
False synchronisation, mal-operation of protection equipment and erroneous switching operations by
operators [7], may lead to considerable damage. Re-strikes at clearing out-of-phase currents with
large out-of-phase angles will also lead to comparable consequences. Developments in modern
networks lead to a higher probability of the phenomena described: distributed generation leading to
large power transfers, systems separation on overloaded OH-lines, large power swings due to tripped
generation, etc.
7. RECOMMENDATIONS
Utilities have to look carefully for such situations and, when applicable, put forward the appropriate
requirements to the protection of involved equipment and switchgear.
The present Standards do not give users clear and adequate support in specifying TRV-values for
out-of-phase conditions, for clearing of fault currents flowing through series capacitors and for
dielectric withstand requirements under synchronisation conditions. The requirements should be
revised or more guidance should be incorporated to improve understanding.
Since increased TRV requirements may lead to increased costs of circuit-breakers, enhanced out-of-
phase requirements should be limited in their application or countermeasures to limit TRV should be
used [9].
8. BIBLIOGRAPHY
[1] IEC Standard 62271-100 (Am.2, 2006) Part 100: [10] IEEE Standard C37.013 (1997)
High-voltage Alternating-current Circuit-breakers Standard for AC High-Voltage Generator
[2] IEEE Standard C37.04 (1999) Circuit Breakers
IEEE Standard Rating Structure for AC High- Rated on a Symmetrical Current Basis
Voltage Circuit Breakers [11] IEC Standard 60694 (2001)
Rated on a Symmetrical Current Basis Common specifications for high voltage
[3] ANSI Standard C37-06 (1997) switchgear and controlgear standards
AC High-Voltage Circuit Breakers Rated on a [12] IEC Standard 60071-1 (1993)
Symmetrical Current Basis. Preferred Ratings Insulation Co-ordination, Part 1:
and Related Required Capabilities Definitions, principles and rules
[4] IEEE Standard C37-09 (1999) [13] IEC Standard 60071-2 (1996)
IEEE Standard Test Procedures for AC High- Insulation Co-ordination, Part 2:
Voltage Circuit Breakers Application guide
Rated on a Symmetrical Current Basis [14] Severe Duties on High-Voltage Circuit
[5] TRV-Networks for the Testing of High-Voltage Breakers Observed in Recent Power Systems
Equipment H.Hamada, Y.Kasahara, T.Shimato,
A.L.J. Janssen, P. Knol, L. van der Sluis K.Hirasawa, K.Suzuki, T.Yoshizumi
CIGRE SC 13 Session 1996, Rep. 13-205 CIGRE SC 13 Session 2002, Rep. 13-103
[6] TRVs and Fault Clearing Stresses in Extra- [15] CIGRE Technical Brochure 305 (2006)
High-Voltage Radial Networks Guide for the application of IEC 62271-100
E. Haginomori, e.a. and IEC 60694, Part 2
EEJ, Vol. 114, No. 4, (1994), pp 50-61 [16] Comparison of Generator Circuit-Breaker Stresses
[7] Distributed Generation in Relation to Phase in Test Laboratory and Real Service Condition
Opposition and Short-circuits I.M. Canay
A.L.J. Janssen, e.a. IEEE-PD, Vol.16, No.3, July 2001, pp.415-421
10th SCC Symposium 2002 in Poland [17] IEC Standard 62271-203 (2003)
[8] Contribution to Q. 3-15 of SC A3 Special Report Gas-insulated metal-enclosed switchgear for
L. Zehnder, M. Kriegel rated voltages above 52 kV
CIGRE Session 2006 Proceedings [18] Flashovers of open circuit-breakers caused by
[9] Long Distance AC Power Transmission and lightning strokes CIGRE TF 33.11.06
Shunt/Series Compensation - lectra 186, Oct. 1999, pp. 114-123
Overview and Experiences [19] Circuit-breaker Requirements for Alternative
Q. Bui-Van, F. Gallon, F. Iliceto, A.L.J. Janssen, Configurations of a 500 kV Transmission System
B. Middleton, M. Waldron J. Amon, S.A. Morais, e.a.
CIGRE SC A3 Session 2006, Rep. A3-206 CIGRE SC 13 Colloquium 1995, Brazil, Rep. 2.3
88
89

Changing Network Conditions AND System Requirements: December 2007

Uploaded by

Copyright:

Available Formats

Changing Network Conditions AND System Requirements: December 2007

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Changing Network Conditions AND System Requirements: December 2007

Uploaded by

Copyright:

Available Formats

335

CHANGING NETWORK CONDITIONS

The impact of distributed generation On Equipment rated above 1 kV

Changing Network Conditions and System Requirements

The impact of distributed generation

HV-equipment is equipment with a rated voltage above 1 kV

A summary of the findings, recommendations and conclusions of CIGRE WG A3.13s studies

Acronyms and abbreviations

2. Technology of distributed generators

3. Regulations and policy with respect to distributed generation

4. Network topologies related to distributed generation

6. Power electronic equipment

7. Possible impact on the Standards

8. Conclusions and recommendations

B. Changing Network Conditions and System Requirements [22]

D. Dielectric, Switching and System Requirements under Out-of-Phase Conditions,

In recent years, in a number of countries, distributed generation has evolved to a substantial or

Fig. 3 General scheme of WEC-controls, incl. DFIG [13]

Fig. 5 Double axis, aero-derivative gas turbine, also available as STIG

Table I: Rough indication of Critical Clearing Times

High speed micro-turbines need a power-electronic interface to be connected to the LV-grid

Fig. 6 Wind farm in the North Sea

3.1 Fault-ride-through requirements

Supergrid Voltage Level

0 1 1.2 2 2.5 3 Time (s)

Fig. 7 Examples of Fault-Ride Through requirements

3.2 Other requirements

Table II: Minimum time periods during which wind

Voltage (pu) note 1 Minimum time

V < 1,0 see Fig. 7

As illustrated in figure 8, DG in distribution, sub-transmission, and industrial networks leads

Fig. 8 Structural change of power flows in distribution grid

DG connected to faulty MV or LV distribution-feeders

DG connected to healthy MV or LV (radial) distribution-feeders in case of short circuits in an

As stated previously, operating in an island mode is rather exceptional as it is very difficult to

Fig. 9 European Energy and Network Policy - Priority Interconnection Plan

As traditionally passive distribution networks become active, an issue of importance is safety;

5.1 Out-of-phase conditions

tf t oop GEN 25 MVA TA=5.08 sec

Fig. 10 Simulation of 25 MVA synchronous generator (CHP) dynamic behaviour

4 || Cables 3x1x150 mm 2 2 || Cables: 25 m 4 || Cables: 25 m

110 kV 20 kV P, Q switched off

Fig. 11 Single line diagram of simulated MV-network

Fig.18 Recovery voltage RV as product of FPCF and voltage factor

Out-of-phase current for several

Fig.19 Ioop/Isc as a function of X1/X2 and out-of-phase angle

5.3 Short-circuit currents

Fig. 20 Saturation effect of current transformers due to the DC-component

Fig. 21 Windmills in the Netherlands connected to a separate 20 kV-network

The short-circuit contribution of double-fed induction generators (DFIG) is very special, as it

Distortion of voltages and currents in power systems is caused by power electronic

A feature of DG, certainly those of a conventional technology, is the ability to absorb

6.2 Filter banks

6.3 Filter bank switching

Another phenomenon to be considered is that at switching off of a capacitor-bank the short-

To summarise; filter-banks have to be switched by switchgear with a very low probability of

6.4 Other aspects

Figure 22: Sample frequency responses of VT (left graph) an CT (right graph)

Distribution networks are changing rapidly, both in a technical sense and in a