Final Report 2015 March B5-C6.26.Compressed
Final Report 2015 March B5-C6.26.Compressed
Final Report 2015 March B5-C6.26.Compressed
26/CIRED
Final Report
26.03.2015
Copyright © 2015
“Ownership of a CIGRE and CIRED publication, whether in paper form or on electronic support only infers right of use for per-
sonal purposes. Total or partial reproduction of the publication for use other than personal and transfer to a third party are
prohibited, except if explicitly agreed by CIGRE and CIRED; hence circulation on any intranet or other company network is
forbidden”.
Disclaimer notice
“CIGRE and CIRED give 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”.
Joint Working Group B5/C6.26/CIRED | Protection of Distribution Systems with Distributed Energy Resources 2
Erreur ! Utilisez l'onglet Accueil pour appliquer Überschrift 1 au texte que vous souhaitez faire apparaître
ici.
Convenor
CIRED Co-Chair
Members
Copyright notice
Figures presented in this report have been collected from the referenced documents
Table of Contents
Members of the Working Group ............................................................................................ 2
Table of Contents.................................................................................................................. 3
1 Preface .......................................................................................................................... 6
2 Scope ............................................................................................................................ 7
3 Glossary and Definitions ................................................................................................ 8
4 Background on Distributed Energy Resources and current practices ............................. 9
4.1 Analysis of CIGRE TB421 ....................................................................................... 9
4.1.1 Content and scope ........................................................................................... 9
4.1.2 Network structure ........................................................................................... 10
4.1.3 Islanding ........................................................................................................ 10
4.1.4 Standardized communication and adaptive protection ................................... 11
4.1.5 Interface protection ........................................................................................ 11
4.1.6 Connection schemes and protection concepts ............................................... 14
4.1.7 Future trends ................................................................................................. 15
4.2 Protection characteristics of Distributed Energy Resources .................................. 16
4.2.1 Basic considerations ...................................................................................... 16
4.2.2 Blinding and sympathetic tripping................................................................... 16
4.2.3 Inverter based short circuit contribution.......................................................... 18
4.2.4 Fault ride through capability ........................................................................... 19
4.2.5 Island capability ............................................................................................. 20
4.3 Review of current practices for distribution system protection ............................... 21
4.4 Country Specific Approaches ................................................................................ 25
5 Protection of Distribution System with Distributed Energy Resources .......................... 29
5.1 Impact of Distributed Energy Resources on distribution system protection ........... 29
5.1.1 Definition of Situation ..................................................................................... 29
5.1.2 Check of fault scenario for several fault locations .......................................... 31
5.1.3 Future Aspects............................................................................................... 32
5.1.4 Summary ....................................................................................................... 34
5.2 Recommended best practices for reliable island detection .................................... 35
5.2.1 Passive local based measurement schemes.................................................. 37
5.2.2 Active detection ............................................................................................. 38
5.2.3 Protection based on a communication network (Communication based transfer
trip schemes) ............................................................................................................... 39
5.2.4 Future trend and recommendations ............................................................... 40
1 Preface
The usage of Distributed Energy Resources (DER) in utilities around the world is expected to
increase significantly. The existing distribution systems have been generally designed for
unidirectional power flow, and feeders are opened and locked out for any fault within. How-
ever, in the future this practice may lead to a loss of significant generation where each feed-
er may have significant DER penetration. Also, utilities have started to investigate islanding
operation of distribution systems with DER as a way to improve the reliability of the power
supply to customers.
This report is the result of 17 months of work of the Joint Working Group B5/C6.26/CIRED
“Protection of Distribution Systems with Distributed Energy Resources”. The working group
used the CIGRE report TB421 “The impact of Renewable Energy Sources and Distributed
Generation on Substation Protection and Automation”, published by WG B5.34 as the entry
document for the work on this report. In doing so, the group aligned the content and the
scope of this report, the network structures considered, possible islanding, standardized
communication and adaptive protection, interface protection, connection schemes and pro-
tection concepts and future trends accordingly.
After this preface (chapter 1), the scope of the technical brochure according to the Terms of
Reference (TOR) is stated (chapter 2). Then, in the glossary, symbols and abbreviations
used throughout the technical brochure are listed (chapter 3).
The first main part of the report starts with a summary of the backgrounds on DER and cur-
rent practices in protection at the distribution level (chapter 4). This chapter contains an
analysis of CIGRE TB421, protection relevant characteristics of DER, a review of current
practices for distribution system protection and a summary of country specific approaches
for protection in distribution systems. These country specific protection approaches are given
in more detail in the Annex in an aligned format.
The second main part of the report describes issues of protection of distribution schemes
that include DER (chapter 5). It features the impact of DER on distribution system protection
and sums up recommended best practices for reliable island detection.
The third main part offers a proactive approach to technology trends (chapter 6), under con-
sideration of new applications of teleprotection and the use of communication (e.g. IP, PLC,
BPL…). Moreover, protection schemes for future distribution networks that include DER are
presented.
The report closes with a summary and an outlook (chapter 7), as well as a bibliography
(chapter 8).
2 Scope
The scope of this technical brochure is to study the impact of DER on distribution system
protection taking into account DER characteristics and the possibility of islanding operation,
and to provide the guidelines to protect distribution systems with DER. The following topics
are presented within the technical brochure.
The review and the recommendations are applied to the different types of existing medium
voltage and low voltage distribution systems.
AD Active Demand
AMR Automatic Meters Reading
BPL Broadband Power Line
COROCOF Comparison of rate of change of frequency protection
DA Distribution Automation
DER Distributed energy resources
df/dt frequency change rate protection
DG Distributed generation/ distributed generators
DNO Distribution network operator
DSO Distribution system operator
FDIR Fault Detection, Isolation and network Recovery
FRT Fault ride through
GOOSE Generic Object Oriented Substation Events
HAN Home Area Network
HV High voltage
IDMT Inverse Definite Minimum Time
IED Intelligent electronic devices
IP Internet Protocol
ITA Inverse Time Admittance
LoM Loss of mains detection
LV Low voltage
MV Medium voltage
NAN Neutral Access Network
NDZ Non detection zones
OC Over-current
PCCN Protections et Contrôle-Commande Numérique
PLC Power Line Communication
PSTN Public switched telephone network
RAN Regional Action Network
RES Renewable energy sources
ROCOF Rate of change of frequency
SV Sampled values
THD total harmonic distortion
TOR Terms of reference
U/I Under-impedance
WAN Wide Area Network
With an increasing number of distributed energy resources, basic assumptions for the design
and implementation of protection systems are complicated, and protection concepts need to
be adapted accordingly. Both CIGRE and CIRED are looking into these developments, and
as one example of the outcome of these activities CIGRE report TB421 “The impact of Re-
newable Energy Sources and Distributed Generation on Substation Protection and Automa-
tion” was generated. This report is aimed to provide further insight into the matter.
The usage of DER in utilities around the world has increased and is expected to increase
further especially with the realization of the move to smart grids. Distribution systems are
generally designed for unidirectional power flow and feeders are opened and locked out for
any fault within. However, this practice may lead to loss of significant generation in future
where each feeder may have significant DER infeed. Additionally, utilities have started to
look into islanding operation of distribution systems with DER as a way to improve the relia-
bility of the power supply to customers. However, the difference between short circuit levels
for cases where the distribution system is connected to the grid and while it is islanded, can
be significant. This may result in malfunctioning of overcurrent (OC) protection or other pro-
tection schemes. In addition, the “plug and play” DER will continuously change the short
circuit level and thus may impact today’s overall protection schemes. Furthermore, the short
circuit power contribution varies with the DER technology applied. Wind turbines contribute
less current when their internal protection (crowbar protection) is activated and power elec-
tronic interfaced DER do not contribute as much fault current as conventional synchronous
generation.
The working group used the CIGRE report TB421 “The impact of Renewable Energy
Sources and Distributed Generation on Substation Protection and Automation”, published by
WG B5.34 as the entry document and coordinated their outputs with ongoing B5.43 activities
“Coordination of protection and control of future networks”.
Though distribution level DER connections are beyond the scope of TB421, the impacts on
the HV system must always be assessed2. This technical brochure thus focuses on distribu-
tion level DER connections (and, to some extent, their possible impact on the HV system).
However, as TB421 also states, in order to further increase the security of supply, DER units
could be allowed to operate in an islanded mode in a meshed distribution network4. This is a
reason that ring operation and protection issues related to ring operation should be consid-
ered. If more active network management schemes are used in future distribution networks,
it will also possibly mean increased topology variations (including for example radial or ring
operation of MV feeders) in which protection principles and settings adaptation may be re-
quired. In Denmark (and in South Africa at 11kV and 22kV in some cases), even though the
distribution system can operate in ring configurations, they are almost always operated in
radial configurations. Ring operation is not currently allowed by the French DSO ERDF.
ERDF, as well as the DSO in Romania, where island detection and trip is mandatory, do not
consider operating in an islanded mode as a possible evolution of the protection scheme.
4.1.3 Islanding
With reference to the topic of islanding, TB421 mentions that there is an increased need for
high performance anti-islanding protection5. The greater capability to withstand external
faults without nuisance tripping should not affect the sensitivity for the islanding detection, so
a compromise between sensitivity for islanding detection and stability under external disturb-
ances is required6. Growing DER penetration levels of all sizes and connection voltages
causes an increased need for high performance anti-islanding protection7. This technical
brochure thus proposes recommended (best) practices for reliable islanding detection, e.g.
communication based transfer trip schemes and passive local measurements based
schemes.
1
Cigre working group B5.34 TB421, p. 12
2
Cigre working group B5.34 TB421, p. 24
3
Cigre working group B5.34 TB421, p. 14
4
Cigre working group B5.34 TB421, p. 109
5
Cigre working group B5.34 TB421, p. 14
6
Cigre working group B5.34 TB421, p. 25
7
Cigre working group B5.34 TB421, p. 108
This technical brochure thus considers both usage of IEC61850 in protection circuits as well
as adaptive protection schemes, because in the future both short-circuit and earth-fault pro-
tection settings of MV feeder Intelligent Electronic Devices (IED) may need to adapt to
changes in network topology resulting from increased utilization of active distribution network
management schemes (depending on the utilized protection scheme). To support improved
supply reliability, to deal with topology changes and to disconnect faulted sections rapidly,
directional OC, distance and differential protection with high-speed communication based
blocking schemes will be utilized increasingly in the short-circuit protection of future Smart
Grids. The required future performance for transmitting blocking signals and voltage and
current data from sensors could be achieved by utilization of IEC 61850 GOOSE and sam-
pled value (SV) services, with increased usage of wireless 4G technologies in addition to
fiber-optic based communication. However, the problem related to synchronization of SV
data and the cost of SV usage and 4G technologies needs to be considered.
This is very difficult to achieve for smaller DER and possibly not essential. Stability during
faults can be of less a concern (or not at all) for distributed generation of relatively smaller
8
Cigre working group B5.34 TB421, p. 96
9
Cigre working group B5.34 TB421, p. 15
10
Cigre working group B5.34 TB421, p. 16
11
Cigre working group B5.34 TB421, p. 47
12
Cigre working group B5.34 TB421, pp. 73 & 74
13
Cigre working group B5.34 TB421, p. 93
14
Cigre working group B5.34 TB421, p. 18
Another important issue identified in TB421 is that the interface protection must be coordi-
nated with the remote utility protection15. This technical brochure thus considers selectivity
and co-ordination of DER unit protection (islanding detection vs. FRT requirements) with MV
network (feeder) protection. This aspect is related to reclosing co-ordination which has also
been reported in chapter 5.5 of TB42116.
Concerning chapter 5.4 of TB421 “Interface protection requirements and settings”, there has
been some modification of the French legislation regarding FRT requirements for producers
with maximum power above 5 MW (Figure 4-1).
Figure 4-1: Modification of the French legislation regarding FRT requirements for producers with maxi-
mal power above 5 MW
Wind and photovoltaic generation over 1 MW are required in Romania to adapt to FRT con-
ditions according to the following diagram (Figure 4-2).
Figure 4-2: Requirements for FRT capability of wind and photovoltaic generation over 1 MW in Romania
15
Cigre working group B5.34 TB421, p. 27
16
Cigre working group B5.34 TB421, p. 89
Figure 4-3: Requirements for FRT capability of RES < 100kVA and connected to LV in South Africa
RES greater or equal to 100 kVA are required to adapt to FRT conditions according to the
following diagram (Figure 4-4) in South Africa.
Figure 4-4: Requirements for FRT capability of RES ≥ 100kVA in South Africa
Directional OC protection will most likely not work correctly in the future due to insufficient
starting currents. Distance protection with under-impedance starting (U/I starting) might
solve the problem, or most likely under-impedance starting. This would require the network
topology and short circuit capacity at the relay to be known in advance. Differential protec-
tion could solve many problems, but will have other limitations (incl. cost).
A detailed comparison (with benefits and drawbacks) of possibly different applicable protec-
tion schemes in chosen example configurations (e.g. those presented in chapter 5.3 of
TB421) could be interesting. These protection methods could be for example:
17
Cigre working group B5.34 TB421, p. 97
Table 5-2 of TB421 also mentions “dedicated lines” and “non dedicated lines” as the most
typical arrangements of DER connection. This technical brochure uses these configurations
as the basis for recommendations.
This technical brochure thus also deals with how the distribution networks become more
similar to transmission networks, especially with regard to protection.
One of the most important changes to the distribution network due to the massive integration
of DER is the bi-directional nature of fault currents. This implies the need to develop more
complex protection schemes for the distribution network, which are able to deal with several
contributions to the fault current similar to the situation in transmission networks. Neverthe-
less, it is not obvious that future distribution system protection should be the same as in the
transmission system. The power contributions from different ends (sources) will usually be
different compared to transmission networks, so distribution networks will have some specif-
ic features when compared to transmission networks. The future distribution network might
still require a different protection scheme to that of a transmission network
18
Cigre working group B5.34 TB421, p. 92
• Firstly, the location of DER is different, i.e. distributed rather than central. Thus, short
circuit contributions from DER originate from directions not necessarily considered in
conventional protection schemes. This issue is similar to unexpected load flows un-
der normal conditions and can result in phenomena such as “blinding” or “sympathet-
ic tripping”.
• Secondly, many DER are not directly connected synchronous or asynchronous ma-
chines, but coupled to the network via inverters. The magnitude of the short circuit
current of these inverters is usually limited to values not much higher than the nomi-
nal current to protect the inverter itself. Accordingly, the short circuit capacity of grids
dominated by inverter short circuit current sources is significantly lower than that of
grids with rotating machines of the same rating.
• Thirdly, this lower short circuit contribution is also connected to a different time char-
acteristic of the short circuit current. While rotating machines behave like a voltage
source under short circuit conditions, inverters act more like a current source limited
to nominal current, and the time characteristic of the current determined by the con-
trol scheme of the inverter.
Additionally, the capability to operate during and after faults in the system has not been a
strong requirement for decentralized generation – on the contrary, it was and is often re-
quired that DER immediately disconnect under faulty grid conditions so that the well-
established protection schemes can be maintained.
However, with a growing number of DER and an increasing importance of DER to provide
short circuit capacity both during and after the fault, the Fault Ride Through (FRT) capability
is of major concern.
Finally, some distributed generation connected to low voltage level may be hidden to the
network operator, i.e. not officially declared and announced by the users. If the amount of
hidden LV DER is high, its impact to LV level protection and safety may be critical.
However, with DER connected to the system, the short circuit current contributions may orig-
inate from different paths than the central connection point and may result in “blinding” and
“sympathetic tripping”.
Blinding is caused by the fact that DER fault currents have an impact on the voltage at their
connection point. Thus, while the overall fault current at the fault location increases due to
the DER fault current contribution, the fault current provided via the central connection point
decreases.
Figure 4-5: Fault currents without (left) and with (right) distributed generation (Index “D”), only positive
sequence system shown.
I FL (1) =
c ⋅ E (1)
=
[
c ⋅ E (1) ⋅ Z N (1) + Z D (1) ] > I F 'L(1) ,
Z L (1) + Z N (1) || Z D (1) Z L (1) Z N (1) + Z N (1) Z D (1) + Z D (1) Z L (1)
i.e. the overall fault current with DER contribution is higher, the fault current provided from
the central connection point is lower according to
Z D (1) c ⋅ E (1)
I FN (1) = I FL (1) ⋅ = < I F 'N (1) .
Z N (1) + Z D (1) Z L (1) + Z N (1) + Z L (1) Z N (1) Z D (1)
In addition to the effect of “blinding”, with a significant increase of DER, the short-circuit ca-
pacity of the transmission system will probably decrease because the number of large con-
ventional power stations that provide the high fault levels will be reduced [56]. This would
negatively impact the protection coordination in distribution networks with high DER penetra-
tion. [13]
In Figure 4-5, “sympathetic tripping” would be caused by the fact that the DER fault current
contribution
Z N (1)
I FD (1) = I FN (1) ⋅
Z N (1) + Z D (1)
may initiate a relay trip when the relay is placed in the connection path of the DER, even if
the fault is not downstream to the path. Sympathetic tripping may be overcome by the appli-
cation of directional relays.
The fault current contribution of DER is strongly technology dependent [55]. Distributed gen-
erators based on rotating generators (synchronous or induction machine) will produce a rela-
tively high current during a fault (a current level of about 200 to 400% of nominal current in a
few cycles after the fault inception); while generators interfaced through power electronics
will limit the DER current magnitude to a maximum of 1 to 2 p.u. during the fault [57]. There-
fore coordination problems during normal operation are less likely with DER interfaced
through inverters and with a strong main substation [13].
DER units coupled by full converters allow injecting negative-sequence current if required
[75]. This behavior would make the fault detection easier. Negative-sequence control of the
DER as described in [58] allows for reduction of the overvoltages in the healthy phases and
increase of the unbalanced current for easier fault detection. It can be expected that the fu-
ture grid codes will specify asymmetrical current injection.
Furthermore, strict over-current limits are applied in order to protect the converters. Accord-
ing to [59] these limits can reach 1,3 p.u. for stator reactive current and 0,4 p.u. for the line-
side converter reactive current. State-of-the-art balanced fast voltage control of converter-
coupled generation for unbalanced faults might impact network protection, either by hinder-
ing fault detection/clearance (impact on sensitivity of the protective system) or by triggering
undesired disconnection of the generation (impact on security of the protective system) due
to:
As a result of the need for secure operation of the power system, provided by protection
based on three-phase over- and under-voltage, over-current, distance (impedance, angle),
and differential protection relays or any combination of those, then state-of-the-art balanced
fast voltage control of converter-coupled generation for unbalanced faults requires further
investigation [58].
A further difference between conventional networks and DER connected networks is the
behavior of the fault transient. For instance, different transients generated by inverter con-
trollers could affect some relays, e.g. the direction determination [13].
Short circuit calculation programs normally use traditional generator models, with a voltage
source behind an impedance. These programs may not be able to give reliable results for
Moreover, DC networks or more and more use of convertors and invertors in the distribution
grid will become a trend in DER related networks
Traditionally, DER protection has been set to automatically disconnect whenever faults in the
network are detected. With increasing importance of DER with regards to supporting the grid
operation this approach is no longer feasible. Moreover, DER protection settings that auto-
matically disconnect units whenever frequencies above 50,2 Hz (or 49,8 Hz) are detected in
European grids, pose a serious threat to system stability in European grids as the total
amount of installed DER versus base load increased significantly.
New requirements for FRT capability have already been described in chapter 4.1.5.
Figure 4-6 shows a typical fault ride through capability curve. It is obvious that DER following
this requirement will have an impact on protection scheme applications and settings.
Moreover, the fault behavior of the network will change over time if the distributed generators
are switched or if the topology is altered to minimize the losses or prevent voltage problems.
In a microgrid, the change of fault behavior is very large when the microgrid switches from
the grid-connected mode to the islanded mode of operation. The inertia and short-circuit
capacity of these islanded systems is much lower. [13]. Permanent relay settings will be-
come less effective in some situations and methods for adaptive resetting of protection char-
acteristics may be needed.
In a microgrid, most of the sources are connected via power electronic converters. These
converters do not supply sufficient currents to operate current based protective devices in
islanded mode because they have been designed to limit the fault current. Therefore protect-
ing a converter dominated microgrid is a challenging technical issue under the current limited
environment [60].
As mentioned above, due to the low thermal inertia of semiconductor switches, inverters are
actively current limited and, because of their small fault current contribution, they lead una-
voidably to various problems that have to be considered by the protection system [61]:
• Characteristics of the inverters under fault conditions may not be consistent with the
existing protection devices;
• Throughout the whole microgrid, there may be different inverters with different char-
acteristics;
• Even in the case of an individual inverter, its basic characteristics may differ depend-
ing on its design or application;
• There may be difficulties in characterizing inverter behavior for short-circuit studies,
since this depends on the control strategy applied;
• There may be a significantly reduced fault current level when changed from grid-
connected to islanded mode of operation.
Based on the above, one of the most important issues is to ensure that the behavior required
from DER units, including fault-ride-through needs, is compatible with the developed LV mi-
crogrid protection system during faults in microgrid. In other words this means that when the
protection of a microgrid operated in island mode is designed, one of the most important
questions to answer is how converter based DER units will contribute to the fault current
[15][25].
This chapter gives a short overview of the similarities and differences of the current practices
for distribution system protection. An overview of the individual protection functions is pro-
vided by Table 4-1. In addition the country-specific protection schemes are shown in Appen-
dix A - Country specific protection approaches.
Despite a general common protection scheme, which can be extrapolated from most of the
contributions from countries, some general differences can be noticed. Here are the main
themes which can be source of disparities between each country’s practices:
• The functions developed within the protection scheme: for instance, some countries
allow operating under a certain level of fault current whereas others prefer to elimi-
nate every fault on the distribution network.
• The neutral treatment (isolated, low impedance, compensated).
• The network structure (meshed or non-meshed network).
• National regulatory legislation.
• The voltage level operated by the DSO.
• The kind of protection functions used in each protection scheme, distributed different-
ly on the network.
• Telecommunication facilities at the utilities disposal.
• Protection functions’ configurations (thresholds, temporization…), which depend on
many factors (quality commitments of each DSO / DNO, penetration rate of DER…)
Basically the protection function can be classified into two categories. Firstly there is short
circuit protection, which is supposed to prevent thermal and mechanical asset stress and
damage caused by short circuit currents. Second, there is system protection, which protects
power grid from inadmissible operating conditions.
In most cases short circuit protection against faults on the MV busbar is taken care of by an
over-current protection function within the transformer protection system. Usually over-
current protection is sufficient to protect feeders in radial networks, but some countries use
distance protection due to meshed networks (e.g. Austria, Germany, Spain and Denmark).
As a special feature some DSOs also use reverse interlocking. This works by the feeder
protection operating in a quicker tripping time for busbar faults, in case they do not detect a
network fault in the respective feeder.
With long MV lines with low short circuit current at the end of the line (typical for rural areas)
the transformer protection is not able to support backup protection for the feeders. This is
why in some countries a local backup protection with circuit breaker failure detection at each
affected feeder is used (e.g. Germany).
The short circuit protection of the DER is comprehensively taken care by over-current pro-
tection (in some countries directional because of meshed operation). Alternatively fuses
might be in place for low-power generators.
To realize the system protection, all countries have implemented different levels of over- and
under frequency protection as well as voltage protection. These protection functions discon-
nect the DER from the distribution network once a deviation from the operational parameters
is detected. However there are differences in the specific use of df/dt protection and ROCOF
respectively. While some countries make these an integral part of their system protection
(e.g. South Africa, Romania, Denmark and Australia), other countries do not use them be-
cause they have concerns of negative influence at future black-starts.
Another unique feature of some countries such as Austria and Germany is that an additional
undervoltage/reactive power protection disconnects DER that receive inductive reactive
power during faults in the transmission grid (e.g. wind power plants with induction ma-
chines).
Moreover, Italy makes use of further levels of frequency protection, which evaluate the zero,
positive and negative sequence voltage as an extra criterion.
A (automatically) detected islanded network is treated very differently. In some countries one
can find loss of mains (LoM) as additional protection functions (e.g. with impedance meas-
urement or voltage vector shift protection ) or a tele decoupling signal that trips the circuit
breakers of large DER at the decoupled MV feeder (e.g. France).
Yet in some countries such islanded networks do not cause the DER to be disconnected
provided that the frequency and voltage remain in the allowed bandwidth.
LV DER de-
neutral treat- grid op- MV busbar pro- MV feeder protec- MV DER overcur- MV DER decoupling treatment of islanded
Country coupling pro-
ment eration tection tion rent protection protection grid
tection
Transfer trip where load
U>, U<, f>, f<, ROCOF,
non- I> with reverse I>, Ie> matching is possible and
Australia solid earthed I>, Ie> voltage vector shift,
meshed interlocking AR other LoM protections
neutral displacement
may be unreliable.
meshed
Z< with reverse
Austria comp. or non- Z< I> U>, U<, f>, f<, Q-U
interlocking
meshed
isolated / com- I>,
non- U>, U<, f>, f<,
China pensated / low Ie> I>>, I>, Ie> I> U>, U<, f>, f<, ROCOF transfer trip
meshed ROCOF
impedance arc protection
non- U>>, U>, U<, f>, f<,
Denmark comp. I>directional I>directional I>
meshed ROCOF
I> with reverse
non- U>>, U>, U<<, U<, f>, U>, U<, f>, f<,
Finland comp./isolated interlocking, Ie>, I>, Ie>directional I>, Ie>, fuse
meshed f<, LoM LoM
arc protection
U0>, U>, U<, f>, f<, U< U>, U<, f>, f<, tele decoupling (MV),
low imp. or non- I>, Ie>, P0>,
France I>, Ie> I>, Ie> + FRT requirements + LoM (DIN VDE impedance measurement
comp. meshed I>directional
teleprotection 0126-1-1/A1) (LV)
meshed Z<, I> (backup) or
low imp. or Z<, f< & P> direc-
Germany or non- I>directional, I> I>directional U>, U<, Q-U U>, U<, f>, f<
comp. tional
meshed (backup)
U>, U<, Uen>, f>*, f<*,
non-
Italy comp. I>, U> I>, I>directional, I>, Ie> f>, f< *with voltmetric transfer trip
meshed
release: U0>, Ud>, Ui>
I>> I>, Ie> , some
Nether- Low imp./ iso- non-
cases with reverse I>>, I>, Ie> I>, Ie> U>, U<, f>, f<
lands lated meshed
interlocking
non- I> with reverse
Norway comp. /isolated Z< or I> I> U>, U<, f>, f<,
meshed interlocking
I>, I>>, I>>>, Ie>, U>, U<, U<<,
non- I>, I>>, I>>>, Ie>, Ie>>
Portugal low imp. Ie>> and Ie> direc- I>, I>> U<<<, U0>, U0>>, U>, U<, f>, f<
meshed and Ie> directional
tional f>, f>>, f<, f<<
I>, sometimes
low imp., comp. non- earth fault reverse I> or I>directional,
Romania I>, Ie> U>, U<, f>, f<, df/dt
or isolated meshed interlocking and arc Ie> or I>directional
protection
South solid/resistively meshed or I>, Ie>, arc protec- I>, Ie> (dir/non-dir) I>, Ie> or pilot wire transfer trip for specific
U>, U<, U0>, f>, f<, df/dt
Africa grounded non- tion or pilot wire relay- relaying network criteria
LV DER de-
neutral treat- grid op- MV busbar pro- MV feeder protec- MV DER overcur- MV DER decoupling treatment of islanded
Country coupling pro-
ment eration tection tion rent protection protection grid
tection
meshed ing
Solid earthed,
low imp., iso- meshed I>, Ie> or busbar I>>, I>, Ie>>, Ie>,
Spain lated, comp. or non- differential Z< (for meshed I>, Ie> U>, U<, U0> f>, f<, U>, U<, f>, f< Transfer trip
(very few instal- meshed (meshed MV) MV), 67
lations)
*with several levels and delay times
This chapter gives an overview of the relay protection settings used by the listed coun-
tries/companies in medium voltage distribution networks with DER. Many of the protection
schemes are examples and the intention of the list is to provide a state of the art overview of
relay protection. In practical situations adaptations of the settings are necessary. Detailed
description of the country specific protection practices are given in the Appendix “Country
specific protection approaches”.
The over-current protections schemes of the feeders with DER have starting currents in the
range 1,2- 1,67 of rated feeder currents. Coordination times of 0,3 seconds are typical. Dis-
tance protection (with over-current start) is used in Germany, Spain and Austria, and rec-
ommended in Norway. Directional over-current relays are used in Denmark and Romania as
alternative solutions.
The DER decoupling protection is primarily composed of under/over voltage and under/over
frequency relays. The settings here vary significantly and also depend on the DER ratings
and fault-ride-through requirements. Instantaneous under-voltage protection (U/t<<) in the
range of 0,4-0,8 pu referred to voltage at point of common coupling is reported.
Table 4-2: Overview of the relay protection settings used by the listed countries/companies
Due to current regulatory frameworks in countries and technical developments, the penetra-
tion of DER in distribution networks has increased continuously and it can be expected that
this growth will continue in the future. This huge increase has an important impact on the
protection schemes of the distribution networks.
Based on this principle structure, all questions regarding the protection – both the current
practice as well as any future recommendations – must relate to it.
Figure 5-1 shows an outgoing feeder of a typical MV distribution grid supplied from the bus-
bar of a HV/MV-substation. It can be a ring (loop) which is usually operated radially. This
single feeder (which can be subdivided by one or more switching stations) supplies a num-
ber of MV/0,4-kV-transformer stations for customers (public supply). It is also
• the grid connection to and the in-feed from DER-plants (e.g. small co-generation
plants, small windmills, …)
and
Larger wind farms (green) are usually directly connected to the busbar. It should be noted
that substations used exclusively for the infeed of wind farms are expressively not covered
by this CIGRE/CIRED-JWG report.
transformer
HV / MV
substation
MV
busbar
CB B
transformer station
MV / 0,4 kV G
for consumers
B wind farm
with direct connection
to substation
transfer station
with DG input G
(owner )
Switching
station
CB D CB A
D Legend:
HV … High Voltage level (typically > 35 … 150 kV)
MV … Medium Voltage level (usually 10 / 20 / 30 kV)
transfer station DG … Distributed Energy Resources
with DG input G …. Circuit breaker (CB) with protection
(owner )
transfer station
for private grid
private grid
(micro grid)
(micro grid)
incl. DG
G
transformer station
MV / 0,4 kV
for consumers
Figure 5-1: Principle structure of the “distribution system” covered by this working group
Assuming that the wind farm (green) as well as DER and have sufficient FRT capabili-
ties, with respect to chapters 4.3 “Review of current practices for distribution system protec-
and
Table 5-1: fault scenario for several fault locations (common practice / future aspects)
sufficient)
C Generally: a fault in the HV-grid should not cause a trip of the substation.
1. Fault C (e.g. fault location is within the transformer differential protection zone)
will be tripped by the transformer protection.
2. A) If the remaining grid is not able to keep in island mode, e.g. too less genera-
tion: f<- and/or U<-criterion will trip the DERs
B) If the remaining grid remains in stable condition unwanted islanding.
Remark: If the sample grid given in Figure 5-1 consists only of radial outgoing feeders (i.e.
without it being meshed) then the fault scenario of fault location A is equivalent to fault loca-
tion B.
All protection schemes where DERs are involved must be checked: There must be a large
enough fault current to start the relevant protective devices:
HV/MV Transformer
MV Feeders
DER facility
1. In case of a fault in the supply grid all DER will decouple because of their U/f-protection.
2. The borders of the remaining grid are given due to its neutral treatment:
a. The neutral impedance ZN is directly connected to the neutral of the supplying
transformer:
In case of a transformer fault (fault location C) the remaining grid (brown dotted)
will lose its neutral treatment. A continuation of the operation is only possible if
the earth fault protection of the remaining grid is adapted to the new situation.
transformer ZN
HV / MV
substation
MV
busbar
CB B
Figure 5-3: neutral treatment of the grid with neutral impedance directly connected to the star point of the
transformer
transformer
HV / MV
grounding
(earthing)
transformer
substation
MV
busbar ZN
CB B
Figure 5-4: neutral treatment of the grid with neutral impedance connected to a grounding transformer
A serious issue introduced by the current high penetration of DER is the so called “uninten-
tional islanding” phenomenon. “Unintentional islanding” happens when a portion of the distri-
bution network with installed DER is disconnected from the rest of the network. In such cas-
es, voltage problems may arise and the localized frequency may not be in synch with the
frequency of the main transmission network.
The probability of having an unintentional island is related to the amount of power generation
connected to the distribution network and to the local loads. If there is a balance between
dispersed production and load, the risk of unwanted islanding exists in cases where the
feeder breaker is opened.
The basic problem with unwanted islanding after system disconnection is that parts of the
network are energized whereas de-energisation is expected.
In general, instead of the term “unwanted islanding”, it’s more appropriate to talk about “un-
controlled islanding”. In fact, during the islanding operation the main electric parameters are
not under direct control of the DSO (voltage) or the TSO (frequency).
For safety reasons, European standard EN 50110-1 clause 6.2 Dead working summarized
below has five essential rules and does not allow operation of the network under such condi-
tions. For this reason no distinctions are usually drawn between detection of or protection
against the islanding.
The actual rules require clear identification of the work location. After the respective electri-
cal installations have been identified the following five essential requirements shall be under-
taken in the specified order unless there are essential reasons for doing otherwise:
1. disconnect completely;
2. secure against re-connection;
3. verify that the installation is dead;
4. carry out earthing and short-circuiting;
5. provide protection against adjacent live parts
Depending on the power flow conditions, there are two different levels for the uncontrolled
islanding: the first level is at the MV feeder level, the second level is at the HV/MV substation
level. Depending on these levels, different approaches are possible in order to detect uncon-
trolled islanding condition. For instance, in the HV/MV substation level, in case of a grid op-
erated in open ring, uncontrolled islanding detection is possible through the voltage signal
presence on the network side of the feeder with the open breaker condition.
In order to avoid the unintentional islanding and out of sync reconnection which can cause
large damages to the DG and the grid, the protection system must operate in a very short
time by disconnecting the generators in a time that is shorter than the time of the first auto-
matic reclosure.
For many reasons some portions of the network may be subjected to the risk of islanding: for
example a network portion can be disconnected due to a fault protection intervention. Each
country has a defined set of rules in order to achieve an adequate level of safety and reliabil-
• The DER must be disconnected if the voltage or the frequency are out of a contrac-
tual range;
• The DER must be disconnected if one or more phases of the prevailing network
(transmission network) are missing;
• In case of an automatic reclosure, the DER must be disconnected before the first re-
closure.
Up to now, according to the state of the art, there is no islanding detection system which is
recognized as really efficient and shared by all the countries. Nevertheless, here is the state
of the art.
• Passive protection;
• Active protection
• Network communication based protection
• Under and over frequency protection function (ANSI/IEEE C37.2 code 81);
• Under and over voltage protection function (ANSI/IEEE C37.2 code 27 and 59);
Frequency measurement is usually performed by phasor vector change, [62], which makes it
very susceptible to voltage variations like those arising from faults. Therefore, it is usual to
use the line-to-line voltage measurement instead of single phase voltage measurement. The
aim is to feed the algorithm with voltage with low transient component (there are more single
phase faults than phase to phase faults). The ideal way should be frequency measurement
for positive sequence voltage. Some manufacturers have a minimum operating time of about
80ms to prevent unwanted trips. This has proven to be a good practice according to the Por-
tuguese experience.
The problems presented above also occur with other frequency measurement algorithms like
“zero-crossing”.
This type of relay has been known to have some issues which arise from each individual
manufacturers implementations, [63]. The performance of ROCOF relays is related to the
chosen algorithm for frequency measurement (“zero crossing” or Fourier Transform), the
length of the measuring window (longer measuring windows have a better response to volt-
The change in power output from the generator causes a shift of the voltage vector. This
protection is based on measuring the period of the voltage, which is compared with the pre-
vious measure. In island operation, the duration of the period, which is proportional to the
phase, changes due to the unbalance between generation and load at the first opening. The
relay phase shift is sensitive to disturbances such as faults on other feeders or transmission
network transients and therefore it is difficult to coordinate with other protections.
The reverse reactive power flux protection is installed in the connection point between the
distribution network and the generator.
The reverse active power flux protection is installed in the connection point between the dis-
tribution network and the producer network.
The rate of change of the active power protection is used in addition to the under/over volt-
age and under/over frequency protection.
The operation occurs when there is a difference between the exported reactive power and
the reference value for longer than a settable time. To avoid unwanted tripping in case of
fluctuations of the source, the set interval is chosen greater than the duration of possible
fluctuations.
The major nodes are connected by a communication link. Via this link the devices interact in
order, for example, to disconnect a DER from the HV/MV substation.
The method is based on the comparison of frequency change rates in two or more point of
the network.
The method is able to discriminate the frequency variations due to loss of network from other
changes due to other causes.
A lock signal is transmitted by the transmitter relay to all relays located near the DER. The
transmitter relay transmits this signal if the frequency change is generated by events not
related to the loss of network.
The DTT is more reliable compared to the previously described local measurement based
methods (passive methods) because it is not affected by the local power balance and it is
also immune to the remote system events. However, it needs to be noted that the cost of
providing adequate communication media may be prohibitive in some cases. Additionally,
the network topology may require the transfer trip to be supplied from more than one location
which adds to the cost and complexity of the scheme.
In the past this problem was dealt with by intertripping of the distributed generators. With an
increasing number of generators involved, this is not practical in modern MV and LV grids
due to the need for low latency, redundant and highly available communication lines.
The preferred solution to prevent islanding networks is a reliable detection system at the
DER connection point. As showed, there are several approaches developed to achieve this
objective. To implement a really reliable detection, some sort of information about the net-
work topology and/or the switching condition has to be provided. One way is to use separate
communications with additional complexity. Another example could be a “grid alive signal”
modulated on the voltage at the MV busbar and detected by the DER.
Another option is using the communication links which are a part of future smart grid and
advanced meter scenarios: ping the meters and ask for a voltage reading.
Yet another option is to minimize the need for communication by measuring voltage at the
place where the switching is done and expand the logic in the protection device of the MV
feeder.
Today MV islands are typically unwanted, whereas larger HV islands may be used in grid-
restoration scenarios after system blackouts.
With the increasing frequency of natural disasters these days, and the growing energy de-
mand in developing countries with poor infrastructure, the need to retain microgrids in the
system (as opposed to unwanted islanding), can get stronger. To follow this approach, it
would be necessary to change rules such as redefining the DSO’s responsibility; and rede-
sign of technologies mainly for the generators that uses static converters (inverters).
Future European Grid Codes (especially ENTSO-E Network Codes – Requirement for Gen-
erators), currently in committee consideration within the European Union, can also have
huge impacts on protection schemes. Enlargement of the frequency range, ability to provide
static power regulation depending on the frequency, definition of specific Fault Ride Through
requirements for different kind of generation and fast reactive current injection during fault
are among the most important ramifications of these network codes.
The new communication technologies are providing new opportunities for the development
of the Smart Grids. In order to define the requirements for these technologies it’s important
to notice that protection schemes are also designed in order to protect the network from un-
solicited islanding (like in France, Romania, Spain and Italy).
In general it’s possible to distinguish between wired technologies and radio technologies.
• Wired technologies:
o Copper Pair communications technologies
o Power Line communications technologies
o Fibre Optic communications technologies
• Radio technologies:
o VHF/UHF
o TETRA
o WIFI
o ZigBee
o Z-WAVE
o WIMAX
o Cellular data services and Satellite
With the assumption that, in case of islanding, the protection system has to act in the mini-
mum possible time, the table below shows the appropriate technology to use in MV network.
With the assumption that, in case of islanding, the protection system have to act in a time up
to 100ms, the table below show the appropriate technology to use in LV network.
WiFi does not apply to long range links between DSO to HV/MV Substation (MV/LV Substa-
tion) and between HV/MV substations and MV/LV substations since for these links, very high
power is used, compared to other wireless technologies such as VHF/UHF radios. WiFi is
applicable in the scenario NAN of the architecture, especially in those areas where there are
high concentrations of nodes. In areas like cities where there are a moderate concentration
of nodes, deployment of a wireless mesh network based on WiFi can be a great solution.
WiFi is also applicable in the scenario HAN architecture.
Situation in Italy:
In Italy, pilot tests are underway using PLC technology on LV network. In MV/LV substation
a PLC transmitter is installed. The islanding condition is detected by the voltage presence
with a circuit breaker opening condition. In case of islanding of the LV network an alarm
message is sent to all LV producers that are connected to the line via a broadcast signal.
The producers are provided with a PLC receiver that is connected to the interface protection.
In case of an alarm being received, the protection switches the frequency thresholds to the
narrower band. Currently there are several issues concerning interference with smart me-
ters, and these will be fixed with an appropriate choice of PLC carrier frequency.
Situation in France:
• Optical Fibre, which seems to be the best solution concerning the transmission time
but has some issues from the administrative point of view,
• Microwave Transmission, which could be a cheap and simple solution for installation
close to the substation,
• Internet Protocol (IP), which is currently field tested.
Concerning the IEC 61850 Protocol, along with the development of Digital Protection and
Control Systems of substations ("PCCN"), more and more protection equipment are compat-
ible with IEC 61850. Along with its adaptive strategy to the market, ERDF allows this equip-
ment to communicate with each other using the IEC 61850 protocol. Indeed, both feeder
protections and transformer protections are used by ERDF with IEC 61850 protocol (in
around 100 substations), via gateways connecting each to the supervising network of the
substation.
The purpose of this section is to highlight in a few words the new challenges of distribution
network protection related to new grid code requirements, active network management
schemes and island operation. Both medium-voltage (MV) and low-voltage (LV) distribution
networks as well as both directly connected rotating generator based DER units and con-
verter interfaced DG units are covered. Also, some of the proposed protection schemes for
smart grids, found in current literature, for island operated distribution networks are re-
viewed.
In the forthcoming ENTSO-E network code (NC) for generators (RfG) [2], it has been stated
that islanding detection should not be based only on the network operator’s switchgear
opening/ closing position signals. Moreover, if high-speed communication is used as a pri-
mary islanding detection method, the passive local islanding detection method is still needed
as a back-up [1].
Larger non-detection zone (NDZ) and unwanted DER trips due to other network events (nui-
sance tripping) have been the major challenges with traditional, passive local islanding de-
tection methods based on frequency (f), df/dt, vector shift (VS) or voltage (U). [1]
If the number of DER units in distribution networks increases, as expected, in the future, the
possibility of achieving power balance in the distribution network will also increase. There-
fore, the risk of distribution system segments operating in the NDZ of the traditional passive
islanding detection methods will increase, too. In addition, the use of f, U and rate-of-
change-of-frequency (ROCOF) for defining DER units’ FRT requirements in the new grid
codes (as in [2]), to enable utility grid stability supporting functionalities from DER units, will
increase. [1]
Recent and forthcoming grid code requirements, such as the active power/frequency (P/f)
regulation during over-frequency for all DER units and during under-frequency for larger
DER units [2], [3] will enable DER units connected to MV and LV networks to control their
active power even after islanding. This means that frequency deviations are instantly cor-
rected and islanding may not be detected with the traditional, passive islanding detection
methods. Therefore, the use of the traditional parameters for reliable and selective islanding
detection may become even more difficult in the future than it is today. [1]
Due to the above-mentioned reasons, a new, future-proof, passive islanding detection algo-
rithm and scheme has been proposed in [4] and [5], which is able to detect very fast and
selectively islanding situations even in a perfect power balance without NDZ, and is also
applicable to different type of DER units. [1]
Traditionally, active network management and adaptive protection functionalities have been
developed and operated independently [6]. However, in the future increasing attention
should be paid to understand the level of active network management and protection func-
tions coupling to be able to create future-proof solutions for the Smart Grids [7]. [1]
Frequency
In NC RfG [2] for DER units (Type A and larger), it has been stated that with regard to fre-
quency ranges, a DER unit shall be capable of staying connected to the network and operat-
ing within the frequency ranges (some difference between synchronous areas) and time pe-
riods specified in Figure 6-1Figure 5-6. [1]
Figure 6-1. Frequency FRT (Nordic area) and support requirements for DER units (Type A and larger). [1],
[2]
ROCOF/df/dt
In NC RfG [2], it is stated for DER units (Type A and larger) that, with regard to the ROCOF
or df/dt withstanding capability, a DER unit shall be capable of staying connected to the net-
Voltage
In NC RfG [5] for Type B and larger DER units it has been stated that with regard to FRT
capability of DER units, each TSO should define a voltage-against-time-profile (low-voltage-
ride-through, LVRT, curve) as shown in Figure 6-2. [2], [1]
Voltage in [2] refers to the root-mean-square (rms) value of the positive sequence of the
phase-to-phase voltages at fundamental frequency in per units (p.u.). [2]
Figure 6-2. Voltage FRT requirements for DG units (Type B and larger). [1], [2]
Reactive power exchange and control between distribution and transmission net-
works
In ENTSO-E NC for demand connection (DC) [8], it has been stated that all transmission
connected distribution networks shall fulfill the requirements related to reactive power ex-
change and control. These include requirements like 1) the actual reactive power range
specified by the DSO shall not be wider than 0,9 power factor of the larger of their maximum
import capability, 2) DSOs shall have the capability at the connection point to not export re-
active power (at nominal voltage) at an active power flow of less than 25% of the maximum
import capability, and 3) TSO shall have the right to require DSOs to actively control the ex-
change of reactive power at the connection point as part of a wider common concept for the
management of reactive power capabilities for the benefit of the entire network. [8], [1]
Advanced automatic functions are also implemented in closed loop in ERDF control centers:
Automatic Fault Detection, Isolation and network Recovery (FDIR) help the operator to effi-
ciently handle faults on the network, minimize duration of outages and improve customers’
quality of service.
With active network management, the capacity utilization of lines can be improved, large
voltage deviations can be avoided, system losses and interruptions can be minimized. Active
network management requires more information (measurements) from different points in the
distribution network (Figures 6-3 and 6-4) as well as utilization of fast and cost-efficient
communication technologies and further development of standardization (IEC 61850 related
standards). Real-time information about distribution network status (voltage, frequency etc.)
is required for example during voltage and frequency deviations to create network supporting
active and reactive power commands for DER units. Information about distribution network
status for control and monitoring purposes will be obtained in the future increasingly from
sensors across the network through high-speed wireless 4G networks and optical fibers. [9],
[10]
Active network management may simultaneously affect the protection settings if for instance
network topology is changed. On the other hand, e.g. due to earth-faults in some network
location, topology may be changed and this may have an effect on active network manage-
ment functionalities such as voltage control or loss minimization. Therefore, dependencies
between active network management and protection functionalities require careful planning
and development to create future-proof solutions for future Smart Grids. In future, it is likely
that these different active network management functionalities like voltage control, island
operation coordination, minimization of losses etc. will be realized through centralized solu-
tions at HV/MV (MV level management by DMS/SCADA or grid automation controller or IED)
and MV/LV (LV level management by IED, RTU or MicroSCADA) substations (Figure 6-3).
Also an intelligent coordination hierarchy between management of MV and LV level active
zones will be essential from a total concept point of view. Centralized monitoring (including
proactive protection) and earth-fault locating as well as different events or measurement
reporting functionalities are becoming more and more important from an asset management
Figure 6-3. An example about some possible centralized functionalities at HV/MV and MV/LV substations.
[9], [10]
Figure 6-4. Smart Grid compatible IEDs with appropriate communication capabilities and new functionali-
ties will play key role in enabling future active network management and protection concepts. [9], [10]
or due to automatic load restoration when the normally open point (NOP) in a meshed distri-
bution feeder is automatically moved for load restoration purposes following a fault. To sup-
port improved supply reliability, to be able to deal with topology changes and disconnect
faulted section very rapidly, distance and differential protection with high-speed communica-
tion based blocking schemes will be utilized increasingly in the short-circuit protection of fu-
ture Smart Grids [11], [12]. In reality, the required future performance for transmitting block-
ings and voltage and current samples from sensors could be achieved by utilization of 61850
GOOSE and SV services and possibly with wireless 4G technologies. From MV feeder IEDs’
earth-fault protection point of view, it is essential that their settings and protection principles
can also adapt to changes in MV network earthing method e.g. when changing from centrally
compensated utility grid connected operation to isolated island operation. [9], [10]
Adaptation of MV feeder IED protection settings can be done by changing pre-defined set-
ting groups or by changing settings in real-time by central controller (e.g. grid automation
controller). In the case of islanding protection adaption of MV feeder IEDs could also be
based on local detection of CB status change or multi-criteria based islanding detection etc.
On the other hand, if the size of the intended island could also be adaptive i.e. dependent on
current power balance situation, then the MV feeder IED could also be “locally aware” of the
transition possibility to island operation, in that point, and thereby activate those protection
functions in IED which are needed for successful transition to island operation (e.g. high-
speed operation in large voltage dips, fault current through IED or not, detection of healthy or
faulty island). They could also be predefined as different setting groups which are activated
centrally by grid automation controller when needed. Therefore, it needs to be defined as
part of the proposed adaptive protection scheme how the logic related to it will be centralized
or de-centralized. [9], [10]
Microgrid protection must respond to both utility grid and microgrid faults. During utility grid
faults, protection isolates the microgrid from the utility grid as rapidly as necessary to protect
the microgrid loads (Figure 6-5). For faults during island operation, the protection isolates the
smallest possible section of the radial feeder to eliminate the fault. After isolation from the
utility grid (Figure 6-5), local generators are the only fault current sources in the electric is-
land and the fault current level depends on the types, sizes and locations of DER; but it is
generally lower than the fault current from the utility grid. [16]
Figure 6-5. Microgrid protection isolates the microgrid from the utility grid as rapidly as necessary to
protect the microgrid loads. [16]
In the development of the new protection scheme for microgrids, several issues must be
considered, for example
In addition, the developed protection scheme for microgrid must be supported by the tech-
nical choices made in the microgrid operation and control issues. Some of the key issues
related to the LV microgrid protection are briefly reviewed based in [17] from which more
detailed information can be found. The size and number of microgrid protection zones will
define the needed amount of protective devices (PDs) for microgrid protection. The size of
microgrid protection zone must be such that it fulfils the requirements of customers and at
the same time is economically feasible. [15]
As stated by [18], the protection issues for microgrids cannot be properly resolved without a
thorough understanding of microgrid dynamics before, during and after islanding. Specifically
directly connected rotating machines are very sensitive to lose stability in voltage dips
caused by faults in island operated microgrid and so they may jeopardize the stability of the
whole microgrid. For that reason, if there are directly connected rotating machines connect-
ed, protection should operate in islanded microgrid rapidly for every kind of fault and e.g. if
microgrid customers have fuses with high rated currents, there is a risk that customer protec-
tion may operate too slowly in island operation due to low fault currents, which in turn may
cause instability in island operated microgrid after fault clearance. [15]
In cases where overcurrent based protection is utilized during island operation, protection
and control functions of IEDs in microgrid may need real-time information about network
topology, the status of DER units (on or off), state of charge of storage systems, and also
number and size of loads connected to the microgrid. These conditions have to be updated
and checked continuously in order to guarantee that protection settings are suitable for cur-
rent configuration [14]. [15]
Based on the above and as mentioned in [18], the high-speed operation of the protection
devices is very crucial for reliable operation of the microgrid protection system. Utilization of
high-speed telecommunication is expected to be an essential part of future smart grid pro-
tection system to achieve fast and selective protection both in grid connected and islanded
modes of operation. The same communication protocols and standards used in HV / MV
network can be applied directly to the LV microgrids. However, due to the smaller scale of
LV microgrid‘s, the costs of protection devices must also be lower than the cost of devices
used in the HV/MV network. [15]
One important issue, which is required to enable stable transition from normal grid connect-
ed operation to island operation, is coordination of IED protection settings with DER unit FRT
requirements (especially low-voltage-ride-through, LVRT). To prevent unnecessary tripping,
faulty lines should be disconnected first by the protection system and only after that should
the DER units be disconnected according to their FRT / LVRT profile. Fast operation of pro-
tection is required and emphasized if there are several protection zones. To achieve this
communication based protection methods and schemes are often required to ensure selec-
tive operation [19]. Some further discussion about issues related to protection of microgrids
can be also found in [13], [15], [20]-[24].
DER Unit Fault Behavior and Effect on Protection Scheme during Island Operation
It is also important to ensure that the behavior required from DER units, including fault-ride-
through needs, during faults in microgrid is compatible with the developed LV microgrid pro-
tection system. In other words this means that when protection of island operated microgrid
is designed one of the most important questions to be resolved is how converter based DER
units will contribute to the fault current feeding. [15], [25]
Usually control mode change of one or more DER units connected to the distribution network
is required after changing from normal to island operation. Traditionally this means that un-
der normal operation DER unit is in active(P)/reactive(Q) power control and after islanding
the control mode is change to voltage(U)/frequency(f) control (or voltage/speed control).
However, control schemes which do not require changing after transition to / from island
operation have also been proposed. For example in [24], an enhanced control strategy was
proposed which improves the performance of a DER unit under network faults and transient
disturbances, in a multi-unit microgrid setting. The proposed control strategy does not re-
quire the detection of the mode of operation and switching between different controllers (for
grid-connected and islanded) modes, and it enables the adopted DER units to ride through
network faults, irrespective of whether they take place within the host microgrid or impact the
upstream grid. [24]
From the island operated LV microgrid protection point of view, it is essential to know how
converter based DER units will contribute to the fault current. Grid codes and standards for
smart grids with island operation capability are necessary for the development of future
smart grids protection solutions to reduce complexity and to avoid the need for too many
case specific alternatives. [15], [25]
6.2.2 Example Case of CIGRE / CIRED JWG – Protection Needed for Successful
Transition to Intended Island Operation
In following, the same example case of this CIGRE/CIRED JWG presented in Chapter 5
(Figure 5-1) has been used to define the protection needs (functions, time selectivity) when
intentional island operation and especially successful transition to island operation is consid-
ered.
From island operation perspective (in addition to grid code FRT requirements from DG units
during normal grid connected operation) it is required that the wind farm (green) as well as
DER and have sufficient FRT capabilities. Figure 6-6 shows possible intended islands
and MV feeder short-circuit protection at CB1-CB4 (Figure 6-6) which is assumed to be di-
rectional like the earth-fault protection. In the following, the time selectivity issues (Figure 6-
7) are discussed, with different fault scenarios (faults A-E in Figure 6-6), regarding success-
ful transition to island operation.
In Figure 6-7 protection time selectivity issues, general time delay setting principles and the
role of high-speed communication when a) islanding is not allowed and b) islanding is possi-
ble are shown. Figure 6-7 also illustrates the role of high-speed communication based inter-
lockings/blockings (as well as transfer trip based islanding detection) in reliable and selective
In Figures 6-6 and 6-7 the idea is that the possible operation principles of directional short-
circuit protection in forward direction can be
1) Directional overcurrent protection with fixed time delay (and high-stage / low-stage
settings)
2) Distance protection with fixed time delay (in forward direction)
Similarly in Figures 6-6 and 6-7 the possible operation principles of directional protection in
reverse direction (for intentional islanding) can be
1) Undervoltage with fixed time delay (and high-stage / low-stage settings) AND current
direction detection in reverse direction. Function pick-up/start is only based on un-
dervoltage (i.e. not in overcurrent, because fault current levels of inverter-based DER
units can be fairly low as discussed in previous chapters)
2) Distance protection with fixed time delay (in reverse direction)
From Figures 6-6 and 6-7 it can be seen that selectivity problems can be possible if commu-
nication based interlockings/blocking for example are NOT used (Figure 6-7a)), because
coordination between LVRT curve of DER units (defined by grid codes) and required time
differences between CB2 and CB3 in forward direction may be hard to be achieve. This nat-
urally depends from the number of consecutive protection zones and the allowed time differ-
ence between operation time delays of CB2 and CB3.
Transition to intentional island operation IS only possible (Figure 6-7b)) if active and reactive
power unbalance at CB1, CB2, CB3 or CB4 is small enough (or enough, rapidly controllable
active and reactive power units exist in the possible island (1.-4.b in Figure 6-6). before pro-
tection start/operation of CB1-CB4 in reverse direction). If this is not the case transition to
island operation should not be allowed. Here it is worth mentioning that the new grid codes
(like forthcoming ENTSO-E Network Code RfG) enable / support transition to intentional is-
land operation because of P/f-droop control requirements of DER units during overfrequency
situations (underfrequency based load shedding schemes could have similar kind of effect)
and possibly also due to voltage control (Q/U-control) requirements.
Figure 6-6. Possible intended islands 1.-4.b (see also Figure 6-7).
Figure 6-7. Protection time selectivity issues, setting principles and the role of high-speed communica-
tion when a) islanding is NOT allowed and b) islanding IS possible (see also Figure 6-6).
In the above discussion and in Figure 6-7, only short-circuit protection has been considered,
but naturally also earth-fault protection principles and settings have to be proper during both
normal and island operation. Therefore it should be noted here that after opening CB 2, CB 3
In the following, the protection operation principles during different faults A-E in the example
network shown in Figure 6-6 are described. It is also assumed that high-speed communica-
tion is assumed to be available/possible and islanding IS possible (Figure 6-7b) 2. Time se-
lectivity (with communication) if power generation / consumption are close to each other be-
hind the possible island connection point CB.
- CB1 will operate in reverse direction and disconnect the LV microgrid intentionally
from the utility network
o If active and reactive power unbalance at CB1 is small enough (i.e. stable
transition to island operation is possible) as stated above
o CB1 could also sent signal to LV microgrid DER units to change their control
mode etc. after operation
- CB 2 will operate in forward direction
o CB 2 sends simultaneously interlocking signal to CB 3 (and CB4) to prevent
their false operation and
o CB2 can also send communication based transfer trip (faulty island) discon-
nection signal to DER unit
Wind farm and DER unit will remain connected according to LVRT curve of DER units
(Figure 6-7).
- CB1 will operate in reverse direction and disconnect the LV microgrid intentionally
from the utility network
o If active and reactive power unbalance at CB1 is small enough
o CB1 could also sent signal to LV microgrid DER units to change their control
mode etc. after operation
- CB2 may also operate in reverse direction and disconnect part of the MV feeder in-
tentionally from the utility network to island operation
o If active and reactive power unbalance at CB2 is small enough
o CB2 could also sent signal to DER unit to change control mode etc. after
operation
- CB 3 will operate in forward direction
o CB 3 sends simultaneously interlocking signal to CB 4 to prevent false opera-
tion and
o CB3 can also send communication based transfer trip (faulty island) discon-
nection signal to DER unit
Wind farm will remain connected according to LVRT curve (Figure 6-7).
- MV busbar fault
- CB1 will operate in reverse direction and disconnect the LV microgrid intentionally
from the utility network
o If active and reactive power unbalance at CB1 is small enough
- HV network fault => intentional islanding should not take place, MV and LV network
connected DER units should ride-through HV network faults and possibly also sup-
port HV network according to grid code requirements (for example by reactive power
injection)
o Possible indication about HV network fault could be sent from HV/MV trans-
former protection IED to MV feeder IEDs to indicate that in this case inten-
tional islanding is not possible/allowed
Instead FRT and HV network support of DER units is preferred
o However, if very sensitive customers (sensitive to voltage dips) are connected
to MV or LV network then similar actions could take place as described above
for fault
The adaptive feature was also proposed in [14] for the protection of microgrids; where direc-
tional overcurrent relays are reconfigured by a remote central unit in case of grid separation
(islanding) or grid reconnection to consider for the change of short-circuit level. The central
unit is constantly aware of the network topology and the connected generators. [13]
Paper [27] has also presented an adaptive protection scheme for distribution networks that
include large number of DER and are able to dynamically change configuration. Based on
the real-time analysis of the current state of the distribution system, it adjusts the active set-
ting groups of the field protection devices according to the changes in the network topology.
A centralized controller installed at the substation with IEC 61850 based communications is
used for data integration from the medium voltage feeder and DER IEDs and performs the
control function. The decision to modify the settings is made based on the logic programmed
into this controller (substation computer or RTU) with IEC 61131-3 compatible programming
languages. [27]
The adaptive protection scheme presented in paper [27] together with microgrid control sys-
tem has also been developed and adapted for a real demonstration at Hailuoto island in Fin-
land. [28]
As a result of the different proposals, it has been concluded in [29] that the main challenges
with regard to a possible implementation of an adaptive protection system may be:
For example, in [30], a possible solution for fault detection in islanded microgrids based on
current symmetrical components has been presented. It proposes to use zero-sequence
current detection in the event of an upstream phase-to-earth fault (coordinated with unbal-
anced loads) and negative sequence current for phase-to-phase faults. [29]
In [31], a pilot instantaneous overcurrent protection scheme has been briefly described,
based on two routines, which can perform instantaneous protection for local line and remote
bus-bar, regardless of the DER location. In [32], a communication assisted protection selec-
tivity strategy with three structural levels has been proposed, to be applied with voltage-
restrained directional overcurrent protection [29].
In paper [19], comparison between directional overcurrent and distance protection has been
simulated considering 3-phase short-circuit faults only. The simulations showed that distance
protection is able to selectively distinguish between faults in the MV and LV networks. Fur-
ther, based on the simulations and the analysis, it was stated that overcurrent protection
cannot be used if fault current levels within the island are close to the maximum load current.
Therefore, distance protection may be suitable to be used in island operation. However,
when the scope is expanded to cover also two phase and possibly also earth faults the via-
bility of distance protection must be further studied [19]
In [41] the same authors as in [39] also proposed to utilize the total harmonic distortion
(THD) of voltage to improve the protection system in microgrids with inverter-interfaced DER
units, for ground faults. After identifying the type of fault by monitoring the variation of the
fundamental frequency (50 Hz), the voltage THD of different feeder relays was analyzed in
order to determine the faulted zone. [29]
In order to avoid the difficulty of the previous methods associated with detecting the oscilla-
tion waveform of the voltage variation, instead of using the voltage magnitude, [42] propose
to use only voltage positive sequence. In [43], it has been claimed that a distinction among
the three fault types can be made only considering the positive/direct and negative/inverse
sequence voltage components, without using the zero sequence / homopolar information.
Reference [44] proposed a very similar approach by determining the fault occurrence and
the fault zone, based on a busbar voltage measurement and its subsequent transformation
from abc coordinates to dq coordinates. [29]
In [45], the reduction in system voltage has been also used to implement the under-voltage
back-up protection scheme for current differential protection.
- Minor differences in voltage drop among the relays located at both ends of short lines
lead to protection operation failures, due to reduction of the voltage gradient
- Problems with practical application of some of these methodologies, as well as with
possibly needed communication infrastructure, when high number of DER units are
present
- Methods may be strongly dependent on the network architecture and on the defini-
tion of “protection zone” for the relay associated with each generator
- Problems in detection of high impedance faults.
The use of differential protection for microgrids with low fault current level has been sug-
gested also in [23], [46], [47], [48] and [49] to protect inverter-dominated microgrids. Howev-
er, differential protection might be expensive since protective devices must be placed on
every line segment of the network. Therefore [49] proposes to form protection zones consist-
ing of several line segments [13].
In addition, here it is worth mentioning that the topology of the sample network i.e. schematic
diagram in [45] (Figure 6-8) is chosen to be very well suited for differential protection also
during island operation. In general, current differential protection is not very well suitable for
protection of islanded part of distribution network having radial topology and many protection
zones (i.e. for protection of ‘last’ protection zones with open end / CB).
In addition, in references [15], [17], [20], [21] and [34] microgrid protection schemes based
on use of both voltage and directional overcurrent are analyzed. However, these schemes
are based on utilization of high-speed communication for interlocking / blocking purposes to
ensure selective operation of protection during island operation. In the following these pro-
posed schemes for LV and MV microgrids are presented.
LV microgrids
Different kind of protection methods and principles for microgrids has been proposed. One
problem in some of the proposed solutions for LV microgrid protection is that their applicabil-
ity is limited to microgrids with only converter connected DG units. Therefore, these solutions
may overlook others e.g. requirements on operational speed of protection to maintain stabil-
ity of LV microgrid with directly connected rotating machines after fault clearance.
According to [15] key fundamental properties required from the future LV microgrid protec-
tion systems include,
i) Adaptivity,
ii) Utilization of fast standard based communication (IEC 61850),
iii) Fast operation in deep voltage dips due to faults to maintain stability in healthy
part of LV microgrid,
iv) Fast operation to fulfill needs of very sensitive customers,
v) Selective operation in every kind of faults and
vi) Unnecessary operation of PDs and disconnection of DER units must be avoided.
In the following, two different proposed protection schemes for LV microgrids based on utili-
zation of both voltage and current (with direction detection) are shortly presented and more
details can be found from the references. The main difference of the proposed protection
schemes is that the first one (Proposed Scheme 1) relies on extensive use of high-speed
communication and the other (Proposed Scheme 2) is not based on use of communication.
Proposed Scheme 1
In references [15], [17], [20] and [21] the following scheme for protection of LV microgrids
has been proposed.
The main structural choices of the proposed LV microgrid protection system are summarized
in Figure 6-9. In Figure 6-9, type of protection devices (PD 1-4) chosen are presented. Here
it should be mentioned that when measurements from active and reactive power flow be-
tween utility grid and LV microgrid are needed during normal operation, also current meas-
urements needs to be included in PD 1 (Figure 6-9). However, from the proposed protection
system point of view the current measurements at PD 1 were not necessarily needed. Prop-
erties of the examined LV microgrid e.g. type, number and location of fault current feeding
Figure 6-9. Number of protection zones and type of protection devices (PD 1-4) needed in normal and
island operation of LV microgrid. [15], [17], [20] and [21]
After LV microgrid transition from normal to island operation, MMS will send state changed
signal from normal to island operation to different PDs of the microgrid to adapt to the
changed microgrid configuration (Figure 6-9). PD 1 is changed to be ready for future syn-
chronized re-connection back to utility grid and the re-synchronization requires that phase
voltages are measured from both sides of PD 1. Protection settings of PD 2s are changed to
the ones needed in island operation. To avoid malfunction of PD 2s, the protection settings
of PD 2s are not changed from normal to island operation settings before all possible transi-
ents and oscillations in voltages, currents and frequency are stabilized after transition to is-
Role of MMS is also important in power balance management of island operated microgrid
e.g. after fault F2 at LV feeder, MMS must immediately send, after operation of LV feeder
protection (PD 2), new set point values for those DER units that are still connected at the
healthy part of the microgrid or disconnection signal to some less critical customer loads.
During islanded operation of a microgrid, possible oscillations due to sudden changes in the
microgrid configuration need to be taken into account for the protection concept to achieve
selective protection and to avoid unnecessary tripping of protection. This could be done by
using communication based interlocking signals.
Fast real-time communication is needed for microgrid protection purposes between protec-
tion devices (PD 1 and 2) and also with master unit and DER units during microgrid island
operation. In addition MMS needs to be able to communicate in real-time with all these mi-
crogrid components as well as with customer loads. Communication should be based on
common standard like IEC 61850. Active microgrid components in the PCC of microgrid (PD
1, master unit and MMS) are also responsible for synchronized re-connection of the mi-
crogrid back to utility grid (Figure 6-9).
More information for example about the functions needed for LV microgrid protection in the
proposed scheme as well as details about the operation curves of PDs in the proposed LV
microgrid protection system during normal and island operation can be found from [15], [17].
Proposed Scheme 2
In [24] and [52], strategies for the coordination of protective devices, in typical radial distribu-
tion networks with DER, were proposed. Expanding on the idea presented in [24], in [33] and
[52] protection strategies based on microprocessor based relays for low-voltage microgrids
has been proposed. One of the salient features of this protection scheme is that it does not
require communications or adaptive protective devices. In addition, it is stated in [33] that the
proposed scheme is to a large extent independent of the fault current magnitude, the mi-
crogrid operational mode, and the type and size of the DER, subject to the modified relay
setting for the grid-connected mode of operation. [24]
MV microgrids
In [24] and [53], communication-assisted protection strategies have been proposed, which
should be implementable by commercially available microprocessor-based relays, for protec-
tion of inverter-based medium-voltage microgrids. Proposed scheme also includes a backup
protection strategy to manage communication network failures. In [24] it is also stated that
the proposed protection strategy is independent of the fault current levels; type, size, and
location of the DERs; and the operational mode of the microgrid. [24]
- Selectivity with LV network faults must be ensured during MV network island opera-
tion and therefore (in addition to selection of pick-up limit different from high load cur-
rent) traditional directional overcurrent based schemes are more difficult to apply dur-
ing island operation with more than one / multiple protection zones in the island op-
erated MV network
- Probably the most suitable protection schemes during island operation, especially
with inverter-based DER units having low short-circuit current contribution, are im-
pedance/admittance based schemes or schemes based on both voltage and current
(directional)
CIGRE TB421 “The impact of Renewable Energy Sources and Distributed Generation on
Substation Protection and Automation”, established by WG B5.34 served as entry document
for the work on this report. The joint working group aligned content and scope, the network
structures to be considered, possible islanding, standardized communication and adaptive
protection, interface protection, connection schemes and protection concepts and future
trends accordingly, even if the scope was quite different (distribution networks instead of
transmission networks in WG B5.34).
There are some differences in protection relevant characteristics of DER compared to “con-
ventional” energy resources and the possible impact on current and future protection
schemes.
Generally, the short circuit current contribution of distributed energy resources differs from
“conventional” energy resources in three ways: First, the location of DER is different, i.e.
distributed rather than central. Short circuit contribution of DER comes from directions not
necessarily considered in conventional protection schemes. Second, many DER are not
synchronous or asynchronous machines directly connected to the grid, but coupled via in-
verters. The magnitude of the short circuit current of these inverters is usually limited to val-
ues not much higher than the nominal current to protect the inverter itself. Accordingly, the
short circuit capacity of grids dominated by inverter short circuit current sources is signifi-
cantly lower than from rotating machines. Third, this lower short circuit contribution is also
connected to a different time characteristic of the short circuit current. While rotating ma-
chines behave like a voltage source under short circuit conditions, inverters act more like a
current source limited to nominal current, and the time characteristic of the current deter-
mined by the control scheme of the inverter.
Additionally, the capability to operate during and after faults in the system has been less of a
requirement for DG – to the contrary, it was and is often required from DER to disconnect
under faulty grid conditions so that the well-established protection schemes can be main-
tained. However, with a growing number of DER and an increasing importance of DER to
There are various similarities and differences of the current practices of distribution system
protection. This report focused on the main reasons for disparities (e.g. network structure,
national legislation or neutral treatment) and typical protection function for busbar, feeder or
DER protection. Moreover an overview of the individual protection function of different coun-
tries is given in a tabular listing.
The over-current protections schemes of the feeders with DER have starting currents in the
range 1,2- 1,67 of rated feeder currents. Coordination times of 0,3 seconds are typical. Dis-
tance protection (with over-current start) is used in Germany, Spain and Austria, and rec-
ommended in Norway. Directional over-current relays are used in Denmark and Romania as
alternative solutions.
The DER decoupling protection is primarily composed of under/over voltage and under/over
frequency relays. The settings vary significantly and depend also on the DER ratings and
fault-ride-through requirements. Instantaneous under-voltage protection (U/t<<) in the range
of 0,2-0,8 pu referred to the voltage at the point of common coupling is reported.
The principle structure of the “distribution system” which is covered by this working group
has been defined as follows: private grids are connected in different ways to a typical medi-
um voltage outgoing feeder of a substation for public supply.
By means of this defined structure the current practice in case of a fault is discussed. The
interaction between overcurrent/time-protection and the decoupling protection (consisting of
over-/under-voltage and over-/under-frequency-protection) is shown for different fault loca-
tions.
An important question is the source of the short circuit current in such structures: it can be
assumed that its main part, which is able to start the protection devices, is coming from the
grid. Depending on the contribution of the short circuit current of the DER it can be possible
that a micro grid with DER will disconnect selectively because of the overcurrent/time-
protection. However, in any case it is common usage that the disconnection of DER will ap-
pear by means of the decoupling protection.
One further aspect is the neutral treatment of the grid. Under normal operation the grid
works with its designated neutral treatment. Depending of the fault location it can happen
that remaining parts of the grid will lose their original neutral treatment. In these cases it is
necessary to adapt the setting of the protection devices according to the “new” neutral
treatment. In other words: the borders of the remaining grid are determined because of its
neutral treatment.
One of the most important changes caused by the massive integration of DER into the Dis-
tribution network is the bi-directionality of fault current (if existing sufficiently). This implies
The high penetration of DER is also changing the design paradigm of the Medium Voltage
networks. Initially the direction of the active power was from the high to the low voltage level,
but due to the increasing generation connected to the medium voltage (low voltage) network,
this assumption is no longer true.
New issues are arising concerning the new scenarios. One of them is that of “unintentional
islanding”. “Unintentional (or uncontrolled) islanding” happens when a portion of the distribu-
tion network with a relevant DER is disconnected from the rest of the network. In such case,
voltage problems arise and frequency is no more synchronous with the frequency in the
transmission network.
Depending on the power flow conditions, there are two different levels for the uncontrolled
islanding: the first level is at the MV feeder level, the second level is at the HV/MV substation
level.
Different approaches are possible in order to detect a non-controlled islanding. To avoid the
unintentional islanding the protection system must operate very quickly by disconnecting the
generators in a time shorter than the time of the first automatic reclosure.
Each country has defined a set of rules in order to achieve an adequate level of safety and
reliability of the power system. In general the system must provide protections that avoid
unwanted islands by disconnecting the generation units connected to the network portion of
an island as quickly as possible.
Up to now, there is no islanding detection system, especially for low voltage networks, which
is recognized as really efficient and used by all the countries. There are passive and active
techniques to detect an unintentional island; furthermore a communication network can be
used if present.
Future European Grid Codes can have huge impacts on protection schemes, in particular:
enlargement of the frequency range, ability to provide static power regulation depending on
the frequency, definition of specific Fault Ride Through requirements for different kinds of
generation and fast reactive current injection during fault.
The new communication technologies are providing new opportunities for the development
of the Smart Grids. Two technologies are being considered: wired technologies and radio
technologies.
With the assumption that, in case of islanding, the protection system has to act in the mini-
mum possible time, the appropriate technologies to use on MV networks are Fiber Optic
(currently field tested in France), Copper Pair, VHF/UHF radio (cheap and simple solution for
installation close to the substation) and Public Data service. For LV networks, PLC technolo-
gy (currently field tested in Italy and also used in Spain), Wifi (if the system is cyber secured)
and Public Data service seem to be the most appropriate technologies. However, it is highly
required to use reliable communication media for protection.
Concerning Substation Automation Systems, more and more protection equipment is com-
patible with the new protocol IEC 61850. For instance, both feeder protection and transform-
Future distribution network protection, both in MV and LV networks, will meet increasing
amount of challenges due to new grid code requirements, active network management
schemes and intended island operation.
The trend in new grid codes is to require FRT capability from DER units and possibly also
allow island operation. Therefore, reliable islanding detection will be increasingly important in
the future, to be able to make the correct decisions, like for example change the setting
group of IEDs or change the control principles and parameters of DER unit.
Active network management may simultaneously affect the protection settings if for instance
network topology is changed and protection adaptation may be required. On the other hand,
e.g. due to earth-fault in some network location, topology may be changed and it may have
an effect on active network management functionalities such as voltage control or loss mini-
mization. Therefore, dependencies between active network management and protection
functionalities require careful planning and development to create future-proof solutions for
future Smart Grids.
During intended island operation high fault currents from the utility grid are not present. Be-
sides, most of the DER units connected to microgrids will be converter interfaced and have
limited fault current feeding capabilities. Due to that, during intended island operation, con-
ventional distribution network protection schemes cannot guarantee correct operation of pro-
tection in terms of detection time, sensitivity and selectivity.
It is also important to ensure that the behavior required from DER units during faults in island
operated microgrid is compatible with the applied microgrid protection scheme. From the
protection scheme perspective, it is essential to know how the converter based DER units
will contribute to the fault current.
Several schemes have been proposed for protection of intentional island operated mi-
crogrids in current literature. For example schemes that are based on directional overcurrent
or symmetrical components, distance protection, voltage, current differential or combination
of both voltage and directional overcurrent functions.
In general, it can be stated that selectivity with LV network faults must be ensured during MV
network island operation and therefore (in addition to selection of pick-up limit different from
high load current) traditional non-directional overcurrent based schemes are more difficult to
apply during island operation with more than one / multiple protection zones in the island
operated MV network. Probably most suitable protection schemes for intended island opera-
tion, especially with inverter-based DER units having low short-circuit current contribution,
are impedance/admittance based schemes or schemes based on both voltage and current
(directional).
8 Bibliography
8.1 Standards
[1] Laaksonen, H. “Reliable Islanding Detection with Active MV Network Management.” CIRED
Workshop 2014, Rome, Italy, June 11-12, 2014.
[2] ENTSO-E, 2013, “ENTSO-E Network Code for Requirements for Grid Connection Applicable
to all Generators (NC RfG),” 8 March 2013.
[3] CENELEC, WG DRAFT, 2012, Combined document CLC/FprTS 50549-1 and CLC/FprTS
th
50549-2, Version 15 of December 2012.
[4] H. Laaksonen, 2012, "New Multi-criteria-based Algorithm for Islanding Detection in Smart
Grids", IEEE PES ISGT Europe 2012, Berlin, Germany.
[5] H. Laaksonen, 2013, "Advanced Islanding Detection Functionality for Future Electricity Distri-
bution Networks", IEEE Transactions on Power Delivery, vol. 28, no. 4, 2056-2064.
A.1 Australia
Itransformer
Uen Transformer
VT UxE Protection
ILx Device
CT
OFF cmd
substation Busbar protection:
MV Zone interlocked bus blocking
busbar Ifeeder scheme.
OFF cmd
Feeder protection:
ILx I>> ≈ 400A
CT IE>> ≈ 240A
UxE Feeder ISEF>> ≈ 4A (sensitive earth)
VT Uen Protection Normal load: 240A
Device Emergency load: 360A
(tdiscr ≈ 0,4 s for old prot. and CBs,
tdiscr ≈ 0,3 s for new prot. and CBs)
transformer
station
MV / 0.4 kV
for Generator protection:
Generator: 415V/11kV, 1.25MVA
consumers
Transformer: Y/Y/D. 11kV star point
unearthed for parallel operation, and
earthed for islanded operation
U>/t 1,15·Uc / 5,0 s
U</t 0,80·Uc / 5,0 s
transfer f>/t 52,0 Hz / 5,0 s
DG Grid
station VT f</t 49,5 Hz / 2,0 s
Protection
MV UxE
Device VVS 10deg
for OFF cmd
DG ROCOF 0,2Hz/s, 0,5s
NVD 15%, 7,0s
I>> ≈ 80A
t>> = IEC SI, TMS 0,15
G IE>> ≈ 20A
CT tE>> = IEC SI, TMS 0,20
ISEF>> ≈ 4A
t>> < 10s
At new sites, the busbar is most often protected by a zone interlocked busbar blocking
scheme. At some sites, high impedance or low impedance busbar differential protection is
employed.
The distribution feeders are protected by Inverse Definite Minimum Time (IDMT) overcurrent,
IDMT earthfault (residual overcurrent) and fixed time sensitive earthfault (residual overcur-
rent) protection systems. At new sites the former 2 are duplicated. Auto reclose is enabled
on all feeders with any overhead component, and is set to single shot at 10 seconds. Many
feeders have line reclosers installed and the recloser utilises the same functions as on the
feeder, but with more recloses (up to 3 shots), and sometimes with an attempt at fuse sav-
ing.
Where a feeder has embedded generation which can exceed the zone interlocked busbar
blocking scheme pickup, a directional overcurrent relay needs to be installed to prevent the
generator fault contribution from blocking high speed operation of the busbar blocking
scheme for a busbar fault. Distribution substations are generally protected by fuses on the
HV side which range from 3,15 A to 100 A. Larger 1 MVA or 1,5 MVA distribution substa-
tions are protected by HV side circuit breakers.
Embedded generation is normally connected via a delta winding (on the distribution feeder)
side to ensure that sensitive earthfault protections will not operate spuriously. Embedded
generators which are not capable of matching/supporting the minimum load on the segment
of the feeder (any segmentation following protection device operation) are generally required
to have most of the following protections installed:
• Overcurrent
• Neutral displacement
• Rate of change of frequency
• Voltage vector shift
• Underfrequency
• Overfrequency
• Undervoltage
• Overvoltage
Embedded generators which are capable of matching/supporting the minimum load on the
segment of the feeder (any segmentation following protection device operation) are general-
ly required to have communication systems to the supply substation to ensure that the gen-
erator or generator infeed connection is disconnected from the network whenever the supply
substation circuit breaker opens.
A.2 Austria
transfer
Protection
station VT
Device
MV UxE
U/t & f/t
for
OFF cmd
DG
CT
G
ILx
Protection
OFF cmd Device
I>/t
ILx
CT
UxE Protection
VT Uen Device:
I/t
core IΣ U/t- and f/t-protection:
balance U>/t 1,15·Uc / < 0,1 s
CT U</t 0,70·Uc / 0,7 s
f>/t 51,5 Hz / < 0,1 s
f</t 47,5 Hz / < 0,1 s
transformer (Uc = voltage at PCC)
station I>/t-protection:
MV / 0,4 kV I>> ≈ 1,4·Ir,DG
for t>> = “0”
consumers
transfer
Protection
station VT
MV UxE Device
U/t & f/t
for OFF cmd
DG
CT
G
ILx
Protection
Device
OFF cmd
I>/t
A.3 China
main
transformer
HV / MV
Protection
MV protection of main trans-
Itransformer Device former:
Uen
VT UxE TP I>, 0,3s
ILx I>, 0,6s
CT
OFF cmd
substation
MV
busbar Ifeeder
OFF cmd
ILx
CT MV protection of Feeder
UxE Protection (FP):
VT Uen Device: I>>, 0s
FP I>, 0,3s
core IΣ I>, 0,6s
balance
CT IDIFF> or Z<(optional)
transformer
station
MV / 0,4 kV
for
consumers
transfer
Protection
station VT
Device Interface protection:
MV UxE
IP U>/t 1,15·Uc / < 0,1 s
for
DG U</t 0,80·UN / 2,0 s
OFF cmd ILx U</t 0,40·UN / 0,2 s
CT f>/t 50,2 Hz / < 0,2 s
G f</t 48,75 Hz / < 0,2 s
Anti-islanding, less than 2s
Protection
Device
GP
A.3.2.1 General
In China, DG is defined as a generation located in the vicinity of the user, which power can
be locally consumed on 10 kV (distribution system voltage) and below (220V/380V) system
that connect to the main grid, and the total installed capacity at individual POINT OF DG
CONNECTION is not more than 6 MW.
- Buchholz protection;
- differential protection;
- instantaneous overcurrent protection;
- (voltage controlled) time-delayed overcurrent protection.
For transformer with capacity less than 10MVA, differential protection may not be applied
under the assumption that the sensitivity of overcurrent protection is good enough.
Undervoltage and negative sequence overvoltage control element may be applied on over-
current protection for sensitivity purpose. Typical settings for voltage control element are:
Upp<0,7UN, U2>0,06UN.
MV (10KV) protection of main transformer include two stages (voltage controlled) time-
delayed overcurrent protection, with time delay of 0,3s and 0,6s respectively.
General
An interface protection relay shall be installed at the POINT OF DG CONNECTION with the
following functions:
Anti-islanding protection
For inverter type DG, which operating character determines that they can run continuously
under islanding, anti-islanding protection is essential to prevent islanding and protect the
maintenance staff and primary equipment. Anti-islanding protection shall be able to detect
the islanding state and disconnect DG from main grid quickly when islanding happens. Anti-
For synchronous or asynchronous generator type DG, which operating character determines
that they cannot run continuously under islanding, anti-islanding protection is not necessary.
While the clearance time of DG shall coordinate with interconnection line protection in order
to prevent asynchronous reclosing. The power system security and stability can be guaran-
teed suppose the fault clearance time coordinates with interconnection line protection.
Voltage protection
The settings of voltage protection for DG shall meet the requirements described in the fol-
lowing table.
Requirements on DG voltage
Voltage on POINT OF DG
Requirements
CONNECTION
U < 50% U N Maximum clearance time shall not exceed 0,2s (0,1s)
50% U N ≤ U < 85% U N Maximum clearance time shall not exceed 2,0s
85% U N ≤ U < 110% U N Run continuously.
110% U N ≤ U < 135% U N Maximum clearance time shall not exceed 2,0s
135% U N ≤ U Maximum clearance time shall not exceed 0,2s (0,05s)
Note:
1. UN is the rated voltage point of interconnection
2. Maximum clearance time is the duration from the instant that abnormal condition occurs to
the instant that DG stop outputting.
Frequency protection
The settings of frequency protection for DG shall meet the requirements described in the
following table.
Requirements on DG frequency
Frequency Requirements on fault-through ability
Inverter type DG: according to the lowest frequency permitted by inverter itself,
or requirement of control center.
<48Hz
Synchronous or asynchronous generator type DG: not less than 60s, or shorter
when necessary provided that the power grid stability is ensured.
48Hz-49,5Hz Not less than 10 min
49,5Hz-
Run continuously
50,2Hz
DG can run at least 2 min and has the capability to reduce active power output
50,2Hz-
following the command from control center, or stop supplying in 0,2s.
50,5Hz
Re-connecting of DG in outage is not permitted.
Stop supplying to main grid in less than 0,2s.
>50,5Hz
Re-connecting of DG in outage is not permitted.
Typical settings for frequency protection are: >50,2Hz, 0,2s; <48,75Hz, 0,2s.
Re-closure
Auto-reclosing function is equipped in the interface protection relay:
A.3.2.5 DG protection
General
Power transformer, inverter, synchronous or asynchronous generator of DG shall be
equipped with reliable relaying protection.
In addition, relaying protection for DG shall be able to detect short-circuit fault (including sin-
gle phase to ground fault) and phase open fault that occurs at main grid side, and disconnect
DG from main grid in this case.
Inverter type DG
At DG side of POINT OF DG CONNECTION, the following protection functions are neces-
sary:
- (directional) overcurrent protection;
- undervoltage / overvoltage protection;
- underfrequency / overfrequency protection;
- anti-islanding protection;
There are two types of anti-islanding protection for inverter type DG – active one and pas-
sive one. Schemes of active anti-islanding protection include frequency offset, active power
variation, reactive power variation, impedance change caused by pulse current injection, etc.
Schemes of passive anti-islanding protection include voltage phase angle jitter, 3rd harmonic
voltage variation, rate of change of frequency, etc.
Re-connecting
After disconnection caused by a disturbance, re-connecting of DG is not permitted until the
grid voltage and frequency return to normal limit. When the grid voltage and frequency return
to normal limit, DG which outputs at 380V can be re-connected automatically after a certain
intended delay, which is not less than 20s and shall be specified by control center; DG which
outputs at 10kV (6kV) - 35kV level shall be re-connected only if it is commanded by control
center.
A.4 Denmark
DG Protection (DGP):
According to grid code – See sec-
tion 4B
19
Communication between relays via 200V DC wires
20
Communication between the breakers if there is problem with upstream breaker
A.4.2.3 Line/Feeder
• 2 step directional relay
o Trips CB for 1,2·I for 0,4s (in country side) and 0,8s (in city area) in forward
direction
o Trips CB for 20·I for 20-30ms in forward direction
o Trips CB for 1,2·I for 1s in any direction when VT malfunctions
A.4.2.4 DG protection
• Generation unit protection
o Refer to grid code in section 4B
• Lines and cable protection
o Trips CB for 1,4·cable rating for 20 ms towards the DG side from PCC
o Trips CB for 1,4·cable rating for 0,5 s towards the utility side from PCC
A.4.2.5 Others
• Single phase faults need to be located within 2 hours
• Petersen coil is used to ground the system and it is short circuited to locate the fault
A.5 Finland
A.5.1.1 Typical Protection Scheme for Primary Substation and MV Network (Urban)
MV Neutral Grounding: Isolated
MV/LV
secondary
substation LV feeders
LV feeder protection:
Overcurrent protection with fuses
MV/LV
secondary
substation LV feeders
Generation installations of up to 50 kVA must be equipped with protection devices that dis-
connect the generation installation or isolated generation fed by the generation installation
from the public network if the feed to the network is cut off or if the voltage or frequency at
the equipment connection point deviates from the normal reported values. The set values of
protection are presented in following table, in which Ur means the normal nominal rated volt-
age of the distribution network. The system operator in Finland may deviate from the values
case by case. Before connection of the DG unit to network information concerning the instal-
lation must be delivered to the system operator e.g. information about the implementation of
protection to prevent isolated/island operation (method and operating time of Loss of mains
protection). [66]
If the generation installation disconnects from the network due to the functioning of the pro-
tection device, it can reconnect to the network only after the network voltage and frequency
have returned back to the limits permitted by the set values for protection and have re-
mained within these limits for a certain minimum period. This minimum period is 20 seconds
for installations connected to the network with an inverter and 3 minutes for other generation
installations. In addition to the equipment that fulfil the requirements set in [66], also equip-
ment < 50 kVA that fulfil the technical requirements set in German application guide VDE-
AR-N-4105, are accepted to be connected to the distribution network. [66]
The operating delays must be based on the generation installation’s ability to operate at un-
der- and overfrequency and at under- and overvoltage conditions. The operating delays of
protection must be agreed separately with the system operator of the connection point. In-
formation about the protection settings of the generation installation must be delivered to the
system operator of the generation installation as part of the documentation delivered on the
installation. The generation installation must be equipped with protection that prevents island
operation. A description of protection that prevents island operation must be delivered to the
system operator of the connection point, and the protection settings must be agreed on sep-
arately with the system operator of the connection point. [67]
No different recommendations are given for DG units connected via dedicated lines or con-
nected directly to the substation.
ENTSO-E grid code RfG [69] divides requirements for four type/size of DG units (Power
Generating Modules) i.e. Type A (DG units > 0,8 kW connected to voltage levels below 110
kV) and B, C and D:
In addition, ENTSO-E grid code RfG [69] further divides the requirements into the two cate-
gories. Category-1-requirements (exhaustively described by RfG) include: frequency ranges
(including limited frequency sensitive mode) and voltage ranges. Category-2-requirements
(not exhaustively described by RfG, specified in national level) include among others: reac-
tive power and FRT. Based on ENTSO-E grid code RfG [69] voltage related FRT voltage
One interesting issue in the future is to see that what impact CENELEC (EUROPEAN
COMMITTEE FOR ELECTROTECHNICAL STANDARDISATION) possible forthcoming
standards
1. CLC/FprTS 50549-1 “Requirements for generating plants larger than 16 A per phase
to be connected in parallel with a low-voltage distribution network” and
2. CLC/FprTS 50549-2 “Requirements for generating plants to be connected in parallel
with a medium-voltage distribution network”
may have on national grid codes in Europe when compared to the requirements set by the
forthcoming ENTSO-E grid code RfG [69].
G
DG unit
(<50 kVA)
connection to
LV network
A.6 France
In addition to classical phase and earth faults protections (OC, PWH, PAH, PVH).
AN
D
OR
There is therefore a correct coordination between Undervoltage and Fault Ride Through
requirements.
MV decoupling
Type H.1 Type H.2 Type H.3 Type H.4 Type H.5
protections
Functions
Teleprotection
Time = feeder
60 ms < Time < 100 ms Time = 1.2 s
trip delay + 0,5 s
LV
decoupling protections ENS DIN VDE ENS DIN VDE 0126-1-1 and DIN VDE
B.1
0126 0126-1-1/A1 VFR 2013-VFR 2014
A.7 Germany
OFF cmd
substation
MV
main bus
busbar Ifeeder MV protection of feeder:
OFF cmd Z< protection:
tend,directional = 0,9 s
ILx tend,nondirectional = 1,2 s
CT
CBF I>> ≈ 1,2·Ir,feeder max 80% Ifault,2p
UxE Protection Back-up I>/t-protection:
VT Uen Device MV: & tBack-up I>/t=0,9 s
Z< or I>/t
core IΣ & Back - up I>/t MV decoupling protection:
balance & & U/t & (f/t) U/t- and f/t-protection:
CT & UQ→&U< U>>/t 1,15·Uc / ≤ 0,1 s
U>/t 1,08·Uc / 1 min
U</t 0,8·Uc / 2,7 s
(Uc = voltage at PCC)
Reactive power low voltage pro-
tection:
Q&U< 0,85 Uc / 1,5s (MV CB)
signal line
transfer
station
MV
for
LV decoupling protection:
DG
Protection U/t- and f/t-protection:
Device LV: U>>/t 1,15·ULV / ≤ 0,1 s
G & U/t & f/t U</t 0,8·ULV / 1,5 s
U<</t 0,45·ULV / 0,3 s
f>/t 51,5 Hz / ≤ 0,1 s
f</t 47,5 Hz / ≤ 0,1 s
(ULV = Lowvoltage)
OFF cmd
substation
MV
main bus
busbar Ifeeder
OFF cmd
MV protection of feeder:
Z< protection:
ILx
CT tend,directional = 0,9 s
CBF tend,nondirectional = 1,2 s
UxE Protection
VT Uen & I>> ≈ 1,2·Ir,feeder max 80% Ifault,2p
Device: Back-up I>/t-protection:
Z< or I>/t tBack-up I>/t=0,9 s
core IΣ & Back - up I>/t
balance
CT
transformer I>/t-protection:
station I>> = various
MV / 0,4 kV t>> = 0,1s
for
consumers MV decoupling protection:
U/t- and f/t-protection:
U>>/t 1,15·Uc / ≤ 0,1 s
U>/t 1,08·Uc / 1 min
U</t 0,8·Uc / 2,7 s
(Uc = voltage at PCC)
transfer Reactive power low voltage protec-
Protection tion:
station VT Device MV:
MV UxE & U/t & (f/t) Q&U< 0,85 Uc / 1,5s (MV CB)
signal line
OFF cmd
substation
MV
main bus
busbar Ifeeder
OFF cmd
MV protection of feeder:
Z< protection:
ILx
CT tend,directional = 0,9 s
CBF tend,nondirectional = 1,2 s
UxE Protection
VT I>> ≈ 1,2·Ir,feeder max 80% Ifault,2p
Uen Device: &
Back-up I>/t-protection:
Z< or I>/t tBack-up I>/t=0,9 s
core IΣ & Back - up I>/t
balance
CT
transformer I>/t-protection:
station fuse
MV / 0,4 kV
for MV decoupling protection:
consumers U/t- and f/t-protection:
U>>/t 1,15·Uc / ≤ 0,1 s
U>/t 1,08·Uc / 1 min
U</t 0,8·Uc / 2,7 s
(Uc = voltage at PCC)
In MV grids with less short-circuit power, a second over-current protection relay and a CB-
failure protection may be needed at the feeders.
Protection scheme at the DER transfer station can be divided into three categories:
Short-circuit protection
Short-circuit protection of the generating plant is required for clearing of short-circuits near
the connection point. In addition, it serves as back-up protection in the event of faults within
the generating units and in the power grid. The minimum requirements to support short-
circuit protection is a directional (definite time-delay) over-current protection relay – a dis-
tance relay is recommended.
The following devices are required as primary protective disconnection equipment at the
connection point:
• reactive power plus under-voltage protection Q->& U<: The generating plant is dis-
connected from the network after 0,5 s, if all three line-to-line voltages at the network
point of connection are below 0,85 Uc (logical AND connection) and if the generating
plant simultaneously consumes reactive power. Optional the time delay can be ex-
tended to 1,5s if there is a trip command transfer to the generator’s CBs with a max-
imum time delay of 0.5s.
• rise-in-voltage protection U>> and U>: Three phase voltages between conductors,
logical OR
• under-voltage protection U<: Three phase voltages between conductors, logical OR
Generally: In the case of automatic reclosure, the plant operator must be prepared for the
recovery voltage at the network connection point may be asynchronous to the voltage of the
generating plant. [64]
• to stay at the medium-voltage grid with no feeding of any power (zero power mode)
• or if zero power mode isn’t available: to disconnect with U<< in less than 100ms
A.8 Italy
transformer
station
MV / 0,4 kV
for
consumers
• Buchholz transformer
• Buchholz on load tap changer
• Oil overtemperature
• Overcurrent
• Zero-sequence overvoltage;
• MV overvoltage
The protection system to be adopted on MV feeders against the phase to earth faults, de-
pends on the neutral point condition. In case of neutral earthed impedance the system is
more complex than in insulated neutral point condition.
• maximum phase current protection, double-pole version at least, with three thresh-
olds, one dependent-time and the other two, independent-time; since the first thresh-
old is used to prevent overload, the second is used to effect delayed tripping and the
A.8.3 Notes
A.8.3.1 67N
67N.S1:
Fault with a single phase to ground: neutral point compensated (Petersen coil)
Uen = (5 ÷ 6)% ∙Ur (PH-E)
IΣ = 2 A
tdiscr ≥ 1 s (depending from the MV network automation
rules)
67N.S2:
Fault with a single phase to ground: neutral point: insulated
67N.S3:
Fault with two phases to ground (cross-country-fault, no depending from neutral state)
Uen = 2% ∙Ur (PH-E)
IΣ = 150 A
tdiscr ≥ 0,1 s
Maximum voltage U>.S1 (ANSI 1,10 Ur Start time ≤ 3 s, not adjusta- Depending on voltage
CODE 59.S1), 10 minutes ble. values during the moving
mean function (according to Delay time setting = 0 ms window. Maximum 603,70
EN 61000-4-30, Class S, but Depending on voltage values s.
adopting a moving window with during the moving window.
refresh time ≤ 3 s) Maximum value 603 s.
Maximum inverse sequence 15% Ur/Ur (PH-E) For voltmetric unlock use Equal to start time
(5)
voltage Ui > (ANSI CODE 59 Vi) (ANSI CODE 81V): 0 ms
(1)
(indicative, de- (equal to start time: 70 ms)
pending on the
network)
Minimum direct sequence 70% Ur/Ur (PH-E) For voltmetric unlock use Equal to start time
(5)
voltage Ud< (ANSI CODE 27 (ANSI CODE 81V): 0 ms
(1)
Vd ) (indicative, de- (equal to start time:70 ms)
pending on the
network)
Threshold active only for inverters and rotating generators connected to distribution network with AC/AC converters.
For rotating generators directly connected U<.S2: operate time 70 ms, threshold value 70%, U<.S1: excluded .
For voltage values below 0,2 Ur, f>.S1, f>.S2 & f<.S1, f<.S2 protections shall be disabled.
Function used both for tripping and for voltmetric unlock function.
Regulation in % of nominal residual voltage Ur (PH-E) in case of a phase to earth fault with 0 Ω fault resistance derived
directly from an open delta winding or calculated internally the IPR from phase to earth voltages derived from non iron
core voltage transducers.
Regulation in % of nominal phase to earth or phase to phase voltage, according to voltage measurements methods.
A.9 Netherlands
VT Ifeeder
Protection device 2:
I> 1,5·Ir,feeder
I>> 0,7 * I”k (phase – phase)
ILx Protection Ie > 120 A for impedance grounded net
CT Device 2 tdiscr = 0,3 s
Protection device 3:
I> 1,5·Ir,feeder
Protection I>> 0,7 * I”k (phase – phase) – 0 s
Device 3 Ie > 120 A – 0 s for impedance grounded
net
tdiscr = 0,3 s
Protection
Device 4 Protection device 4:
In the 10 kV and 13 kV net mostly direc-
tional overcurrent relays for the feeder
In the 25 kV net line differential protection
MV / 0,4 kV In the 23 kV net a mix of directional over-
current and line differential protection
MV
G
Protection
Device
I>/t
Remarks:
• There are no fault ride through requirements in the Dutch grid code for all generator
types with a rated capacity less or equal to 5 MW.
• There are no fault ride through requirements in the Dutch grid codes for wind genera-
tors and parks independent of the rated capacity of the wind park or wind generator
• So far, in practice most of the wind generators under voltage protection are set at
U</t 0,8 p.u , t=100 msec. So far such fast isolation of wind generators during a fault
enabled us to use overcurrent protection in its simplest form.
• For relatively big generators inter-locking is used to avoid island operations. This is
typical in our 25 kV net
A.10 Norway
Neutral treatment: Petersen Coil earthed, Resistance added in case of an earth fault
TP DIFF
Petersen coil
or isolated Idiff
Feeder protection FP:
Z1A: Z<0,8*Zfeeder
x x t= 0,2 sec
FP FP Z1B: t=0,0 sec. (breaker close)
Z< Z< Z2: Z<1,2*Zfeeder
t= 0,3 sec.
A.11 Portugal
IL x
CT
UxE Feeder
VT Uen Protection
Device
core IΣ
balance
CT
DG Grid Protection Device:
U>/t 1,15UC / <0,1s
U</t 0,85|0,80UC / 1,5|1,6s†
U<</t 0,21|0,25UC / 0,0|0,6s†
U<</t N.U.|0,18UC / N.U.|<0,1s†
U0>/t 0,1UC / 1,0s
U0>>/t 0,7UC / <0,1s
F>/t N.U.|50,5Hz / N.U.|1,5s†
F>>/t 51,5Hz/ <0,1s
F</t N.U.|49,5Hz / N.U.|1,5s†
F<</t 47,5Hz/ <0,1s
I>/t 1,3·Ir,DG / 1,5s
transfer
DG Grid I>>/t 4,0·Ir,DG / <0,1s
station VT
Protection
MV U xE
Device
for OFF cmd
DG
G
CT
Remarks:
Transformer MV overcurrent protection is the first level and is mainly intended as a backup
to the feeders phase overcurrent protection. Starting current is set 30% above the rated cur-
rent to allow transformer overloading. There is a SCADA alarm when the current is above
110%IRated and the Dispatch centre will divert loads without the need for tripping.
There is also a backup protection for the feeder’s zero sequence overcurrent protection
which is not represented in the figures. The zero sequence overcurrent backup will trip the
transformer in 3s, for non-resistive faults, and in 180s, for resistive faults.
Earth fault protection for non-resistive faults is insured by the I0>> function which provides a
fast trip to minimize fault damage. Resistive faults are cleared by the I0d> with a higher time
setting to improve correct direction location probability. Detection of broken conductors and
extremely resistive faults are the objectives of function I0>, this leads to large fault clearing
times.
Automatic reclosing in rural (manly aerial) feeders has 3 cycles (one fast and two slow) and
is not used in urban (cable) feeders.
DG protection‘s main goal is to prevent unintentional islanding and the protection functions
associated are chosen and set accordingly. There are factors which may alter the protection
settings such as: Low Voltage Ride Through (LVRT) capability; direct link to the substation;
Direct tripping from the substation feeder through a communication link. DG units connected
to networks, which are also feeding other clients, must be tripped with low time settings. DG
units connected through a dedicated line and with fault ride through capability don’t need
such strict time settings which makes them more unlikely to suffer unwanted trips. In this
case the DG protection is coordinated with the feeder protection to allow a slower tripping. It
is assumed that no fast reclosing cycles are present in the DG feeder.
If the DG has LVRT then interconnection protection is coordinated with the LVRT curve and
is set not to trip with voltage dips within the curve. If the DG is placed in a network with other
clients then there may be the need for a direct trip through a communication channel.
Figure A-1: Fault Ride Trough curve in the Portuguese grid code
Remarks:
1. ‡ - without Direct Tripping | with Direct Tripping | with fault ride through capability
2. N.U. – Not used
A.12 Romania
transfer
Protection
station VT
MV UxE Device
U/t & f/t
for OFF cmd
DG
CT
G
ILx
Protection
Device
OFF cmd
I>/t
Notes:
1. In most of the cases the required protection is the overcurrent one. Other protection
requirements could be imposed if the selectivity of the protection scheme is not
During the voltage sags the DG has to supply active power according to the voltage
level and to maximize the injected reactive current without violating its operating lim-
its. It has to generate the maximum reactive current for at least 3s.
Remark:
1. The gazetted RSA Electricity Regulation Act makes provision for the Regulator to
publish compliance ‘Codes’ which are referred to below.
2. Embedded Generation(EG) > 0MVA connected to MV or HV networks & EG >
10MVA connected at any voltage level to the Distribution network shall in addition to
the Distribution Code also comply with the protection & control requirements of the
Transmission Grid Code. This code is detailed and is applied to all historic large gen-
eration. EG > 50MVA excluding Renewable Power Plants (RPPs), shall comply with
the governor requirements of the Transmission Grid Code. Note that the MVA rating
above refers to the total connected plant rating.
3. All RPPs shall comply with the RPP Grid Code. This code takes precedence.
4. All EG (which includes RPPs) connected to the Utility’s (Eskom) MV or HV networks
shall conform to its interconnection standard for embedded generation.
A.13.3 Requirements
The RPP Grid Code is divided into the following categories:
(i) Cat A1: 0 - x – 13,8kVA (ii) Cat A2: 13,8 < x < 100 kVA (iii) Cat A3: 100 – x < 1 MVA
b) Category B: 1 MVA – x < 20 MVA, also includes RPPs with rated power < 1 MVA con-
nected to MV.
c) Category C: 20 MVA ≤ x
52
51
Continuous
50 MINIMUM OPERATING RANGE FOR RPPs operating range
(49.0 Hz to %51.0 Hz)
49
48
47
46 200ms 4 6 60
0.1 1 10 100 1000
Duration of the incident, Seconds
With reference to the above diagram, when the frequency is less than 47.0Hz for longer than
200ms, the RPP may be disconnected and when the frequency is greater than 52Hz for
longer than 4s, the RPP shall be disconnected from the grid.
The voltage ride through capabilities and maximum disconnection times for RPPs of Catego-
ry A1 and A2 are given by the following:
V< 50 0.2
50 ≤ V < 85 2
85 ≤ V ≤ 110 Continuous
120 ≤ V 0.16
In connection with symmetrical fault sequences in areas B and D above, the RPP shall have
the capability of controlling the reactive power with the following compliance:
Requirements for reactive power support during volt drops or peaks at the POC
The supply of reactive power has the first priority in area B, while the supply of active power
has second priority. If possible, active power shall be maintained during voltage drops, but a
reduction in active power within the RPP’s design specifications is acceptable.
The power-frequency response curve for Category A RPP’s includes mandatory active pow-
er reduction and is given by the following figure:
The power-frequency response curve for Categories B and C is given by the following figure:
Parameter Hz
fmin 47
fmax 52
f1 SO
f2 SO
f3 SO
f4 50.5
f5 52
f6 50.2
A.14 Spain
In this characteristic the values for Il and for Vh are the ones given for Ins> and Vn>. The
typical values for Ih and Vl are: Ih = 3*Il y Vl = 2,2 V (Ur=110/3).
The Spanish distribution network can be briefly described as follows and divided in:
- Meshed: with voltages of 132, 110, 66, 50, 45 and 30 kV. The 30 kV network is con-
nected to the 132 kV / 220 kV network by means of a Y-D transformer and an artificial
ground connection is created with a zig-zag transformer. All the transformer generating
lower voltages are Dy. The 30 kV network is three-wire only grounded with the zig-zag,
connected in the substation,
- Radial:
- 45 kV solidly grounded: connected to the 132 kV / 220 kV network by means of
Yyn transformers with neutral solidly grounded. The transformers generating lower
voltages are Dyn or Dz.
- Solidly grounded below 30 kV (20, 15, 13 and 11 kV): the network is supplied
with a Dyn transformer solidly grounded or with a limiting reactance of 4 ohms.
Load transformers are Dyn or Dz.
- Zig-zag grounded below 30 kV (25, 20, 15, 11 kV): this network is supplied by a
Yd transformer. A zig-zag transformer is connected directly to ground or via a re-
sistor. The fault current is limited to 300, 500, 575 or 1000 A. Load transformers
are Dyn or Dz.
- Resistance grounded (25-20 kV): supplied by Yyn transformer grounded with a 6
ohm limiting resistor (1000 A limitation).
- Ungrounded (25, 20, 15 and 10 kV): supplied by Yyd, Yd or Yy transformers.
Load transformers are Dyn or Dz.
A.15 USA
Published Protection requirements for greater than 2500 KVA to 10 MVA Generators con-
nected to the Distributions System
---and---
a. Total generation is greater than or equal to 50% of the minimum line section
load.
---or---
b. Total generation is greater than 2000 KVA and less than 10 MVA.
---or---
Transformer Protection
Utility Station Utility Line Device:
8
Feeder Protection Device:
19 27
1 7 17 18 6. Device Type: Breaker
Failure Backup Tripping
59G 8 81 27/59
4 (i.e. LBB)
8. Device Type: Over or
Undervoltage Relays
16 Device #:59G or 27
25 27 9. Device Type: Impedance
5
9 12 13 11 Relay Plus Timer Device
3 2
21-2 51 32 51G #: 21-2/2
2
TR Connection to DG Grid Protection Device:
10
be specified by 2. Device Type: Undervolt-
utility engineer age Relay Device #: 27
15 14 3. Device Type: Syn-
chrocheck Relay Device
25-2
#: 25
SCADA RTU 6 14 11. Device Type: Ground
Overcurrent Relay Device
#: 51G
12. Device Type: Overcurrent
Relay Device #: 51
6 14 15 16 13. Device Type: Directional
15 Power Relay Device #: 32
14. Device Type: Synchroniz-
ing Relay Device #: 25
Other Load 17. Device Type: Under/over
Frequency Relay Device
27/59
#: 81U/O
G 25-1 81
18. Device Type: Under/over
14 17 18 voltage relay Device #:
1- These numbers refer to the device type notes
27/59
A.15.2 Requirements
Medium-sized generators (2000 -10,000 KVA) can supply relatively large amounts of energy
to the point of fault; therefore, additional protective functions are required. A generator con-
nected to a utility supply line section with a minimum load less than twice the rating of the
generator also stands a very good chance of islanding after the line protective equipment
opens with or without the fault remaining on the line. There are also instances where special
system constraints may require using this design. If there are multiple generators on a single
utility supply line section a different scheme may be required.
---and---
a. Total generation is greater than or equal to 50% of the minimum line section
load.
---or---
b. Total generation is greater than 2000 kVA and less than 10 MVA.
A few of the more commonly used devices are shown in the following list:
Device Numbers
2 Timer
4 Master Contactor
21 Distance Relay
25 Synchronizing or Synchronism Check
27 Under-voltage
32 Power Direction
40 Loss of Field Detection
46 Current Balance
47 Voltage Phase Sequence
50FD Phase Instantaneous Over-current Fault Detector
51 Time Over-current
51G Ground Time Over-current
51N Neutral Time Over-current
51V Voltage Restrained/Controlled Time Overcurrent
59 Overvoltage
59G Overvoltage Type Ground Detector
67V Voltage Restrained/Controlled Directional Time Overcurrent
79 Reclosing Relay
81O Over-frequency
81U Under-frequency
87 Current Differential
Purpose: Provide voltage for manual and automatic closing of source station breaker.
Purpose: Provide voltage supervision for closing of source station breaker. Breaker may
be closed if all 3 phases are dead.
Purpose: Provide voltage and phase angle supervision for manual and supervisory clos-
ing of source station breaker. Breaker will be manually closed for any of the following
indicated conditions, as requested by:
Purpose: Provide voltage to 59G or 27 relays for faults involving ground on the utility
system. Preferred connection is broken-delta; but if feeder loading is unbalanced to the
point that three times zero sequence voltage is normally significant, then the secondary
side should be connected grounded-wye. If this VT is to provide voltage also for imped-
ance relays or directional relays and is to be used for a 59G relay, then the VT may be a
three winding type with the third winding connected grounded-wye or the broken-delta
connection may be provided by using an aux. VT (grounded-wye/broken-delta) if the
main VT is grounded-wye/grounded-wye. The 27 relay could be part of the 27/59 relay-
ing (specified in note 21) if the 27/59 VT’s are located on the primary side of the inter-
connection transformer. This relaying is to provide a relay failure output to trip a select-
Purpose: Provide tripping of customer breaker(s) in the event of a fault on the utility sys-
tem involving ground. If the VT's are connected broken-delta, then the relay used is the
59G overvoltage type. If the VT's are connected grounded-wye on the secondary, then
the relays used are the 27 undervoltage relays.
Purpose: Provide for tripping of customer breaker in the event of a phase fault on the
utility system. This relay is used in a Zone 2 mode. The timer should be capable of
providing a trip time in the 0,5 second to 2 seconds range.
Purpose: Provide voltage for impedance relays and power directional relays.
Number Required: 1
Purpose: Provide for tripping of customer breaker in the event of a fault involving ground
on the customer system. This relay may be in the transformer neutral.
Purpose: Provide for tripping of customer breaker in the event of a phase fault on the
customer system.
Purpose: Provide for tripping of customer breaker if the transformer size is smaller than
generator, to limit power out if necessary to prevent damage to other customers, or to
limit power out because of utility system constraints. This relay is not used for fault de-
tection.
Purpose: Provide for proper closing of breakers when the customer generator(s) are to
be paralleled to the utility system.
Purpose: Provide voltage for synchronizing relays. May be one connected phase-to-
phase or may be a part of a 3-phase voltage transformer package.
Purpose: Provide voltage for under/over voltage and under/over frequency relays.
These voltage transformers are to be connected on the primary or secondary side of
power Transformer. One location only as specified by the utility engineer.
Number Required: 1
Purpose: Provide tripping of customer breaker in the event the frequency fails to be
maintained. This relay would be expected to operate if the customer should become iso-
lated on the utility line and not be able to maintain the load. The relay is to have a mini-
mum of one over-frequency and two under-frequency elements with a definite time type
characteristic capable of providing a trip time in the 0,1 second to 2 second range. The
setting is to conform to IEEE 1547 section 4.2.4 (Note utility reclosing exception in this
document), unless dictated by other utility system constraints. Frequency relays are to
be connected to VT’s on the primary or secondary side of power Transformer. One loca-
tion only as specified by the utility engineer.
Purpose: Provide tripping of customer breaker in the event the feeder or line voltage
cannot be maintained within acceptable limits. This relay should be a definite-time char-
acteristic or an instantaneous type with a timer. The relay is to have a minimum of two
over-voltage and two under-voltage elements with capable of providing a trip time in the
0,1 second to 2 second range. The setting is to conform to IEEE 1547 section 4.2.3
(Note utility reclosing exception in this document), unless dictated by other utility system
constraints. Voltage relays are to be connected to VT’s on the primary or secondary
side of power Transformer. One location only as specified by the utility engineer.
Purpose: May be a fuse or circuit breaker. Circuit breaker must not be dependent upon
A.C. power for tripping.
Purpose: To provide back-up relay protection should the primary relays fail. The relay
fail contact of microprocessor relays set up to trip the generator breaker of other desig-
nated breaker can fulfil this requirement. If a PLC is installed at the site that performs
critical generator functions, a back-up PLC is to be installed or a health check output is
to be wired in to trip the generator breaker for the failure of the PLC.
Number Required: 1
Purpose: Where the Generation Interconnection Customers is fed from two or more utili-
ty primary lines through a primary selective transfer scheme (ATO), the scheme must
be inhibited from operating until the generator is isolated from the utility line.
Purpose: Where possible the interconnect circuit breaker trip coil is to be monitored to
detect a failed trip coil and alarm the condition
Line Impedance between the Generation Interconnection Customers and the utility
substation:
Line impedance for which the DR must see and clear faults: