1
Review on Islanding Operation of Distribution
System with Distributed Generation
Pukar Mahat, Member, IEEE, Zhe Chen, Senior Member, IEEE, and Birgitte Bak-Jensen, Member,
IEEE
Abstract—The growing environmental concern and various
benefits of distributed generation (DG) have resulted in
significant penetration of DG in many distribution systems
worldwide. One of the major expected benefits of DG is the
improvement in the reliability of power supply by supplying load
during power outage by operating in an island mode. However,
there are many challenges to overcome before islanding
operation of a distribution system with DG can become a viable
solution in future. This paper reviews some of the major
challenges with islanding operation and explores some possible
solutions to overcome those challenges.
Index Terms— Frequency control, grid re-connection,
grounding, islanding detection, over-current protection, under-
frequency load shedding, voltage control.
I. INTRODUCTION
T HE yearly demand growth of less than 3% [1], concern
about the environment, and various benefits of onsite
generation have all resulted in a significant increase in
penetration of dispersed and distributed generation in many
distribution systems. This has also resulted in some power
system operational challenges. But, it has also opened up
some opportunities. One opportunity/challenge is an islanding
operation of a distribution system with DG unit(s). Islanding
is a situation in which a distribution system becomes
electrically isolated from the remainder of the power system
and yet continues to be energized by the DG connected to it.
Currently, it is seen as a major problem and all DG units need
to shut down when a distribution system is islanded. The
IEEE 1547-2003 standard requires islanding to be detected
and DG be disconnected at most within 2 seconds [2].
Similarly, the IEC 61727 standard also requires islanding
detection and DG disconnection at most within 2 seconds [3].
The Danish grid code requires that island operation of power
plants up to 25 MW is avoided [4]. Line workers’ safety, lack
of grounding [5], change in fault power [6], out of phase
reclosing [7], voltage and frequency control [6], etc. are often
portrayed as the reasons to avoid islanding operation.
The work is funded by PSO program, project 2006-1-6316, “Operation
and Control of Modern Distribution Systems”.
P. Mahat, Z. Chen and B. Bak-Jensen are with the Department of Energy
Technology, Aalborg University, Pontoppidanstræde 101, Aalborg, DK-9220,
Denmark (e-mails: pma@et.aau.dk, zch@et.aau.dk, bbj@et.aau.dk).
978-1-4577-1002-5/11/$26.00 ©2011 IEEE
However, with the DG penetration expected to increase
sharply as more and more renewables are integrated to power
system with the realization of smart grid, islanding is an
opportunity to improve the reliability of power supply.
Furthermore, islanding can be economically beneficial [8],
[9]. Islanding operation increases the revenue of DG owner by
additional sale. For the distribution network operators
(DNOs), islanding operation can improve the overall security
of power supply and they may also get additional revenue due
to the improvement in the quality of supply indices [8]. As for
the customers, the main benefit of islanding is the reduction in
the frequency and the duration of interruptions resulting from
outages in the distribution and transmission systems. Islanding
operation of a distribution system with DG is a viable option
provided that the various issues with island operation are
properly addressed. Some of the issues with islanding are state
(islanded or grid connected) detection, control of voltage and
frequency, load control and protection. This paper reviews
these key issues with islanding operation.
II. STATE DETECTION
State detection includes both islanding and grid
reconnection detections. This information is necessary if the
DG control and other operational settings are based on the
state of the distribution system, like setting of the over current
relays. Many islanding detection techniques have been
developed in recent times. Some of those are reviewed in
detail in [10], [11]. Most of the islanding detection techniques
can be grouped into following:
A. Remote Islanding Detection Technique
Remote islanding detection techniques are mainly based on
the communication between the utilities and DG units. It
mainly uses power line communication or Supervisory
Control and Data Acquisition (SCADA) to detect islanding.
Islanding detection techniques based on power line
communication are presented in [12]-[15] and the one based
on SCADA is presented in [16]. Even though, these remote
techniques have better reliability than other techniques, they
are complex and may be expensive to implement, particularly
for small distribution systems [17]. Furthermore, if there is
any problem with the transmitter, it may lead to malfunction
in the entire system. Therefore, local techniques, which
include passive, active and hybrid techniques, are widely used
to detect islanding.
 2

B. Passive Islanding Detection Technique
Passive islanding detection techniques take advantage of
the knowledge that system parameters like voltage, frequency,
etc. vary greatly after islanding. Rate of change of output
power of DG [12], rate of change of frequency [18], rate of
change of frequency over power [19], voltage unbalance [20]-
[22], harmonic distortion [23], data mining [24]-[25], wavelet
[26],[27] and spectral analysis [28], and fuzzy logic based
[29] are few examples of passive detection techniques.
Although passive methods are simple, their main problem is
that it is difficult to detect islanding when the load and
generation in an islanded distribution system closely match.
The limitations of passive detection techniques can be
overcome by active techniques, which can detect islanding
even under a perfect match of generation and demand in an
islanded distribution system.
C. Active Islanding Detection Technique
Active methods directly interact with the power system
operation by introducing perturbations. The main philosophy
of active islanding detection techniques is that a small
perturbation results in a significant change in system
parameters when a distribution system is islanded, whereas
the change is negligible when the distribution system is still
connected to the grid. Reactive power export error detection
system is reconnected back to the transmission grid. A passive
grid reconnection detection technique is presented in [39] and
an active technique is presented in [40]. Even though some of
the islanding detection techniques may be used to detect grid
reconnection, many of them cannot be used directly. As
examples, voltage and frequency of an islanded distribution
system may be easily maintained within under/over voltage
and frequency limits of the passive methods and the active
methods may push these parameters to such an extent that the
islanded distribution system may collapse. After the detection
of islanding, the frequency and voltage in the island have to
be maintained if sustained islanding operation is desired.
III. CONTROL OF DG
DG can be basically classified into three main types based
on their connection to the transmission grid. They are either
directly connected to grid like gas turbine generators (GTGs)
or connected through power electronics like photovoltaic (PV)
or mix of both like wind turbine generator (WTG) based on
doubly fed induction generator (DFIG). Some of the DGs can
be made to control active power or frequency and/or reactive
power/power factor or voltage.
1.05 1

Reactive Power (MVAr)

method [18], impedance measurement method [30], slip-mode
DG Bus Voltage (p.u.)

frequency shift algorithm (SMS) [31], active frequency drift 1.03 -1

(AFD) [32], active frequency drift with positive feedback
(AFDPF) [32], automatic phase-shift (APS) [33], adaptive
logic phase shift (ALPS) [34], and negative sequence current 1.01 -3
injection [35] are few examples of active islanding detection
techniques. The main problem with these techniques is that Reactive Power
they introduce perturbations in the system at predefined 0.99
Voltage
-5
intervals even though it is unnecessary during most operating 0 2
4 6 8 10
Time (s)
conditions and this may degrade the power quality. Fig. 1. Reactive power of DG for voltage rise at DG terminal
D. Hybrid Islanding Detection Technique
1.01 11
Hybrid techniques combine both passive and active
techniques to overcome the problems with both techniques.

Reactive Power (MVAr)

DG Bus Voltage (p.u.)

Average rate of voltage change and real power shift based 0.99 8

method [36], voltage unbalance and frequency set point based

method [37] and voltage/frequency and load switching based 0.97 5
method [38] are few examples of hybrid islanding detection
techniques. 0.95 2
The choice of the islanding detection technique will Reactive Power
basically depend on DG technology. Each islanding detection Voltage
0.93 -1
technique has advantages and disadvantages. However, a 0 2 4 6 8 10
Time (s)
hybrid detection technique with the combination of a faster
Fig. 2. Reactive power of DG for voltage drop at DG terminal
passive technique and a more accurate active technique seem
to be the solution for islanding detection. Even though they Figs. 1 and 2 show the reactive power of a 4.9 MVA DG
may fail to detect islanding for a perfect match of generation connected to a stiff grid when the grid voltage changes. A stiff
and demand, any subsequent change in demand or generation transmission grid is modeled by twenty 255 MVA
will result in the islanding being detected. synchronous generators operating in parallel. The generators’
E. Grid-Reconnection Detection Technique excitation data is presented in Table I.
Although a lot of work has been done in islanding
detection, little work has been done to detect when an islanded
 3

TABLE I in Fig. 4. The isochronous controller with feedback performs

GENERATOR EXCITATION SYSTEM DATA
relatively well under both scenarios. The optimal performance
DG Grid is achieved by operating the DG units with droop control
Excitation Model Type AC5A Type 1 while they are connected to the grid and by operating one DG
Parameters Value with isochronous control and rest, if any, with isochronous
Tr Measurement delay (s) 0 0 control with feedback when islanded. This is because, there
Ka Controller gain (p.u.) 500 175
Ta Controller time constant (s) 0.02 0.03 cannot be more than one isochronous controller in the same
Ke Exciter constant (p.u.) 1 1 system, else they would need to have the same speed set point;
Te Exciter time constant (s) 0.9 0.266 otherwise each generator tries to bring the frequency to its
Kf Stabilization path gain (p.u.) 0.03 0.0025 own reference setting. Similarly, for inverter interfaced DG,
Tf1 1st stabilization path time constant (s) 0.6 1.5 the control strategy has to be changed from power control to
Tf2 2nd stabilization path time constant (s) 0.38
Tf3 3rd stabilization path time constant (s) 0.058 frequency control. This can generally be done by locking the
E1 Saturation factor 1 (p.u.) 5.6 4.5 inverter frequency to grid frequency when the distribution
Se1 Saturation factor 2 (p.u.) 0.86 1.5 system is connected to grid. The inverter can set a frequency
E2 Saturation factor 3 (p.u.) 4.2 6.
Se2 Saturation factor 4 (p.u.) 0.5 2.46 within the required range when the distribution system is
Vmin Controller minimum output (p.u.) -7.3 -12. islanded.
Vmax Controller maximum output (p.u.) 7.3 12. The choice of the control strategy will be based on
The increase in terminal voltage leads to under excitation of available DG technologies and different control strategies may
the DG and may result in loss of synchronism [41]. On the be required for different DG technologies to achieve better
other hand, low terminal voltage may cause overloading and performance. The shift between the control strategies can be
excessive heating of the DG. According to [42], small easily done by correctly determining the states of the
generators’ operation at VAr/power factor control mode is distribution system by choosing the appropriate state detection
justifiable. However, when the distribution system is islanded, technique.
the voltage may go beyond the power quality limits. In such a
case, small generators have to maintain the islanded
distribution system’s voltage by operating in voltage control 50

mode. The power electronics interfaced DGs are operated to

Frequency (Hz)

produce power at the pre-set value when they are connected to

the grid [43], [44]. Inverters act as current controlled voltage 49

source inverters (VSI) when they are connected to grid [45].

However, they also have to control the voltage when the Droop control
distribution system is islanded [46],[47]. Hence, the inverters 48 Isochronous control
Isochronous control with feedback
act as voltage controlled VSI while they are islanded [45].
0 5 10 15 20
Apart from controlling the voltage, the DG also need to
control the frequency of the islanded distribution system. Time (s)

For the directly connected conventional DG units, like

Fig. 3. Performance of controllers during islanding
GTG, the output power of the DG and hence the frequency of
7
the islanded distribution system can be controlled with a Droop control
droop controller, an isochronous controller or an isochronous 6
Isochronous control
Turbine power (MW)

controller with feedback [39]. The test system presented in 5 Isochronous control with feedback
[39] is islanded with a power deficiency of 5 MW and a 4
generator reserve margin of 6 MW to test the controllers. The 3
controller time constant (T1) is 0.1 seconds. 2
Fig. 3 shows the distribution system frequency with three
1
different types of controllers when the distribution system is
0
islanded at time (t) =0 seconds(s). Similarly, Fig. 4 shows the
GTG’s turbine power, with three different controllers, when it -1
0 5 10 15 20
is connected to a grid whose frequency is above its reference Time (s)
frequency at t = 0s. Fig. 4. Performance of controllers while connected to grid
With a droop controller, the GTG finds a new operating
point close to the initial power set point while it is connected An islanded distribution system itself is a small power
to grid. But the frequency can go below the acceptable limit system. With the proper voltage and frequency control, it can
after islanding. Even though the isochronous controller bring operate without the support of the main grid. Even though the
the frequency back to reference value after islanding, it drives reliability of the electricity supply is improved by allowing
the output power of the turbine to the limits even with a slight islanding operation, the reliability of the overall system in an
difference in grid frequency and reference frequency as shown islanded operation is less compared to grid-connected
 4

operation [48]. Hence, it is desirable to reconnect the some loads have to be shed to stabilize the frequency in the
distribution system back to the transmission system when the islanded distribution system.
transmission system is back to normal operating conditions.
V. LOAD SHEDDING
IV. RECONNECTING ISLANDED DISTRIBUTION SYSTEM If the frequency of an islanded distribution system goes up,
Reconnecting an energized distribution system to a it can be controlled by reducing the output power of the
transmission grid is similar to synchronizing a DG to a generators. Similarly, when the frequency goes down, it can
distribution system or a central power station to a transmission be brought back by increasing the output power. However,
system. The difference between the voltages, frequencies and most DG units operate at maximum power to increase their
phases of the distribution system and the transmission system revenue. Photovoltaic generators use maximum power point
should be within certain limits before they can be reconnected tracking, variable speed wind turbines optimize power co-
such that they do not affect the network operation. Some of efficient (Cp) to produce maximum power and combined heat
the synchronization requirements for DG and/or energized and power plants are operated at maximum power. Therefore,
distribution system are summarized in Table II [2],[49]. ΔU, the only way to sustain an islanding operation when the
Δf and Δδ represents differences in voltages, frequencies and frequency goes down, while all DG units are operating at
phases, respectively. maximum power, is by shedding some loads. Under-
frequency load shedding has been used in power system for a
TABLE II
long time. Under-frequency load shedding techniques can
GENERATOR SYNCHRONIZATION CRITERIA
basically be classified into static and adaptive techniques. An
Standards example of static under-frequency load shedding which shed a
IEEE 1547 Belgium France fix amount of load at each frequency step is presented in [50].
On the other hand, adaptive or dynamic load shedding sheds
DG less than 500 kVA
ΔU = ±10% loads based on rate of change of frequency (RoCoF) like in
Δf = ±0.3Hz [51],[52]. The load shedding problem for the islanded
Δδ = ±20º distributed system should be treated differently from the large
ΔU, Δf, Δδ must power system because of differences in characteristics.
DG of 500-1500 kVA
ΔU = ±5%
be such that they ΔU = ±10% Islanded distribution systems often have small generators and
do not cause any Δf = ±0.1Hz hence small inertia. Thus, the frequency tends to decay more
Δf = ±0.2Hz
sudden variation Δδ= ±1 º
Δδ= ±15 º rapidly. Fig. 5 shows the distribution system frequency when
>6% in voltage
the test distribution system, presented in [40], is islanded
DG larger than 1500
kVA while all the DG units are operating at maximum power and
ΔU = ±3% the distribution system has a 1.84 MW power deficiency.
Δf = ±0.1Hz Loads are assumed as constant power loads. Islanding is
Δδ= ±10 º simulated at t=0s. The distribution system collapses within
0.24 seconds of islanding. Hence, it is critical to shed loads
With proper controls, frequency, phase and voltage can be quickly to stabilize the frequency.
controlled and the distribution system can be synchronized to 51
the transmission system. This is possible if the synchronizing
switch is close to the DG. But the synchronizing switch can be 50
far away from the DG and there may be multiple DG units in 49
Frequency (Hz)

the islanded distribution system. Hence, the DG can be made

to change the voltage, frequency and phase with the use of 48
communication. If implementing a communication system is
expensive for a distribution system, synchronization check 47
relays can be installed to check the synchronization criteria.
46
Since the DG units are able to control the voltage and
frequency of the islanded distribution system, the distribution 45
system can be made to wait for the voltage, frequency and 0 0.05 0.1 0.15 0.2 0.25
Time (s)
phase differences to be within the limits before it is connected
back to grid. If the control strategy of the DGs was switched Fig. 5. Frequency of islanded distributions system without load shedding
during the islanding operation, it has to be switched again for
grid connected condition. The control strategies discussed in Since the convention load shedding algorithms cannot be
section III will be able to maintain the frequency in the directly implemented in the distribution system as real time
islanded distribution system only if the total demand is less load and frequency data are rarely available, and frequency
than the generation. If there is more demand than generation tends to decay rapidly, an adaptive under frequency load
shedding for a distribution system based on customers’

willingness to pay, load histories and RoCoF is presented in
[53]. It sheds an optimal number of loads when the DG units
are operating at maximum power and frequency decays.
Impact of delays in load shedding on frequency profile of the
distribution system is presented in [54]. It also proposes
shedding different amount of loads when the frequency falls
below certain values. The amount of load needed to be shed is
calculated according to the swing equation and this
information is transmitted to the relays using slower
communication. In [55], the rate of change of voltage
information has been used to calculate the amount of load to
be shed when the distribution system has been islanded. A two
step optimization for under-frequency load shedding is
presented in [56]. It tries to shed load optimally by taking cost
into consideration. Communication can be used to co-ordinate
the load shedding and stabilize the frequency of the islanded
distribution system.
The shed load can be connected back to the distribution
system when the distribution system is reconnected back to
the grid. Grid reconnection can be identified through simple
grid-reconnection detection techniques or with the use of
more expensive and sophisticated communication system.
Furthermore, islanding detection takes some time. The time
lost in detecting islanding, in some cases, may become the
difference between a sustained island operation and a
distribution system collapse. Hence, it is reasonable to prepare
the distribution system for island operation by shedding loads
as soon as the frequency decay exceeds certain value without
waiting to conform if the distribution system is islanded. With
state detection algorithms, the state of the distribution system
is always known and loads can be reconnected whenever the
distribution system is not islanded.
VI. OVER CURRENT PROTECTION
The fault current depends on available sources of short
circuit power. DG doesn’t contribute as much fault power as
the transmission grid. Hence, when the distribution system is
islanded, the current seen by the protection relays, for a fault
in the distribution system, is less compared to the case when
the distribution system is connected to the transmission grid.
Bolted three phase faults are simulated at the end of each line
in the test distribution system presented in [40]. Maximum
fault currents in the faulted line sections, for grid connected
and island conditions, are presented in Fig .6.
Fig. 6 shows that there is a significant difference in the fault
currents during these two conditions. Therefore, an over-
current protection system designed to operate in a grid
connected mode may take longer time to clear faults when the
distribution system is islanded. As a result, electrical
equipment (like motors connected at low voltage level) may
be disconnected when the voltage stays below some value for
a certain period of time. This may also apply to power sources
like wind turbine generators, which are becoming common in
many distribution systems. Thus, by not the clearing faults
fast, not only the loads but also some generating units might
also be lost, which is undesirable especially in case of
5
islanded condition. On the other hand, if the over current
protection is designed for the islanded operation, then a fault
in the distribution system might result in unnecessary
operation of the breakers due to excessive fault currents from
the transmission grid when the distribution system is
connected to the grid. Hence the distribution system should be
protected against the changing fault current to be able to
operate in island mode.
8
Grid connected
7 Island
Fault Current (kA)
6
5
4
3
2
1
0
Z05-07 Z07-08 Z08-09 Z09-10 Z10-11 Z11-12 Z12-13 Z13-14
Lines
Fig. 6. Fault current in the lines for a three phase fault at the end of the
respective lines for grid connected and island conditions
One of the possibilities to overcome the challenges with the
changing fault power is the use of an adaptive protection.
Adaptive protection is “an online activity that modifies the
preferred protective response to a change in system conditions
or requirements in a timely manner by means of externally
generated signals or control action” [57]. Adaptive protection
of a distribution system with distributed generation can be
realized with the use of micro-processor based directional
over-current relays (DOCR). In the simplest form, the state
detection technique can be used to detect the state of the
distribution system and choose the appropriate tripping
characteristics. Fig. 7 shows schematics of a simple adaptive
directional over-current relay based on state detection.
Fig. 7. A simple adaptive protection relay
The adaptive protection system can also be designed to take
into account the plug & play DG units that may connect and
disconnect from the distribution system any time. An adaptive
protection that takes changes in network conditions into

consideration and updates the protection setting is presented
in [58]. Also, a communication system and high speed
processors can be used to locate the fault and update
protection settings [59]. The problem of changing fault power
can also be solved with the use of more expensive differential
relays [60]. Furthermore, the challenge with over current
protection can also be overcome by the use of some storage
devices that will increase the fault power of the islanded
distribution system so that faults will be cleared on time with
relays set for normal operation[61],[62]. Storage units require
some investment and can guarantee that the faults are cleared
without any delay if they can match the short circuit power of
transmission grid. However, storage will also help to improve
the voltage and frequency profile of the islanded distribution
system. Another possibility is to set over current relays for
islanded condition and island the distribution system when
there is a fault within the distribution system [63]. The method
is very simple and very useful for the distribution system that
always has more generation than demand. Otherwise, loads
have to be shed unnecessarily from time to time.
The choice of protection system will depend on network
characteristics and cost. If the total load is always less than
generation, islanding the distribution system whenever there is
a fault in it and using the same protection setting seems better
choice. On the other hand, if there are not many plug and play
generators and the total load can be more than generation at
times, adaptive protection with 2 settings (one for grid
connected condition and another for island condition), with
the state detection algorithm selecting the setting, can be used.
Similarly, if there are many plug and play generators and the
fault power constantly changes, then adaptive protection with
a number of settings can be used using communication to
choose an appropriate setting.
VII. GROUNDING AND UNSYMMETRICAL FAULTS
The most common type of fault in power system is a single
line to ground (SLG) fault. Unlike the three phase fault, the
fault current during a SLG fault depends on the grounding.
The three main types of grounding are isolated ground,
resonant grounding and low impedance grounding. In isolated
ground, connection to ground is through the large zero
sequence capacitance of the lines. Since the zero sequence
impedance, which is mostly capacitive, is much larger than the
short circuit positive sequence impedance, single phase to
ground fault currents are very small [64]. The healthy phases,
in this case, see a rise in voltage during SLG. In resonant
grounding, the neutral is connected to the ground through an
inductance in such a way that the capacitive fault current
component is compensated to a small residual current. The
inductor is often referred to as Petersen-Coil. Many countries
in Europe, including Denmark, use this type of grounding,
which allows a sustained operation of the distribution system
without opening of the feeder [65]. In low impedance
grounding, the fault current is the largest and the voltage in
the healthy phases rises the least compared to other grounding
for SLG fault.
6
The islanding of the distribution system may leave the DG
source without any kind of grounding. The SLG fault in such
a system could result in customers on the un-faulted phases
supplied with voltage that can increase up to 173%, at worst
case, of the pre-fault voltage level for an indefinite period [5].
Such a high voltage can result in damage of utility and
customer equipment. For example, marginally rated lightning
arresters may be destroyed if the voltage is 150% or more of
the pre-fault condition [66]. Moreover, during unsymmetrical
faults, negative sequence currents are generated. The ability of
machines to withstand these negative sequence current for
short duration depends on the size. The smaller DG has less
handling capacity compared to larger machines [67]. Hence, it
is recommended that the distribution system is grounded for
the island operation. Furthermore, effective grounding also
ensures that faults are cleared with the use of only over
current protection and thus the need of other forms of
protection for earth faults, like zero sequence protection, can
be avoided. The ungrounded distribution system can be
grounded though a grounding transformer that is usually of
wye-delta or zig-zag configuration [68]. Grounding of a
simple ungrounded islanded distribution system with a zig-zag
transformer is shown in Fig. 9.
Fig. 8. A zig-zag transformer used for grounding of an islanded distribution
system
VIII. CONCLUSION
Significant penetration of distributed generation has opened
the possibility of islanded operation of distribution systems to
increase the reliability of power supply. However, there are
still various issues with islanding. State detection (grid
connected or island), control of voltage and frequency, under-
frequency load shedding, over current protection and
grounding are few examples of the issues with island
operation. This paper reviewed these issues with islanded
operation and discussed the possible solutions. There are
various techniques available for islanding detection and
hybrid techniques stands out as the most promising one.
Furthermore, DG can be optimally operated by switching
from one control strategy to another when the distribution
system switches from grid connected to islanded condition or
vice versa. Some loads can be shed if the demand in the
islanded distribution system is higher than the total generating

capacity of the island to sustain the islanded operation. Fast
load shedding techniques are needed to shed loads in
distribution systems. Furthermore, adaptive protection can be
used to overcome the challenges with changing fault power.
Also, the distribution system should be grounded to avoid any
over voltage problem that may arise due to unsymmetrical
faults during island. In conclusion, islanding is a realistic
solution to improve the readability of power supply if the
issues are properly addressed. In fact, the solutions are already
there and its now just a matter of choosing the best one based
upon the network characteristics.
IX. ACKNOWLEDGMENT
Authors would like to thank Himmerlands Elforsyning
A.m.b.a, SEAV-NVE and DEFU, Denmark for their support
for the work.
X. REFERENCES
[1] R. E. Brown, “Reliability assessment and design optimization in electric
power distribution systems,” Ph.D. Dissertation, Univ. Washington,
USA, May 1996.
[2] IEEE Standard for Interconnecting Distributed Resources into Electric
Power Systems, IEEE Standard 1547TM, June 2003.
[3] Photovoltaic (PV) systems - characteristics of the utility interface, IEC
61727 Standard, December 2004.
[4] K. Christensen, et. al., “Technical regulation for thermal power station
units of 1.5 MW and higher,” Energinet.dk, Fredericia, Denmark, Tech.
Rep. TF 3.2.3, 2008.
[5] P. P. Barker and R. W. de Mello, “Determining the impact of distributed
generation on power systems: part 1 – radial distributed systems,” in
IEEE Power Eng. Soc. Summer Meeting, Seattle, WA, USA, 2000, vol.
3, pp. 1645-1656.
[6] S. P. Chowdhury, S. Chowdhury, P. A. Crossley, and C. T. Gaunt, “UK
scenario of islanded operation of active distribution networks with
renewable distributed generators,” Renewable Energy, vol. 34, no. 12,
pp. 2585-2591, Dec. 2009.
[7] R. A. Walling and N. W. Miller, “Distributed generation islanding-
implications on power system dynamic performance,” in IEEE Power
Eng. Soc. Summer Meeting, Chicago, IL, USA, 2002, vol.1, pp. 92-96.
[8] F. Castro-Sayas and G. D. Clarke. (2002). The costs and benefits of
embedded generation islanding operation [Online]. Available:
http://webarchive.nationalarchives.gov.uk/tna/+/http://www.dti.gov.uk/r
enewables/publications/pdfs/kel00284.pdf/.
[9] S. Abu-Sharkh, et. al., “Can microgrids make a major contribution to
UK energy supply?,” Renewable and Sustainable Energy Reviews, vol.
10, no. 2, pp. 78-127, Apr. 2006.
[10] P. Mahat, Z. Chen, and B. Bak-Jensen, “Review of islanding detection
methods for distributed generation,” in Third Int. Conf. Elect. Utility
Deregulation and Restructuring and Power Technologies, Nanjing,
China, 2008, vol. 1-6, pp. 2743–2748.
[11] R. S. Kunte and G. Wenzhong, “Comparison and review of islanding
detection techniques for distributed energy resources,” in 40th North
American Power Symposium, Calgary, AB, Canada, Sept. 2008, pp. 1-8.
[12] M. Ropp, K. Aaker, J. Haigh, and N. Sabhah, “Using Power Line
Carrier Communications to Prevent Islanding,” in 28th IEEE
Photovoltaic Specialist Conf., Anchorage, AK, USA, 2000, pp. 1675-
1678.
[13] W. Xu, G. Zhang, C. Li, W. Wang, G. Wang, and J. Kliber, “A power
line signaling based technique for anti-islanding protection of distributed
generators—part I: scheme and analysis,” IEEE Trans. Power Del., vol.
22, no. 3, pp. 1758-1766, July 2007.
[14] G. Wang, J. Kliber, G. Zhang, W. Xu, B. Howell, and T. Palladino, “A
power line signaling based technique for anti-islanding protection of
distributed generators—part II: field test results,” IEEE Trans. Power
Del., vol. 22, no. 3, pp. 1767-1772, July 2007.
7
[15] G. Wang, J. Kliber, and W. Xu, “A scalable power-line-signaling-based
scheme for islanding detection of distributed generators,” IEEE Trans.
Power Del., vol. 24, no. 2, pp. 903-909, Apr. 2009.
[16] M. A. Refern, O. Usta, and G. Fielding, “Protection against loss of
utility grid supply for a dispersed storage and generation unit,” IEEE
Trans. Power Del., vol. 8, no. 3, pp. 948-954, July 1993.
[17] S. Barsali, M. Ceraolo, P. Pelacchi, and D. Poli, “Control techniques of
dispersed generators to improve the continuity of electricity supply,” in
IEEE Power Eng. Soc. Winter Meeting, New York, NY, USA, 2002, vol.
2, pp. 789-794.
[18] J. Warin and W. H. Allen, “Loss of mains protection,” in ERA Conf.
Circuit Protection for industrial and Commercial Installation, London,
UK, 1990, pp. 4.3.1-12.
[19] F. Pai and S. Huang, “A detection algorithm for islanding-prevention of
dispersed consumer-owned storage and generating units,” IEEE Trans.
Energy Convers., vol. 16, no. 4, pp. 346-351, 2001.
[20] S. I. Jang and K. H. Kim, “An islanding detection method for distributed
generations using voltage unbalance and total harmonic distortion of
current,” IEEE Trans. Power Del., vol. 19, no. 2, pp. 745-752, Apr.
2004.
[21] S. I. Jang and K. H. Kim, “A new islanding detection algorithm for
distributed generations interconnected with utility networks,” in IEE Int.
Conf. Developments in Power System Protection, Amsterdam,
Netherlands, Apr. 2004, vol.2, pp. 571-574.
[22] S. Jang and K. Kim, “Development of a logical rule-based islanding
detection method for distributed resources,” in IEEE Power Eng. Soc.
Winter Meeting, New York, NY, USA, 2002, vol. 2, pp. 800-806.
[23] H. Kabayashi, K. Takigawa, and E. Hashimato, “Method for preventing
islanding phenomenon on utility grid with a number of small scale PV
systems,” in 2nd IEEE Photovoltaic Specialists Conf., Las Vegas, NV,
USA, Oct.1991, vol.1, pp. 695-700.
[24] K. El-Arroudi, G. Joos, I. Kamwa, and D. T. McGillis, “Intelligent-based
approach to islanding detection in distributed generation,” IEEE Trans.
Power Del., vol. 22, no. 2, pp. 828 – 835, Apr. 2007.
[25] K. El-Arroudi, and G. Joos, “Data Mining Approach to Threshold
Settings of Islanding Relays in Distributed Generation,” IEEE Trans.
Power Syst., vol. 22, no. 3, pp. 1112 – 1119, Apr. 2007.
[26] N. W. A. Lidula, N. Perera, and A. D. Rajapakse, “Investigation of a fast
islanding detection methodology using transient signals,” in IEEE Power
Eng. Soc. General Meeting, Calgary, AB, Canada, July 2009.
[27] A. Pigazo, M. Liserre, R. A. Mastromauro, V. M. Moreno, and A.
Dell’Aquila, “Wavelet-based islanding detection in grid-connected PV
systems,” IEEE Trans. Ind. Electron., vol. 56, no. 11, pp. 4445-4455,
Nov. 2009.
[28] J. Yin, C. Diduch, and L. Chang, “Islanding detection using proportional
power spectral density,” IEEE Trans. Power Del., vol. 23, no. 2, pp.
776-784, Apr. 2008.
[29] S. R. Samantaray, K. El-Arroudi, G. Joos, and I. Kamwa, “A fuzzy rule-
based approach for islanding detection in distributed generation,” IEEE
Trans. Power Del., vol. 25, no. 3, pp. 1427-1433, July 2010.
[30] P. O’Kane and B. Fox, “Loss of mains detection for embedded
generation by system impedance monitoring,” in Sixth Int. Conf.
Developments in Power System Protection, Nottingham, UK, March
1997, pp. 95-98.
[31] G. A. Smith, P. A. Onions, and D. G. Infield, “Predicting islanding
operation of grid connected PV inverters,” IEE Proc. Electric Power
Applicat., vol. 147, pp. 1-6, Jan. 2000.
[32] M. E. Ropp, M. Begovic, and A. Rohatgi, “Analysis and performance
assessment of the active frequency drift method of islanding
prevention,” IEEE Trans. Energy Convers., vol. 14, no 3, pp. 810-816,
Sep. 1999.
[33] G. Hung, C. Chang, and C. Chen. “Automatic phase shift method for
islanding detection of grid connected photovoltaic inverter,” IEEE
Trans. Energy Convers., vol. 18, no. 1, pp. 169-173, Mar. 2003.
[34] J. Yin, L. Chang, and C. Diduch, “A new adaptive logic phase-shift
algorithm for anti-islanding protections in inverter-based DG systems,”
IEEE Power Electron. Specialists Conf., 2005, Recife, Brazil, pp. 2482-
2486.
[35] H. Karimi, A. Yazdani, and R. Iravani, “Negative-sequence current
injection for fast islanding detection of a distributed resource unit,”
IEEE Trans. Power Electron., vol. 23, no. 1, pp. 298-307, Jan. 2008.

[36] P. Mahat, Z. Chen and B. Bak-Jensen, “A hybrid islanding detection
technique using average rate of voltage change and real power shift,”
IEEE Trans. Power Del., Volume 24, Issue 2, pp. 764 – 771, Apr. 2009.
[37] V. Menon, and M. H. Nehrir, “A hybrid islanding detection technique
using voltage unbalance and frequency set point,” IEEE Trans. Power
Syst., vol. 22, no. 1, pp. 442-448, Feb. 2007.
[38] W. Chang, "A hybrid islanding detection method for distributed
synchronous generators," in Int. Power Electron. Conf. (IPEC), Sapporo,
Japan, June 2010, pp. 1326-1330.
[39] P. Mahat, Z. Chen and B. Bak-Jensen, “Control strategies for gas turbine
generators for grid connected and islanding operations,” in IEEE PES
Transmission and Distribution Conf. and Exposition, New Orleans, LA,
USA, Apr. 2010.
[40] P. Mahat, Z. Chen and B. Bak-Jensen, “Control and operation of
distributed generation in distribution systems,” Electr. Power Syst. Res.,
vol.81, no. 2, pp. 495-502, Feb. 2011.
[41] T. W. Eberly and R. C. Schaefer, “Voltage versus VAr/power-factor
regulation on synchronous generators,” IEEE Trans. Ind. Appl., vol. 38,
no. 6, pp. 1682-1686, 2002.
[42] J. D. Hurley, L. N. Bize, and C. R. Mummert, “The adverse effects of
excitation system VAr and power factor controllers,” IEEE Trans.
Energy Convers., vol. 14, no. 4, pp. 1636-1641, Dec. 1999.
[43] J. Selvaraj and N. A. Rahim, “Multilevel inverter for grid-connected PV
system employing digital PI controller,” IEEE Trans. Ind. Electron., vol.
56, pp. 149-158, Jan. 2009.
[44] J. M. Espi Huerta, et al., "A synchronous reference frame robust
predictive current control for three-phase grid-connected inverters,"
IEEE Trans. Ind. Electron., vol. 57, pp. 954-962, Mar. 2010.
[45] H. Zeineldin, M. I. Marei, E. F. El-Saadany, and M. M. A. Salama, “Safe
controlled islanding of inverter based distributed generation,” IEEE
Power Electron. Specialists Conf., Aachen, Germany, vol. 4, June 2004,
pp. 2515-2520.
[46] M. E. Haque, et al., “A novel control strategy for a variable-speed wind
turbine with a permanent-magnet synchronous generator,” IEEE Trans.
Ind. Electron., vol. 46, pp. 331-339, Jan.-Feb.2010.
[47] T. C. Green and M. Prodanovic, “Control of inverter-based microgrids,”
Electr. Power Syst. Res., vol. 77, pp. 1204-1213, July 2007.
[48] M. Bollen, J. Zhong, O. Samuelsson, and J. Björnstedt, “Performance
indicators for microgrids during grid-connected and island operation,” in
IEEE Bucharest Power Tech Conf., Bucharest, Romania, June 2009.
[49] V. V. Thong, J. Driesen, and R. Belmans, “Overview and comparisons
of existing DG interconnection standards and technical guidelines,” in
Int. Conf. Clean Electrical Power, Capri, Italy, May 2007, pp. 51-54.
[50] P. G. Harrison, “Considerations when planning a load-shedding
programme,” Brown Bovcri Rev, no. 10, pp. 593-598, Sep. 1988.
[51] D. Prasetijo, W. R. Lachs, and D. Sutanto, “A new load shedding
scheme for limiting underfrequency,” IEEE Trans. Power Syst., vol. 9,
no. 3, pp. 1371-1378, Aug. 1994.
8
[52] V. V. Terzija, “Adaptive under-frequency load shedding based on the
magnitude of the disturbance estimation,” IEEE Trans. Power Syst., vol.
21, no. 3, pp. 1260-1266, Aug. 2006.
[53] P. Mahat, Z. Chen, and B. Bak-Jensen, “Under-frequency load shedding
for an islanded distribution system with distributed generators,” IEEE
Trans. Power Del., vol. 25, no. 2, pp. 911 -918, Apr. 2010.
[54] S. Hirodontis, H. Li, and P. A. Crossley, "Load shedding in a
distribution network," Int .Conf. Sustainable Power Generation and
Supply, Nanjing, China, Apr. 2009, pp. 1-6.
[55] I. Balaguer-Alvarez, et. al., “Control for grid-connected and intentional
islanding operations of distributed power generation,” IEEE Trans. Ind.
Electron., to be published.
[56] P. Du, and J. K. Nelson, “Two-step solution to optimal load shedding in
a micro-grid,” IEEE PES Power Syst. Conf. Exposition, Mar. 2009,
Seattle, WA, USA, pp. 1-9.
[57] G. Rockefeller, et al., “Adaptive transmission relaying concepts for
improved performance,” IEEE Trans. Power Del., vol. 3, no. 4, pp.
1446-1458, Oct. 1988.
[58] H. Cheung, A. Hamlyn, Y. Cungang and R. Cheung, “Network-based
Adaptive Protection Strategy for Feeders with Distributed Generations,”
in IEEE Canada Electrical Power Conf., Montreal, Canada, Oct. 2007,
pp. 514-519.
[59] S. A. M Javadian,. R. Tamizkar, and M. R. Haghifam, “A protection and
reconfiguration scheme for distribution networks with DG,” in IEEE
Bucharest Power Tech Conf., Bucharest, Romania, June 2009, pp. 1-8.
[60] H. H. Zeineldin, E. F. El-Saadany, and M. M. A. Salama, “Distributed
generation micro-grid operation: control and protection,” in Power Syst.
Conf.: Advanced Metering, Protection, Control, Communication, and
Distributed Resources, Clemson, SC, USA, Mar. 2006, pp. 105-111.
[61] N. Jayawarna, N. Jenkins, M. Barnes, M. Lorentzou, S. Papthanassiou,
and N. Hatziargyriou, “Safety analysis of a microgrid,” in Int. Conf.
Future Power Syst., Amsterdam, Netherlands, Nov. 2005.
[62] M. Pedrasa and T. Spooner, “A survey of techniques used to control
microgrid generation and storage during island operation,” in Australian
Uni. Power Eng. Conf. (AUPEC), Melbourne, Australia, Dec. 2006.
[63] H. Nikkhajoei and R. Lasseter, “Microgrid protection,” in IEEE Power
Eng. Soc. General Meeting, Tampa, USA, June 24-28, 2007.
[64] P. Hansen, J. Østergaard, and J. S. Christiansen, “System grounding of
wind farm medium voltage cable grids,” in Nordic Wind Power
Conference, Roskilde, Denmark, 2007.
[65] CIRED working group WG03, “Fault management in electrical
distribution systems,” Finland, December 1998.
[66] IEEE Guide for the Application of metal Oxide Surge Arresters for
Alternating Current Systems, IEEE Standard C62.22-1997, 1998.
[67] Dromey Design. Synchronous generation in distribution systems
[Online]. Available:
http://www.essexenergy.ca/downloads/SynchronousGeneration.pdf
[68] M. Shen, L. Ingratta, and G. Roberts, “Grounding transformer
application, modeling, and simulation," in IEEE Power Eng. Soc.
Summer Meeting, Pittsburgh, PA, USA, July 2008.