ArticlePDF Available

Abstract and Figures

This paper describes and evaluates the feasibility of control strategies to be adopted for the operation of a microgrid when it becomes isolated. Normally, the microgrid operates in interconnected mode with the medium voltage network; however, scheduled or forced isolation can take place. In such conditions, the microgrid must have the ability to operate stably and autonomously. An evaluation of the need of storage devices and load shedding strategies is included in this paper.
Content may be subject to copyright.
Abstract—The main objective of this paper is to present the
development of microsource modelling and the definition of
control strategies to be adopted to evaluate the feasibility of
operation of a microgrid when it becomes isolated. Normally, the
microgrid operates in interconnected mode with the MV
network, however scheduled or forced isolation can take place. In
such conditions, the microgrid must have the ability to operate
stably and autonomously. An evaluation of the need of storage
devices and load-shedding strategies is included in the paper.
Index Terms—Power system dynamic stability and control;
Renewable energy sources and storage devices; Integration of
distributed generation in the main grids.
HE need of reducing CO
emissions in the electricity
generation field, recent technological developments in the
microgeneration domain and electricity business restructuring
are the main factors responsible for the growing interest in the
use of microgeneration. In fact, the connection of small
generation units – the microsources (MS), with power ratings
less than a few tens of kilowatts – to Low Voltage (LV)
networks potentially increases reliability to final consumers,
brings additional benefits for global system operation and
planning, namely regarding investment reduction for future
grid reinforcement and expansion. In this context, a MicroGrid
(MG) comprises a LV network (for example covering an
urban area, a shopping-centre or even an industrial park), its
loads and several small modular generation systems connected
to it [1].
Examples of MS technologies to be used when building a
MG include renewable power sources, such as wind and
photovoltaic (PV) generators, microturbines working on gas or
bio-fuels and different types of fuel-cells, and also storage
devices (such as flywheels or batteries).
The MG is intended to operate in two different operating
Normal Interconnected Mode – the MG is connected to a
main MV network, either being supplied by it or
This work was supported in part by the European Commission within the
framework of EU Project MicroGrids, Contract No. ENK-CT-2002-00610.
J. A. Peças Lopes is with INESC Porto and FEUP – Faculdade de
Engenharia da Universidade do Porto, Porto, Portugal (e-mail:
C. L. Moreira is with INESC Porto and FEUP, Porto, Portugal (e-mail:
A. G. Madureira is with INESC Porto and FEUP, Porto, Portugal (e-mail:
injecting some amount of power into the main system.
Emergency Mode – the MG operates autonomously, in a
similar way to physical islands, when the disconnection
from the upstream MV network occurs.
It will not be common to find fully controllable
synchronous units in a MG, which are normally responsible
for voltage and frequency control in conventional power
systems. The majority of MS to be installed in a MG are not
suitable for direct connection to the electrical network due to
the characteristics of the energy produced (DC power in fuel-
cells and PV generators or high frequency AC power in
microturbines). Therefore, a power electronic interface
(DC/AC or AC/DC/AC) is required. For instance, in [2] a
control scheme based on droop concepts to operate inverters
feeding a standalone system is presented.
In this paper the droop concept for inverter control is
further explored in different modes of operation. Two inverter
control schemes are combined in order to demonstrate the
feasibility of a seamless transition from Normal
Interconnected Mode to Emergency Mode under specific
conditions, as well as the possibility of stably operating a MG
in islanded conditions. In order to achieve this goal, the
potentialities of the MatLab® Simulink® environment and its
libraries (mainly the SimPowerSystems toolbox) were
employed in order to develop a simulation platform suitable
for identifying MG control requirements and evaluating the
MG dynamic behaviour under several conditions. Different
MS technologies coexist and are operated together in the
simulation platform. Controllable and non-controllable
sources as well as the disconnection of non-essential loads are
used in order to guarantee the continuity of electric supply in a
LV area after scheduled or forced loss of the upstream MV
network connection.
This research is being developed within the framework of
an EU R&D project with the objective of studying the
problems that challenge the integration of large amounts of
different MS in LV grids and involves several institutions and
companies [3].
The control of the MG is based on a hierarchical control
architecture in order to assure a robust operation [3].
Consequently, a MicroGrid Central Controller (MGCC) is
installed at the LV side of a MV/LV substation managing in
an upper level the MG operation through several crucial
Defining Control Strategies for Analysing
MicroGrids Islanded Operation
J. A. Peças Lopes, Senior Member, IEEE, C. L. Moreira and A. G. Madureira
management functions, both technical and economical. At a
second hierarchical level each MS and storage device is
locally controlled by a Microsource Controller (MC) and each
electrical load or group of loads is controlled by a Load
Controller (LC). A communication infrastructure must also be
provided in order to guarantee information exchange between
the MGCC and the other controllers. A typical MG structure is
shown in Fig. 1.
Fig. 1. MG control architecture
The interaction between the control devices is as follows:
the MGCC promotes adequate technical and economical
management policies and provides set-points to LC and MC.
LC will act based on an interruptibility concept and MC are
responsible for the control of the MS active and reactive
power production levels.
It is important to understand that the amount of data to be
exchanged between the several network controllers is small, as
it includes mainly messages with set-points to LC and MC, as
well as information requests sent by the MGCC to LC and MC
about the active and reactive powers and voltage levels.
Another important factor that eases the establishment of the
communication infrastructure is the small geographical span
of the MG.
The communications system can use either Power Line
Communication, which presents some interesting
characteristics for this type of network, or explore other type
of access such as Wireless Communication (a rapid-growing
The analysis of MG behaviour requires the development of
a set of dynamic models able to simulate the response of the
MG under several conditions. For this purpose, MS and
storage devices, together with control systems, have been
Inverter modelling is an important issue, especially
regarding operation control, thus deserving careful analysis
and detailed implementation.
A. Microsource Modelling
The MS models developed within the project include
photovoltaic arrays, wind generators, microturbines and a fuel-
cells. Concerning storage devices, flywheel systems and
batteries have also been modelled [4].
For illustration purposes only details on the dynamic model
developed for the Solid Oxide Fuel-Cell (SOFC) are given
next. The fuel-cell includes a Fuel Processor that converts the
used fuel in Hydrogen, a Power Section, where chemical
reactions take place, and a Power Conditioner that converts
DC to AC power. The SOFC model adopted assumes several
simplifications, such as: fuel gases are considered to be ideal,
it is sufficient to define only one single pressure value in the
interior of the electrodes, the temperature in the fuel-cell is
presumed to be always stable, only ohmic losses are
considered, assuming that the working conditions are far away
from the upper and lower extreme values of current, and the
Nernst equation is assumed to be applicable. The complete
model can be seen in the block diagram in Fig. 2.
The state variables used to model the fuel-cell behaviour
are the reaction current (I
,) the hydrogen input flow (q
and the partial pressure of the reaction components - p
, p
and p
- respectively hydrogen, oxygen and water.
The full dynamic model description for the adopted SOFC
can be found in [5] and [6].
Fig. 2. SOFC block diagram model
A GAST dynamic model was adopted for the microturbines
primary unit, since they are small simple-cycle gas turbines,
[5]. Both high-speed single-shaft units, with a compressor and
turbine mounted on the same shaft as the electrical
synchronous machine, and split-shaft units using a power
turbine rotating at 3000 rpm and a conventional induction
generator connected via a gearbox, were modelled. The single
shaft unit requires an AC/DC/AC converter to connect the unit
to the grid. The wind generator is considered to be an
induction machine directly connected to the network and
represented by fifth-order model available in MatLab®
Simulink®. Concerning the PV generator, it was assumed that
the array is always working at its maximum power level for a
given temperature and irradiance. It is basically an empirical
model based on experimentation results as described in [7].
B. Storage Devices Modelling
Storage devices like flywheels and batteries are modelled
as constant DC voltage sources (taking into account the time
span under analysis) to be coupled with the electrical network
by means of power electronic interfaces. They act as
controllable AC voltage sources (with very fast output
characteristics) to face transients like in load-following
situations. Despite acting as voltage sources, they have
physical limitations and have a finite capacity for storing
Grid interface of the storage devices is done using
AC/DC/AC converters for flywheels and DC/AC inverters for
batteries. Usually, the active power is injected into the MG
using a proportional to frequency deviation control approach
(with a specified droop), with the energy delivered to grid
evaluated through the time integral of the active power
injected by the storage device during the simulation time [4].
Due to the large time constants found in the responses of
several MS, such as fuel-cells and microturbines, storage
devices have to provide the amount of power required to
balance the system.
C. Inverter Modelling
Usually, two kinds of control strategies may be used to
operate an inverter. According to the control strategy
followed, the corresponding model is derived. More details on
these strategies can be found in [8]:
PQ inverter control: the inverter is used to supply a given
active and reactive power set-point.
Voltage Source Inverter control logic: the inverter is
controlled to “feed” the load with pre-defined values for
voltage and frequency. Depending on the load, the
Voltage Source Inverter (VSI) real and reactive power
output is defined.
It is also important to mention that when analyzing the
dynamic behaviour of the MG inverters are modelled only by
their control functions. This means that fast switching
transients, harmonics and inverter losses are neglected.
1) PQ Inverter Control
This kind of control can be achieved using an inverter
control scheme based on a current-controlled voltage source.
In [9] a method for computing single-phase active and
reactive powers is presented. This method was adapted in this
work in order to compute the instantaneous active and reactive
components of the inverter current: the active component is in
phase with the voltage and the reactive component with a 90
degrees (lagging) phase-shift, being both limited in the
interval [-1, 1].
The active component is used to control the DC link
voltage and consequently the inverter active output power in
order to balance MS and inverter active power output, whereas
the reactive component controls the inverter reactive power
output. Power variations in the MS lead to a variation of the
DC link voltage, which is corrected via the PI regulator by
adjusting the active current output This inverter can be
operated with a unit power factor (Set Point = 0 in Fig. 3) or
receive a set-point (locally or from a central controller) for the
output reactive power.
Fig. 3. PQ inverter control system
2) Voltage Source Inverter Control
In this case the frequency and voltage control concept used
in a synchronous machine is transposed for the VSI control.
When a VSI is operating in parallel with a stiff AC system
with angular frequency ω
 (Fig. 4), the output power of the
VSI is defined from the droop equation derived. In order to
change the output power of the inverter, a change in the idle
frequency (ω
) is required. If the AC system is not available,
the output power of the inverter depends on the network load
and droop settings so that the network frequency reaches a
new value. The active power is shared among all the inverters
at the new frequency value according to the droop settings of
each VSI [9]. Similar considerations can be made for the
voltage/reactive power control.
Fig. 4. Frequency / active power droop characteristic
A three-phase balanced model of a VSI including the droop
concepts was derived from a single-phase version presented in
[9] and is shown in Fig. 5.
VSI voltage and current are measured in order to calculate
active and reactive powers, which are delayed for decoupling
purposes and in order to emulate the behaviour of a
synchronous machine. Frequency is determined by the delayed
active power through the frequency/active power droop.
Similarly, the voltage magnitude is determined by the delayed
reactive power by using the reactive power/voltage droop.
This control scheme allows the computation of the voltage
reference signals to control the VSI switching sequence using
a PWM technique.
Fig. 5. VSI control system
3) Inverter Behaviour During Short-Circuit Conditions
Conventional power plants comprising synchronous
machines directly connected to the network are able to provide
large short-circuit currents, which are very helpful for a
and efficient fault detection and elimination. However, in a
MG where generation units are mainly connected to the grid
through power electronic interfaces, it is difficult to obtain
high fault currents. According to the protection guidelines for
a MG presented in [10], a protection scheme based on current
(over current protection) will be a great concern due to the low
short-circuit to load current ratios, during islanded operation.
Power electronic switching devices used in inverters are
selected based on voltage, current carrying capability (under
certain cooling conditions for a defined switching frequency)
and safe operating areas. Based on these considerations, the
short-circuit handling capability of a power electronic
interface can be increased only by increasing its power rating.
Accordingly, the following considerations are made:
The VSI was selected to be up-rated in order to provided
a suitable contribution for short-circuit currents (ranging
from 3 to 5 p.u.)
PQ inverters can provide only a small amount of short
circuit current (around 1.5 p.u.)
The PQ inverter control scheme permits the control over
the output inverter current during short-circuit conditions. In
case of a short-circuit, the voltage drop at the inverter
terminals leads to a reduction of the active power output. As a
consequence, the DC link voltage increases and the PI
controller tries to increase the active output current of the
inverter. Limiting the total gain of the PI controller, the output
current of the inverter can also be limited. Simultaneously, an
increase in the DC link voltage will be experimented.
Being a voltage source, the output current of a VSI tends to
be very high (similarly to what happens in conventions
synchronous machine). In order to limit its output current a
control technique like the one presented in Fig. 3 is used. The
main difference is that in this case the current reference has a
maximum peak value dependent on switching devices
characteristics and its frequency is imposed by inverter
frequency/active power droop. The output current is
continuously monitored, and if its value overcomes the
maximum value, the control scheme is switched accordingly.
When the fault is removed, the VSI returns to Voltage
D. Load Modelling
The dynamic behaviour of MG was evaluated considering
only three-phase balanced operation and two load types:
constant impedance loads (power dependent on frequency and
voltage) and motor loads (an induction motor with constant
mechanical torque). As it will be shown, the load
characteristics influence greatly the dynamic behaviour of the
MG, mainly in short-circuit conditions.
Controllable loads, available for load-shedding have also
been modelled, with the amount of load to be shed defined
from the amplitude of the grid frequency deviation.
A MG is an inverter dominated network, where the power
electronic interfaces are responsible for controlling frequency.
Also, a voltage control strategy is required; otherwise the MG
can experience voltage and/or reactive power oscillations [11].
While the MG is being operated in interconnected mode,
all inverters are operated in PQ mode. However, a sudden
disconnection of the main power supply (the upstream MV
network) would lead to the loss of the MG, since there would
be no possibility for load/generation balancing, and therefore
for frequency and voltage control. The unit that can be used to
achieve these requirements is the VSI. Using its control
capabilities by means of droop settings adjustment, a VSI can
be operated in parallel with the main grid without injecting
active or reactive power. When disconnection from the main
grid happens, the VSI output is automatically determined by
the deviation between load and generation in the MG. After
identifying the key solution for MG islanded operation, two
main control strategies are possible:
Single Master Operation: A VSI or a synchronous
machine directly connected to the grid (with a diesel
engine as the prime mover, for example) can be used as
voltage reference when the main power supply is lost; all
the other inverters can then be operated in PQ mode;
Multi Master Operation: More than one inverter is
operated as a VSI, corresponding to a scenario with
dispersed storage devices; other PQ inverters may also
As already mentioned, the VSI has the ability to emulate
the behaviour of a conventional synchronous generator. It
reacts to power system disturbances (for example, load-
following situations or wind fluctuations) based only on
information available locally at the inverter’s terminals
(voltage and current measurements) [2]. In order to promote
adequate secondary control with the aim of restoring
frequency to the nominal value after a disturbance, two main
strategies can be followed: local secondary control, by using a
local PI controller at each MS, or centralized secondary
control mastered by the MGCC, both defining target values
for active power outputs of the primary energy sources [12].
A. Single Master Operation
In this case, a VSI (acting as “master”) is connected to the
network; the other MS are connected to the grid through an
inverter with a PQ control scheme (“slaves”). Droop settings
of the VSI can be modified by the MGCC according to the
operating conditions and in order to avoid large frequency
Q Set Point
V, I
V, I
Droop Settings
P&Q Settings
Fig. 6. Control scheme for single master operation
Assuring a zero frequency deviation during any islanded
operating conditions should be considered the key objective
for any control strategy. This is especially important since
storage devices have limited capacity and it is necessary to
avoid them from keep injecting (or absorbing) active power
whenever the frequency deviation differs from zero.
B. Multi Master Operation
In a multi master approach, several inverters are operating
as VSIs with pre-defined frequency/active power and
voltage/reactive power characteristics. Eventually, other PQ-
controlled inverters may also coexist.
The correction of frequency deviations can be performed
by changing the idle frequency value as used in
proportional/integral frequency governors of synchronous
generators. The change in the idle frequency can also be
performed centrally by the MGCC, a sort of centralized
secondary control, using the communications infrastructure. In
this control strategy the aim of obtaining zero frequency
deviation is also a driving concern, as in the single master
approach and for the same reasons.
Fig. 7. Control scheme for multi master operation
A simulation platform under the MatLab® Simulink®
environment was developed in order to evaluate the dynamic
behaviour of several MS operating together in a LV network
under pre-specified conditions including interconnected and
autonomous operation of the MG. At this stage only three-
phase balanced operation of the network is considered.
A LV test system, defined by NTUA [13], was used to test
the approaches developed. Fig. 8 shows a single-line diagram
with the different types of MS operated in this MG.
Fig. 8. LV network test system
Fig. 9 illustrates the LV simulation platform developed for
the dynamic simulation studies. It includes models and
controls for microturbines (single-shaft and split-shaft), fuel-
cells, small asynchronous wind generators, PV panels and
storage devices (flywheels and batteries) as well as
controllable loads (available for load-shedding).
Fig. 9. LV network for the Matlab® Simulink® simulation platform
In order to illustrate how MS are represented under the
MatLab® Simulink® environment, the model of the SOFC is
shown in Fig. 10. This representation corresponds to the
transposition of the model described in Fig. 2. All MS have
been implemented here in the same way. Using the “look
under mask” and the “block parameters” options it is possible
to change parameters of the MS and its controls.
Fig. 10. SOFC model for the Matlab® Simulink® simulation platform
Disconnection from the upstream MV network and load-
following in islanded operation was simulated in order to
understand the dynamic behaviour of the MG and to evaluate
the effectiveness of the developed control approaches. The
islanding of the MG was investigated for two different
situations: the scheduled islanding and the forced islanding (in
case of a fault in the MV grid). Scenarios using the single
master operation strategy with a VSI and multi master
operation were tested. Due to space limitations, only the
results for forced islanding for single master operation and for
scheduled islanding for multi master operation are presented.
A. Single Master Operation
The short-circuit occurred at t=10 seconds and was
eliminated after 100 milliseconds with the islanding of the
The initial total load of the MG was around 70 kW and the
microsource generation, prior to the islanding, was around 45
kW. In face of the large initial frequency deviation an amount
of load was automatically shedded in order to aid frequency
restoration. This load was reconnected later in small load steps
allowing also the evaluation of the MG behaviour in load-
following conditions.
It is possible to observe from the frequency behaviour that
MG stability is not lost when facing the short-circuit at the
MV grid side.
In order to preserve MG stability it was necessary to shed
the motor loads since the rotation speed would drop too much
and cause the whole system to collapse. Asynchronous
generators (single-shaft microturbine and wind generator)
were not disconnected in order not to loose generation. After
fault elimination, there is a transient period for restoring
normal operation of these generators, which has a strong
impact on inverter current and voltage, as it can be observed in
Fig. 12 after t=10.1 seconds.
0 20 40 60 80 100 120
Frequency (Hz)
0 20 40 60 80 100 120
VSI P and Q
(kW and kvar)
0 20 40 60 80 100 120
Time (s)
P (kW)
Fig. 11. MG Frequency, VSI active and reactive power and SOFC and single-
shaft microturbine active power
9.9 10 10.1 10.2 10.3 10.4 10.5
Time (s)
SOFC I rms (A)
9.9 10 10.1 10.2 10.3 10.4 10.5
VSI V rms (V)
9.9 10 10.1 10.2 10.3 10.4 10.5
VSI I rms (A)
Fig. 12. VSI current and voltage and SOFC current (rms values)
B. Multi Master Operation
In order to analyze the dynamic behaviour of a MG under a
multi master approach, the dynamics of the primary energy
sources (single-shaft microturbine and fuel-cell) were
neglected due to the high storage capacity assumed to be
installed at their DC link [8]. The considered scenario is
similar to the previous one. The disconnection of the upstream
MV network was simulated for t=10 seconds.
After the islanding, active power is shared amongst several
inverters according to droop settings. At t=20 seconds a local
secondary control (based on an integral frequency control
deviation approach) is applied to correct the steady state
frequency deviation after islanding. It is also possible to
observe that voltage variations are very small. In this case
voltage control through droops is sufficient to maintain
voltage levels within acceptable limits.
0 5 10 15 20 25 30 35 40
Frequency (Hz)
0 5 10 15 20 25 30 35 40
Time (s)
Active Power (kW)
Fig. 13. MG frequency and active power in microsources
0 5 10 15 20 25 30 35 40
L - L Voltage (V)
0 5 10 15 20 25 30 35 40
Time (s)
Reactive Power (kvar)
Fig. 14. Microsources voltages (rms values) and reactive powers
From the analysis performed the following main
conclusions can be drawn regarding the dynamic behaviour of
the MG, given the control strategies presented:
Simulation results indicate that the islanding of the MG,
both scheduled and forced, can be performed safely in
several different importing and exporting conditions.
Simulation results also indicate that both control
strategies tested – the single master and the multi master
approach – are effective and assure efficient MG
Also, the results obtained suggest that storage devices are
absolutely essential to implement good control strategies
for MG operation in islanded mode and the load-
shedding procedure is also of very high importance to
sustain fast and long frequency deviations.
The authors want to express their thanks to the research
team of the MicroGrids project for valuable discussions that
helped developing all this research and to the EU for funding
the project.
[1] R. H. Lasseter, P. Piagi, “MicroGrid: A conceptual Solution”, in Proc. of
the 35
PESC, pp. 4285-4290, Germany, 2004.
[2] M. C. Chandorkar, D. M. Divan, R. Adapa, “Control of Parallel
Connected Inverters in Standalone ac Supply Systems”, IEEE
Transactions on Industry Applications, vol. 29, No 1, 1993.
[3] J. Peças Lopes, J. Tomé Saraiva, N. Hatziargyriou, N. Jenkins,
“Management of MicroGrids”, JIEEC2003 Bilbao, 2003.
[4] G. Kariniotakis, et al, “DA1 – Digital Models for Microsources”,
MicroGrids project deliverable DA1, 2003.
[5] Y. Zhu, K. Tomsovic, “Development of models for analyzing the load-
following performance of microturbines and fuel cells”, Electric Power
Systems Research, 2002.
[6] J. Padullés, G.W. Ault, J. R. McDonald, “An integrated SOFC plant
dynamic model for power systems simulation”, Journal of Power
Systems, 2000.
[7] N. Hatziargyriou et al, “Modelling of Micro-Sources for Security
Studies”, in CIGRE Session, 2004.
[8] S. Barsali, M. Ceraolo, P. Pelacchi, “Control techniques of Dispersed
Generators to improve the continuity of electricity supply”, Proceedings
of PES Winter Meeting, pp. 789-794, 2002.
[9] A. Engler, “Regelung von Batteriestromrichtern in modularen und
erweiterbaren Inselnetzen”, PhD dissertation, Berlin, 2002.
[10] N. Jayawarna, “Task TE2 – Fault Current Contribution from
Converters”, MicroGrids draft report for task TE2, 2004.
[11] R. Lasseter, P. Piagi, “Providing Premium Power through Distributed
Resources”, in Proc. of the 33
Hawaii International Conference on
System Science, vol. 4, 2000.
[12] A. Madureira, C. Moreira, J. Peças Lopes, “Secondary Load-Frequency
Control for MicroGrids in Islanded Operation”, in Proc. International
Conference on Renewable Energy and Power Quality ICREPQ’05,
Spain, 2005.
[13] J. Peças Lopes, C. Moreira, A. Madureira, F. Resende, et al. “DD1 –
Emergency Strategies and Algorithms”, MicroGrids project Deliverable
DD1, October 2004.
J. A. Peças Lopes is an Associate Professor with Aggregation at the Dept. of
EE of the Faculty of Engineering of Porto University. He obtained an EE
degree (5 years course) in 1981 from University of Porto and a PhD. degree
also in EE from the same University in 1988. In 1996 he got an Aggregation
degree. In 1989 he joined the staff of INESC as a senior researcher and he is
presently co-Coordinator of the Power Systems Unit of INESC Porto and a
senior member of the IEEE.
C. L. Moreira was born in 1980, in Portugal. He received the Electrical
Engineer degree (5 years course) in 2003 from the Engineering Faculty of
Porto University (FEUP). In 2003 he joined the staff of the Power Systems
Unit of INESC Porto as a junior researcher. Currently he is with INESC Porto
and FEUP as a PhD student. His research interests focus the integration of
dispersed generation to low voltage grids.
A. G. Madureira was born in 1980 in Portugal. He received the Electrical
Engineering degree (5 years course) in 2003 from the Faculty of Engineering
of Porto University (FEUP). He joined INESC Porto in the Power Systems
Unit in 2004 as a junior researcher. He is currently with INESC and FEUP as
a MSc student. His main interests include micro-generation and the integration
of dispersed generation to low voltage grids.
... Line impedances and DG impedances significantly affect the reactive power sharing during the operating mode connected to the network and during the island mode due to voltage drops [21]. At present, the voltage controllers in the MGs are unable to share the demand for reactive power among even identical inverters operating in parallel [25]. Some researchers have previously worked on this issue as in [22], which proposes an alternative controller for reactive power sharing between parallel inverters with nominal voltages. ...
... The linearized small-signal state space models of the current controller loop are presented in (23)- (25): ...
Full-text available
This paper presents a new autonomous effective power distribution control strategy for three-phase parallel inverters. The proposal uses a controller that can provide the system with accurate power sharing among distributed generators installed in the microgrid once some load variations are presented in the network. The methodology uses a virtual current loop introduced into the current controller of the inverter to optimize the output signal, which goes directly to the PWM. This virtual current is obtained by using a virtual impedance loop. Furthermore, a small-signal model of the system is used to check stability of the proposed control strategy, which was developed for island mode operation of the microgrid. Simulations were performed for a microgrid with two generators and a load with five households and implemented in MATLAB/Simulink software. The results show that the model provides a wide margin of stability and a rapid response when electrical loads change, thus fulfilling the reactive power sharing among generators. The proposed method shows a large margin of stability and a rapid transient response of the system.
... The MG can be defined as a system formed by energy sources, storage, and loads (possibly controllable), managed by a local supervisory controller (LSC) with the ability to operate connected to the grid or as an island [4]. From the point of view of the network, the main advantage is that the MG can be regarded as a manageable entity within the power system, operated as a single aggregated load with the potential to participate in the provision of ancillary services to the utility [5]. ...
... The basic control objective in a MG is to achieve accurate power sharing while maintaining close regulation of the MG voltage magnitude and frequency. Centralized control of a MG, based on a communication infrastructure, is investigated in [7]- [9]. However, in MGs, in remote areas with long connection distance between inverters, it is impractical and costly to distribute the dynamic sharing signals, which are characterized by their high bandwidth. ...
... There are two types of control mode for the islanded MGs [7]. One is the master-slave control mode [8], [9], where the main control unit switches from PQ control to V-f control to provide frequency and voltage support for the MG. ...
Full-text available
A distributed secondary control (DSC) strategy that combines $Q$ -learning and pinning control is originally proposed to achieve a fully optimal DSC for droop-controlled microgrids (MGs). It takes advantages of cross-fusion of the two algorithms to realize the high efficiency and self-adaptive control in MGs. It has the following advantages. Firstly, it adopts the advantages of reinforcement learning in autonomous learning control and intelligent decision-making, driving the action value of pinning control for feedback adaptive correction. Secondly, only a small part of points selected as pinned points needs to be controlled and pre-learned, hence the actual control problem is transformed into a synchronous tracking problem and the installation number of controllers is further reduced. Thirdly, the pinning matrix can be modified to adapt to plug-and-play operation under the distributed control architecture. Finally, the effectiveness and versatility of the proposed strategy are demonstrated with a typical droop-controlled MG model.
Full-text available
The objective of this paper is to present novel control strategies for MicroGrid operation, especially in islanded mode. The control strategies involve mainly the coordination of secondary load-frequency control by a MicroGrid Central Controller that heads a hierarchical control system able to assure stable and secure operation when the islanding of the MicroGrid occurs and in load-following situations in islanded mode.
Conference Paper
Full-text available
The objective of this paper is to present novel control strategies for MicroGrid operation, especially in islanded mode. The control strategies involve mainly the coordination of secondary load-frequency control by a MicroGrid Central Controller that heads a hierarchical control system able to assure stable and secure operation when the islanding of the MicroGrid occurs and in load-following situations in islanded mode.
Conference Paper
Full-text available
The main objective of this paper is to present the development of microsource modelling and the definition of control strategies to be adopted to evaluate the feasibility of operation of a microgrid when it becomes isolated. Normally, the microgrid operates in interconnected mode with the MV network, however scheduled or forced isolation can take place. In such conditions, the microgrid must have the ability to operate stably and autonomously. An evaluation of the need of storage devices and load-shedding strategies is included in the paper.
Evolutionary changes in the regulatory and operational climate of traditional electric utilities and the emergence of smaller generating systems such as microturbines have opened new opportunities for on-site power generation by electricity users. In this context, distributed energy resources (DER)--small power generators typically located at users' sites where the energy (both electric and thermal) they generate is used--have emerged as a promising option to meet growing customer needs for electric power with an emphasis on reliability and power quality. The portfolio of DER includes generators, energy storage, load control, and, for certain classes of systems, advanced power electronic interfaces between the generators and the bulk power provider. This white paper proposes that the significant potential of smaller DER to meet customers' and utilities' needs can be best captured by organizing these resources into MicroGrids.
This report describes test measurements of the behavior of two microturbine generators (MTGs) under transient conditions. The tests were conducted under three different operating conditions: grid-connect; stand-alone single MTG with load banks; and two MTGs running in parallel with load banks. Tests were conducted with both the Capstone 30-kW and Honeywell Parallon 75-kW MTGs. All tests were conducted at the Southern California Edison /University of California, Irvine (UCI) test facility. In the grid-connected mode, several test runs were conducted with different set-point changes both up and down and a start up and shutdown were recorded for each MTG. For the stand-alone mode, load changes were initiated by changing load-bank values (both watts and VARs). For the parallel mode, tests involved changes in the load-bank settings as well as changes in the power set point of the MTG running in grid-connect mode. Detailed graphs of the test results are presented. It should be noted that these tests were done using a specific hardware and software configuration. Use of different software and hardware could result in different performance characteristics for the same units.
Aluminum-free strained-quantum-well In(0.2)Ga(0.8)As lasers employing novel two-stepped InGaAsP confinement layers and InGaP cladding layers on a GaAs substrate are demonstrated for the first time. Threshold current density as low as 58 A/sq cm is obtained with broad stripe lasers. This threshold current density is, to the authors' knowledge, the lowest reported for 0.98-micron lasers grown by organic chemical vapor deposition.
A scheme for controlling parallel connected inverters in a stand-alone AC supply system is presented. A key feature of this scheme is that it uses only those variables which can be measured locally at the inverter, and does not need communication of control signals between the inverters. This feature is important in high reliability uninterruptible power supply (UPS) systems, and in large DC power sources connected to an AC distribution system. Real and reactive power sharing between inverters can be achieved by controlling two independent quantities at the inverter: the power angle and the fundamental inverter voltage magnitude
Conference Paper
This paper presents dynamic analysis of an isolated distribution network. This study has been carried out in a project looking at the consequences of a high penetration of distributed energy resources (DER) and distributed generation (DG). A network of a village (200 customers) in isolated mode is used for the study with DER connected to network via inverters such as microturbines, fuel cells and storage device. Several scenarios were used to dynamic simulations such as load and generation variation, starting of LV motor and short circuit. Influences of various coordinated controls and operating modes on dynamic behaviours are investigated. The results of this study can be used to specify the service systems in the future.
The design process of a SOFC plant dynamic model for a power systems simulation (PSS) commercial software package has revealed the trade-off between the satisfaction of the network dynamic requirements and a safe and durable cell operation that the plant controller should implement. This paper describes the initial fuel cell stack and power conditioner modelling methodologies that have addressed such issues.