Ghosh, Arindam and Majumder, Ritwik and Ledwich, Gerard F. and Zare, Firuz (2009) 
Power quality enhanced operation and control of a microgrid based custom power park. 
  In: 7th IEEE International Conference on Control & Automation (ICCA'09), December 9‐
11, 2009, New Zealand, Christchurch. (In Press) 

          © Copyright 2009 IEEE 

©2009 IEEE. Personal use of this material is permitted. However, permission

to reprint/republish this material for advertising or promotional purposes or
for creating new collective works for resale or redistribution to servers or
lists, or to reuse any copyrighted component of this work in other works must
be obtained from the IEEE. 
 Power Quality Enhanced Operation and Control
of a Microgrid based Custom Power Park
Arindam Ghosh, Fellow, IEEE, Ritwik Majumder, Student Member, IEEE,
Gerard Ledwich, Senior Member, IEEE and Firuz Zare, Senior Member, IEEE

Abstract − This paper describes the operation of a microgrid ics and are somewhat unreliable [7]. However, a power elec-
that contains a custom power park (CPP). The park may contain tronic converter can mitigate harmonic and unbalanced load or
an unbalanced and/or nonlinear load and the microgrid may source problems. In [8] a compensator is added with each in-
contain many distributed generators (DGs). One of the DGs in dividual DG system, which is controlled to achieve an en-
the microgrid is used as a compensator to achieve load compensa-
hancement of both the quality of power within the microgrid
tion. A new method is proposed for current reference generation
for load compensation, which takes into account the real and and the utility grid. The compensator is connected both in se-
reactive power to be supplied by the DG connected to the com- ries and shunt. The series element can compensate for all the
pensator. The real and reactive power from the DGs and the util- unwanted positive-, negative-, and zero-sequence voltage dur-
ity source is tightly regulated assuming that dedicated communi- ing any utility grid voltage unbalance, while the shunt element
cation channels are available. Therefore this scheme is most suit- is controlled to ensure balanced voltages within the microgrid
able in cases where the loads in CPP and DGs are physically lo- and to regulate power sharing among the parallel-connected
cated close to each other. The proposal is validated through ex- DG systems.
tensive simulation studies using EMTDC/PSCAD software pack- All customers in a custom power park benefit from high-
age (version 4.2).
quality power supply which is superior to the normal power
Index Terms: Distributed generation, microgrid, power quali- supply from a utility [9]. The electrical power to the park is
ty, load compensation and power sharing. supplied through two feeders from two independent substa-
tions. Both these feeders are coupled together through a solid
I. INTRODUCTION state transfer switch that can perform a seamless sub-cycle
transfer from the preferred feeder to the alternate feeder. In
MICROGRID contains a cluster of loads and distributed
A generators operating as a single controllable system to
provide power to its local area. Many of the DGs are inter-
addition, a power-converter-based device is used to eliminate
harmonic current and/or to correct voltage sag/swell [10-11].
In this paper, the concept of microgrid and custom power
faced to the microgrid through power electronic converters, are combined together to come up with a microgrid that share
which are controlled to deliver the desired active and reactive their local load power requirement with the utility. Moreover,
power to the system. The most common method used for this one the DGs is used as a compensator to cancel the effects the
purpose is the droop control for load sharing without any harmonics and unbalance in the local load such that the PCC
communication, which can be achieved by controlling two voltage is balanced and distortion-free. The other DGs in the
independent quantities – the power angle, and the fundamental system are used a voltage sources providing part of the power
voltage magnitude [1-4]. Local signals are used as feedback to requirement. The compensating DG is interfaced through a
control to the DGs. voltage source converter (VSC), while the remaining DG can
The Consortium for Electric Reliability Technology be either inertial or inertia-less. The real and reactive power
Solutions (CERTS) microgrid concept assumes a relative through the DGs and the utility source is tightly regulated as-
small system in a localized site, where the microgrid can suming that dedicated communication channels are available.
seamlessly separate from the utility grid during contingencies The efficacy of the proposed scheme is validated through ex-
where the microgrid is separated from the utility and continues tensive simulation studies using EMTDC/PSCAD software
to supply its own local loads [1, 5]. The CERTS microgrid package (version 4.2).
assumes a single point of common coupling (PCC) with the
utility. In this, the DGs are controlled locally by power elec- II. SYSTEM STRUCTURE
tronic devices incorporating the droop characteristics. An in-
teresting feature of the CERTS microgrid is that a DG can be The basic power system model with two DG sources is
inserted in a plug-and-play fashion and not all the DGs are shown in Fig. 1 in which the real and reactive power
required at any given time making the microgrid completely drawn/supplied are denoted by P and Q respectively. The mi-
flexible. A review of various microgrid projects is given in crogrid is connected to the utility grid at the point of common
[6]. coupling (PCC). In Fig. 1, DG-1 is connected through the
The power quality issues are important for DG interconnec- compensator and is connected directly to the microgrid
tions as the power electronic converters can generate harmon- through circuit breaker CB-3, while injecting P1 and Q1. DG-2
is connected to the microgrid through CB-4 and an inductor
A. Ghosh, R. Majumder, G. Ledwich and F. Zare are with the School of L2. It injects P2 and Q2. The microgrid supplies an unbalanced
Engineering Systems, Queensland University of Technology, Brisbane, Qld and nonlinear load that draws PL and QL. The utility supply is
4001, Australia.
Email: a.ghosh@qut.edu.au denoted by vs and the feeder resistance and inductance are
denoted respectively by Rs and Ls. The utility supplies Pg and III. REFERENCE GENERATION AND COMPENSATOR CONTROL
Qg to the microgrid and the balance Ps − Pg and Qs − Qg are
supplied to the utility load, which is shown as motor in Fig. 1. In this section, the reference generation and for both the
The breakers CB-1 and CB-2 can isolate the microgrid and the compensator and DG-2 is presented. Also the control issues of
utility load respectively from the supply. the compensator are presented.

A. Compensator Reference Generation

The main aim of the compensator is to cancel the effects of

unbalanced and harmonic components of the load. In [12], an
algorithm is proposed which assumes that the VSC-based
compensator is supplied by a dc capacitor. Hence it can only
supply a portion of the load reactive power and no active pow-
er. However, DG-1 in Fig. 1 must supply pre-specified amount
of real and reactive powers to the load. Hence the algorithm of
[12] has to be modified.
If proper compensation is achieved, the currents ig and i2
will be balanced and so will be the voltage vp provided that vs
is balanced. Let us denoted the three phases by the subscripts
a, b and c. Therefore since ig and i2 are balanced we have
iga + igb + igc = 0
(1)
i2 a + i2b + i2c = 0
From Fig. 1, the Kirchoff’s current law (KCL) at PCC gives
i1k + i2 k + igk = iLk , k = a, b, c (2)

Therefore combining (1) and (2) by adding the currents of the

Fig.1. The microgrid and utility system under consideration. all the three phases together, we get

The structure of the compensator is shown in Fig. 2. In this i1a + i1b + i1c = iLa + iLb + iLc (3)
DG-1 is assumed to be an ideal dc voltage source supplying a
voltage of Vdc1 to the compensator. The compensator contains Since ig and i2 balanced due to the action of the compensa-
three H-bridge converters. The outputs of the H-bridges are tor, the voltage vp will also become balanced. Hence the in-
connected to three single-phase transformers that are con- stantaneous real powers Pg and P2 will be equal to their aver-
nected in wye for required isolation and voltage boosting [12]. age components. Therefore we can write
The resistance Rf represents the switching and transformer ( ) ( ) ( )
v pa i ga + i2 a + v pb igb + i2b + v pc igc + i2c = Pg + P2 (4)
losses, while the inductance Lf represents the leakage reac-
tance of the transformers. The filter capacitor Cf is connected From the KCL of (2), (4) can be written as
to the output of the transformers to bypass switching harmon-
ics. v pa (iLa − i1a ) + v pb (iLb − i1b ) + v pc (iLc − i1c ) = Pg + P2 (5)

Similarly the reactive powers Qg and Q2 will be equal to

their instantaneous components. Therefore we can write
(v pb − v pc )iga + (v pc − v pa )igb + (v pa − v pb )igc = 3 × Qg
(6)
(v pb − v pc )i2a + (v pc − v pa )i2b + (v pa − v pb )i2c = 3 × Q2

Using the KCL of (2), (6) can be rewritten as

(v pb − v pc )(iLa − i1a ) + (v pc − v pa )(iLb − i1b )
+ (v pa − v pb )(iLc − i1c ) = 3 (Qg + Q2 )
(7)

Equations (3), (5) and (7) form the basis of the algorithm.
From these three, the following can be written

⎡i1a ⎤ ⎡iLa ⎤ ⎡ 0 ⎤
⎢ ⎥
A⎢⎢i1b ⎥⎥ = ⎢ ⎥ (
A⎢iLb ⎥ + ⎢ − Pg + P2 ) ⎥ (8)
Fig. 2. DG-1 with the compensator structure. ⎢⎣i1c ⎥⎦
⎣ (
⎢⎣iLc ⎥⎦ ⎢− 3 Qg + Q2 )
⎥
⎦
where regulator (LQR) design. From uc(k), the switching function is
generated as
⎡ 1 1 1 ⎤
⎢ ⎥ If uc (k ) > h then u = +1
A = ⎢ v pa v pb v pc ⎥ (15)
elseif uc (k ) < −h then u = −1
⎢ v pb − v pc v pc − v pa v pa − v pb ⎥⎦
⎣
where h is a small number.
The determinant of the matrix A is given by With respect to Fig. 3, the state vector is chosen as
( ) (
A = v pa v pc + v pb − 2v pa + v pb v pa + v pc − 2v pb ) [
xT = i1 icf vp ] (16)
(
+ v pc v pb + v pa − 2v pc ) (9)
It is stipulated that the compensator does not regulate the PCC
If vp is balanced, then the following is true voltage. Therefore the positive sequence of the fundamental of
the PCC voltage is extracted as in [13] and it is used as the
v pa + v pb + v pc = 0 (10) reference for vp. Also note that the PCC voltage vp is the same
as the voltage across the filter capacitor vcf. Hence the refer-
Substituting (10) in (9), the determinant of A is given as ence for the current icf can easily be computed by phase shift-
(
A = −3 v 2pa + v 2pb + v 2pc ) (11)
ing the reference for vp by 90° and from the knowledge of the
value of the filter capacitor Cf.
The reference for the current i1 can be obtained from (12).
In similar manner, the solution of (8) is computed as
However we stipulate that DG-1 supplies a fraction of the real
⎡i1a ⎤ ⎡iLa ⎤ ( ) (
⎡3 Pg + P2 v pa + 3 Qg + Q2 v pb − v pc ⎤
⎢ ⎥
)( ) and reactive power demanded by the load. Then noting that
⎢i ⎥ = ⎢i ⎥ + 1
( )
⎢ 1b ⎥ ⎢ Lb ⎥ A ⎢3 Pg + P2 v pb + 3 Qg + Q2 v pc − v pa ⎥( )( ) Pg + P2 = PL − P1
( ) ( )( )
⎢3 P + P v + 3 Q + Q v − v ⎥ (17)
⎢⎣i1c ⎥⎦ ⎢⎣iLc ⎥⎦ Qg + Q2 = QL − Q1
⎢⎣ g 2 pc g 2 pa pb ⎥
⎦
(12) Also note that Pg + P2 is a dc quantity, while PL will contain a
double frequency component due to the unbalance in the load.
B. Compensator Control It will also be distorted due to the nonlinear nature of the load.
Hence it is evident from (17) that P1 should cancel out both
The equivalent circuit of one phase of the converter is these components. Similarly, Q1 should also cancel out both
shown in Fig. 5. In this, u⋅Vdc1 represents the converter output the double frequency and nonlinear component of QL.
voltage, where u = ± 1. The main aim of the converter control Now let us stipulate that P1 supplies λ1P times the average
is to generate u. From the circuit of Fig. 1, a state space de- load power PLav and Q1 supplies λ1Q times the average load
scription of the system can be given as reactive power QLav. Then we can rewrite (17) as
x& = Ax + Buc (
Pg + P2 = PL − λ1 p × PL = PLav 1 − λ1 p = PL1)
where uc is the continuous time version of switching function (
Qg + Q2 = QL − λ1Q × QL = QLav 1 − λ1Q = QL1 ) (18)

u. The discrete-time equivalent of the state equation is

We can then modify (12) as to get the following reference
x(k + 1) = Fx(k ) + Guc (k ) (13) currents for i1

⎡i1a ⎤ ⎡iLa ⎤ (
⎡3PL1v pa + 3QL1 v pb − v pc
⎢
)⎤⎥
⎢ i ⎥ = ⎢i ⎥ + 1
(
⎢ 1b ⎥ ⎢ Lb ⎥ A ⎢3PL1v pb + 3QL1 v pc − v pa )⎥ (19)
⎢⎣i1c ⎥⎦ ⎢⎣iLc ⎥⎦ ⎢3P v + 3Q v − v
⎢⎣ L1 pc L1 pa ( pb )⎥⎥⎦
C. Reference Generation for DG-2

With respect to Fig. 1, let us define the rms voltage of the

PCC as |Vp|∠δp and that of the DG-2 back emf as |V2|∠δ2. Al-
so let us define the DG-2 source impedance as X2 = ωL2. Then
the real and reactive power injected to the microgrid by DG-2
Fig. 3. Equivalent circuit of one phase of the converter.
is given by
The control law is then given by VP V2 sin δ
[
uc (k ) = − K x(k ) − xref (k ) ] (14)
P2 =
X2
(20)

2
where K is a gain matrix and xref is the reference vector. The VP V2 cos δ − V p
gain matrix, in this paper, is obtained through linear quadratic Q2 = (21)
X2
where δ = δ2 − δp. Let us assume that P2 is equal to λ2P times the load, i.e., λ1Q = 1 and λ2Q = 0. Furthermore the real power
the average load power PLav and Q2 is equal to λ2Q times the sharing between the two DGs are set as λ1P = 0.2 and λ2P =
average load reactive power QLav. Then we have 0.3.
The load currents are shown in Fig. 4 (a). In this figure and
P2 = λ2 p × PLav all the subsequent figures showing three-phase quantities, the
(22)
Q2 = λ2Q × QLav phases a, b and c are plotted with solid, dashed and dotted
Equations (20) and (21) can be written as lines respectively. It can be seen that they are both unbalanced
and distorted. The PCC voltages are shown in Fig. 4 (b). They
λ2 p PLav X 2 = VP V2 sin δ attain the steady state within 0.1 s (i.e., 5 cycles). The real and
2
reactive powers are shown in Fig. 5 (a) and (b) respectively. It
λ2Q QLav X 2 + V p = VP V2 cos δ can be seen that Pg and P2 attain constant values in the steady
state in accordance to their power share, i.e., 50% and 30%
From the above equation δ is given by respectively. The compensator supplies 20% power as well as
the nonlinear, unbalanced component of the load. The reactive
⎛ ⎞
−1 ⎜ λ2 p PLav X 2
⎟ powers supplied by the source and DG-2 are zero and hence
δ = tan ⎜ 2 ⎟
(23) the compensator supplies the entire reactive power demand of
⎜ λ2Q QLav X 2 + V p ⎟ the load, as evident from Fig. 5 (b).
⎝ ⎠
Then the magnitude of the back emf is calculated from (21) as
2
λ2Q QLav X 2 + V p
V2 = (24)
VP cos δ

IV. SIMULATION STUDIES

Simulation studies are carried out in PSCAD/EMTDC (ver-

sion 4.2). In this, DG-2 is considered to be an inertia-less dc
source supplied through a VSC. The system data, excluding
DG-2 are given in Table I.
TABLE I. SYSTEM PARAMETERS
System Quantities Values
Systems frequency 50 Hz Fig. 4. The load currents and compensated PCC voltages.
Source voltage (Vs) 11 kV rms (L-L)
Feeder impedance Rs = 3.025 Ω, Ls = 57.75 mH
Unbalanced load RLa = 24.2 Ω, LLa = 192.6 mH
RLb = 12.2 Ω, LLb = 100.0 mH
RLc = 48.2 Ω, LLc = 300.0 mH
Nonlinear load A three-phase rectifier supplying an
RL load with R = 200 Ω, L = 100 mH
Motor load (M) Induction motor rated 20 hp, 11 kV
rms (L-L).
DG-1 and Compesator
DC voltage (Vdc1) 3.5 kV
Transformer rating 3 kV/11 kV, 0.5 MVA, 2.5% reactance
(Lf)
VSC losses 1.5 Ω
Filter Capacitance (Cf) 50 μF

The structure of DG-2 is the same as shown in Fig. 2 except

that and output inductance L2 is connected between the filter
capacitor Cf and the microgrid. It is to be noted that the vol- Fig. 5. Real and reactive power sharing.
tage v2 is the voltage across the filter capacitor. The parame-
ters of DG-2 and the connecting VSC are chosen to be the The steady state waveforms phase-a of the PCC voltage and
same as DG-1, given in Table I. The value of the inductance currents ig and i2 are shown in Fig. 6 (a) and (b) respectively.
L2 is chosen as 72.2 mH. In this, current ig is scaled up 40 times while the current i2 is
The instantaneous version of the voltage reference for DG- scaled up 80 times. It can be seen that both these currents are
2, obtained from (23) and (24) are now tracked using the VSC balanced and the a-phases of the currents are in phase with the
connected to it through an output feedback voltage controller phase-a of the PCC voltage. It also proves the source side and
[11]. It is assumed that the breaker CB-2 is not connected, i.e., DG-2 are injecting unity power factor currents.
the induction machine is not in operation. It is desired that the Let us now assume the following real and power sharing:
compensator supplies the entire reactive power requirement of λ1P = 0.2, λ2P = 0.3, λ1Q = 0.5 and λ2Q = 0.2. The breaker CB-2
still remains open. With the system operating in the steady distortion in the PCC voltages. The torque pulsation however
state, the breaker CB-1 opens suddenly at 0.0511 s, islanding stops and the PCC voltages become balanced sinusoids when
the microgrid. The breaker recloses after 0.2109 s. When the the compensator gets connected.
islanding occurs, the PCC voltage tends to fall and hence the
angle difference δ tends to run away causing total voltage col-
lapse. To prevent this, a rapid island detecting scheme is in-
troduced. The instantaneous power pg injected by the grid is
computed from the following relation
p g = v pa iga + v pbigb + v pc i gc (25)

Fig. 7. (a) RMS PCC voltage and (b) instantaneous grid power during island-
ing and resynchronization.

Fig. 6. Unity power factor operation.

Once this power falls below a threshold value, the reference

for the PCC voltage is held at its previous value till the island-
ing condition is removed. The results are shown in Figs. 7 and
8. The rms PCC voltage is shown in Fig. 7 (a), while pg is
shown in Fig. 7 (b). It can be seen that there is a time lag of
0.0057 s (i.e., about one fourth of a cycle) between the instant
of islanding and its detection. This causes a slight dip in the
PCC voltage. As a consequence of this drop, there is a slight
drop in the load real and reactive powers are shown in Fig. 8 Fig. 8. Real and reactive powers during islanding and resynchronization.
(a) and (b) respectively. During the islanding, both the DGs
share the real and reactive power equally as evident from this
figure. Note that Pg and Qg do not go to zero instantaneously
due to the smoothing time constant of the power measurement
block of PSCAD. Therefore for the islanding detection (25) is
needed. Once the islanding is cleared, the system quantities
return to their nominal values as evident from Figs. 7 and 8.
It is to be noted that it has been assumed that the utility
network configuration has not changed during the islanding.
Hence, (25) is sufficient for resynchronization since the PCC
voltage drop and its phase shift during islanding are minimal.
A more rigorous resynchronization process and the systems
response for faults in the utility grid and inside the microgrid
will be will be investigated in future.
Let us now assume that CB-2 is now closed and the values
of the passive components of the unbalanced and nonlinear
load are twice of those given in Table I. We also assume that Fig. 9. (a) Motor torque and (b) PCC voltages before and after compensator
connection.
the breakers CB-3 and CB-4 are open such that the utility
source supplies both the motor and microgrid loads. These two V. DISCUSSIONS
breakers are closed simultaneously at 0.3 s. Per unit motor
torque and the PCC voltages are shown in Fig. 9 (a) and (b) In the studies presented in the paper it has been assumed
respectively. It can be seen that large torque pulsations occur that only two DGs and a load are connected to the microgrid.
when the compensator is not connected due to unbalance and In general, however, there might be several DGs and loads
connected to the microgrid as shown in Fig. 10. In this there ACKNOWLEDGEMENT
are a total number of n loads and m DGs, with the compensa-
tor being DG-1. A command and control center (CCC) con- The authors thank the Australian Research Council (ARC)
trols the power flow in the microgrid. for the financial support for this project through the ARC Dis-
The total active and reactive powers consumed by the loads covery Grant DP 0774092.
are given by
REFERENCES
PL = PL1 + PL 2 + L + PLn
(26) [1] F. Katiraei and M. R. Iravani, “Power management strategies for a mi-
QL = QL1 + QL 2 + L + QLn crogrid with multiple distributed generation units,” IEEE Trans. on
Power Systems, Vol. 21, No. 4, pp. 1821-1831, 2006.
The command and control center then finds the averages of PL [2] J. M. Guerrero, L. G. de Vicuna, J. Matas, M. Castilla, and J. Miret, “A
and QL. Based on these averages (PLav, QLav) and the rating of wireless controller to enhance dynamic performance of parallel inverters
in distributed generation systems,” IEEE Trans. on Power Electronics,
each DG, CCC then determines the load sharing ratio (λp and Vol. 19, No. 5, pp. 1205-1213, 2004.
λq) for each DG. They are then sent to the controller for each [3] M. C. Chandorkar, D. M. Divan and R. Adapa, “Control of parallel
DG to calculate the individual references, with (19) being used connected inverters in standalone ac supply systems,” IEEE Trans. On
Industry Applications, Vol. 29, No. 1, pp. 136-143, 1993.
by DG-1 along with the measurement of PCC voltage and (23) [4] Consortium for Electric Reliability Technology Solutions (CERTS),
and (24) being used by the other DGs. “Integration of distributed energy resource: The CERTS microgrid con-
cept,” Consultant Report, California Energy Commission, October 2003,
www.osti.gov/bridge/servlets/purl/799644-dfXsZi/native/799644.PDF.
[5] R. H. Lasseter, "MicroGrids," IEEE Power Engineering Society Winter
Meeting, Vol. 1, pp. 305-308, 2002.
[6] N. Hatziargyriou, H. Asano, M. R. Iravani and C. Marnay, “Microgrids,”
IEEE Power & Energy Magazine, Vol. 5, No. 4, pp. 78-94, 2007.
[7] R. C. Dugan, M. F. McGranaghan, S. Santoso and H. W. Beaty, Elec-
trical Power Systems Quality, 2nd Ed., McGraw-Hill, New York, 2002.
[8] Y. Li, D. M. Vilathgamuwa, and P. C. Loh, “Microgrid power quality
enhancement using a three-phase four-wire grid-interfacing compensa-
tor,” IEEE Trans. on Industry Applications, Vol. 41, No. 6, pp. 1707-
1719, 2005.
[9] N.G. Hingorani, “Custom power and custom power park,” Flexible
Power HVDC Transmission and Custom Power, CIGRE Australian Pan-
el 14, Sydney, Australia, 1999.
[10] A. Ghosh and A. Joshi, “The concept and operating principles of a mini
Fig. 10. Microgrid structure with command and control center. custom power park,” IEEE Trans. on Power Delivery, Vol. 19, No. 4,
pp. 1766-1774, 2004.
VI. CONCLUSIONS [11] A. Ghosh, “Performance study of two different compensating devices in
a custom power park,” Proc. IEE − Generation, Transmission & Distri-
bution, Vol. 152, No. 4, pp. 521-528, 2005.
The operation of a power quality enhanced microgrid is dis- [12] A. Ghosh and A. Joshi, “A new approach to load balancing and power
cussed in this paper. In this scheme one of the DGs acts as a factor correction in power distribution system,” IEEE Trans. on Power
load compensator to balance the PCC voltage even if some or Delivery, Vol. 15, No. 1, pp. 417-422, 2000.
all the loads in the microgrid are unbalanced and/or distorted. [13] A. Ghosh and G. Ledwich, “Load compensating DSTATCOM in weak
ac systems,” IEEE Trans. Power Delivery, Vol. 18, No. 4, pp. 1302-
The compensating DG has to be a VSC based to achieve load 1309, 2003.
compensation, while the other DGs can be either inertial or
inertia-less. A new algorithm is proposed for compensator
reference current generation, which can provide load compen-
sation, while supplying a desired amount active and reactive
power simultaneously. The load sharing arrangement has been
shown to work satisfactorily even in the case of an inertial DG
system. Through simulation studies, it has been shown that the
VSC-based compensating DG and the inertial-less DG do not
interact together to pose any stability problem.
The tight load sharing arrangement has been achieved by
assuming the presence of dedicated communication channels.
This is viable when the microgrid constitute a small geograph-
ical area when all the loads and DGs are placed. Hence in the
paper, the line impedance is assumed to be negligible. It is
however to be noted that the proposed algorithm will work
satisfactorily even in the presence of a small line impedance,
provided that the line impedance values are incorporated in the
power sharing arrangement.