Micro-grid operation of inverter based distributed

generation with voltage and frequency dependent loads

The MIT Faculty has made this article openly available. Please share
how this access benefits you. Your story matters.

Citation

Zeineldin, H. H., and J.L. Kirtley. Micro-grid Operation of Inverter

Based Distributed Generation with Voltage and Frequency
Dependent Loads. Proceedings of the IEEE Power & Energy
Society General Meeting, (PES '09), 2009. 16. Copyright
2009 IEEE

As Published

http://dx.doi.org/10.1109/PES.2009.5275831

Publisher

Institute of Electrical and Electronics Engineers (IEEE)

Version

Final published version

Accessed

Fri Jun 13 00:58:39 EDT 2014

Citable Link

http://hdl.handle.net/1721.1/72986

Terms of Use

Article is made available in accordance with the publisher's policy

and may be subject to US copyright law. Please refer to the
publisher's site for terms of use.

Detailed Terms

Micro-grid Operation of Inverter Based

Distributed Generation with Voltage and
Frequency Dependent Loads
H. H. Zeineldin, IEEE Member, IEEE, and J.L. Kirtley, Jr., Fellow IEEE

Abstract Distribution systems are experiencing increasing

penetration of Distributed Generation (DG). One attractive
option is to use the available DG capacity during utility outages
by forming planned micro-grids. Load sharing among different
DG, during micro-grid operation, could be accomplished by
equipping each DG with a P-f and Q-V droop characteristic. In
this paper, we analyze the impact of the loads voltage and
frequency dependence on the micro-grids frequency and voltage
deviation during a planned islanding condition. The system was
modeled and analyzed using PSCAD/EMTDC. The simulation
results and mathematical analysis show that the loads voltage
and frequency dependence is an important factor when choosing
the DG droop characteristic. The results also show that some
loading conditions could lead to an unstable micro-grid
operation.

Index Terms Distributed Generation, Inverter, Micro-grid,

Static Load Models.

I. INTRODUCTION

HE Micro-grid could be viewed as a group of DG that

could operate either connected to the grid or intentionally
islanded. Although both the IEEE Std. 929 and UL 1741
standard do not allow islanded operation; the IEEE 1547
standard presents it as one of the topics that needs to be
addressed in the future[1]-[3]. Micro-grids can offer many
advantages which include increase system reliability, ability to
isolate during utility outage, and reduce need for central
dispatch [4].
Distributed Generation are designed to operate close to
unity power factor as per the IEEE Std. 1547 [3]. It would not
be possible to operate a micro-grid with such control [4]. The
operation of a micro-grid requires a robust control strategy
that would facilitate load sharing between different DG units.
Two main strategies have been proposed in literature for
controlling DG on a micro-grid which include the
Master/Slave approach and the droop control approach. In the
master/slave approach, one DG (master) operates as a voltage

This work was supported by the Masdar Institute of Science and

Technology, Abu Dhabi, UAE.
H. H. Zeineldin is with the Masdar Institute of Science and Technology,
Abu Dhabi, UAE (e-mail: [email protected]).
J. L. Kirtley is with the Massachusetts Institute of Technology,
Cambridge, MA

978-1-4244-4241-6/09/$25.00 2009 IEEE

source where it is responsible for controlling the micro-grid

voltage while the other DG (slaves) operate as a current source
[5]-[6]. The DG designed to operate as slaves are responsible
for supplying active power to meet the load demand. The
distributed generators communicate with each other for load
sharing. In the droop control approach, the DG are equipped
with a P-f and Q-V droop characteristic [7]-[9]. The amount of
load supplied by each DG depends on the DG droop. The
main advantage of this technique is that it does not require any
communication links. In [6] and [9], the master slave and
droop approach were both implemented where some DG were
designed to supply a predefined active and reactive power and
others were equipped with a droop characteristic.
In this paper, we focus on the micro-grid load and its
impact on the voltage and frequency deviation during microgrid operation for inverter based DG equipped with droop
controls. The load is modeled as a static load with both
voltage and frequency dependence. The static load model
parameters are varied and both the frequency and voltage
waveforms are examined.
The paper in organized as follows: Section II provides a
brief description of the load model and DG droop
characteristic. Section III presents the micro-grid system and
DG interface model under study. Section IV highlights the
simulation results for the different load model parameters.
Lastly, conclusions are drawn in Section V.
II. MICRO-GRID LOAD AND INVERTER CHARACTERISTICS
The frequency and voltage at which the micro-grid will
stabilize depends on both the load and DG characteristic. Both
the active and reactive power of the load are expressed as
functions of voltage and frequency [10]. The loads voltage
dependence could be modeled through either a polynomial or
exponential relationship. The loads frequency dependence is
represented by multiplying either the polynomial or
exponential load model by a factor. An exponential load
model with both voltage and frequency dependence is given
below in (1) and (2)

P=

V
Po
Vo

NP

(1 + k pf ( f f o ))

(1)

V
= Qo
Vo

NQ

(2)

(1 + k qf ( f f o ))

where Vo represents the initial operating voltage (rated

voltage) and Po and Qo represent the active and reactive
power corresponding to the initial operating voltage. The
parameter f and fo represent the operating and rated frequency
respectively. The parameter V represents the operating
voltage. NP, NQ, kpf and kqf correspond to the parameters of
the load model [10].
As mentioned earlier, the DG is equipped with a P-f and QV droop characteristic as shown in Fig. 1 and Fig. 2. The P-f
characteristic is designed such that the DG delivers its
maximum power at rated frequency. In a similar manner, the
Q-V characteristic is designed such that the DG supplies zero
reactive power at rated voltage. Since both the voltage and
frequency are controlled by the utility during parallel
operation, the DG would operate at unity power factor.
During micro-grid operation, the amount of voltage and
frequency deviation will depend on the amount of active and
reactive power mismatch, the load characteristic and the DG
droop model. The P-f and Q-V droop characteristic could be
mathematically represented as follows
PDG = a1 ( f f o ) + b1

(3)

QDG = a2 (V Vo ) + b2

(4)

where PDG and QDG are the DG active and reactive power
output. Parameters a1, a2, b1, and b2 represent the droop
characteristic parameters. The frequency and voltage at which
the micro-grid will stabilize could be found by equating (1)
and (2) to (3) and (4) respectively as follows
V
Po
Vo

NP

Vo

(1 + k pf ( f f o )) = a1 ( f f o ) + b1

(5)

NQ

(1 + k qf ( f f o )) = a2 (V Vo ) + b2

Qo

(6)

1 p .u

Qmin

Q max

Fig. 2. Q-V Droop Characteristic

III. SYSTEM UNDER STUDY

The system under study consists of a load represented by a
static load model and two inverter based DG as shown in Fig.
3. The distribution system was modeled by an equivalent
source behind an impedance. Each DG is equipped with the
interface control presented in Fig. 4. The frequency and
voltage at the Point of Common Coupling (PCC) are measured
and through the droop characteristic the active and reactive
power reference values (Pref and Qref) are determined. The
power reference values are then compared with the DG output
active and reactive power (PDG and QDG) and the error is
passed to a Proportional Integral (PI) controller to generate the
current d-axis and q-axis reference values (Idref and Iqref).
Through Parks transformation, the DG a-b-c current are
transformed into the d-q axis frame and are compared with the
current d-q axis reference values. The error is passed through a
second PI controller to determine the inverter switching
signals. Table I presents the system, DG and droop
characteristic parameters.

Medium Voltage

It can be seen from (5) and (6) that deviations in voltage are
dependent on deviations in frequency and vice versa. For
constant power loads (NP=NQ=0) with no frequency
dependence (kpf = kqf = 0), the voltage deviation and
frequency deviation are decoupled.

y
c
n
e
u
q
re
f

f3

Standard droop
control

f1

Microgrid Isolating
Point

PCC

Maximum Power
control

AC

AC

DC

DC
DG2

DG 1

Static Load

P1
Fig. 1. P-f Droop Characteristic

P max

Real Power

Fig.3. System under Study.

Low Voltage

3
QDG
V PCC

Qref

K'p +

I DG a
I DGb

f PCC

iq

abc
dq

I DGc
Pref

id
K'p +

K 'I
s

K 'I
s

iqref
+

Kp +

KI
s

Kp +

KI
s

Vqref

Vd ref

id ref

PDG

Fig. 4. DG Interface Control Design.

PCC frequency and voltage. Since the two DG presented in

this work are identical with the same droop characteristic, the
DG share the load equally. During grid connected operation
both DG supply a 100 kW of active power. Since the PCC
voltage is less than 1 p.u during grid connected operation,
each DG will supply 15 kVAR of reactive power which is
dictated by the droop characteristic. From Fig. 5, it can be
seen that the frequency will stabilize at a value of 60 Hz and
the voltage will stabilize at approximately 0.91 per unit. This
result coincides with the mathematical results which are
presented in Table II in the Appendix.

TABLE I
SYSTEM, INVERTER AND DROOP DATA

Grid Parameters

Voltage (line to line)

480V

Frequency

60Hz

Grid Resistance

0.02

Grid Inductance

0.3 mH

DG Inverter Controller Parameters

k p = 5

k I = 0.07

kp = 3

k I = 0.08

a1 = 0.1

b1 = 0.1

a2 = 0.5

b2 = 0

Fig. 5. Active and Reactive power output of each DG with a constant power
load.

The load on the micro-grid was modeled using (1) and (2).
In order to examine the effect of the load model on the microgrid frequency and voltage deviations, the different load
parameters were varied individually while fixing the
remaining load parameters.
IV. SIMULATION RESULTS WITH DIFFERENT LOAD MODELS
The system under study is implemented on the
PSCAD/EMTDC simulator. A micro-grid case is simulated by
opening the utility breaker at the Point of Common Coupling
(PCC) at t = 5 seconds. Both the PCC voltage and frequency
are monitored to determine the effect of load parameters on
the micro-grid voltage and frequency deviations. The next
subsections highlight the simulation results for each case
study.
A. Load Sharing with Droop Control
The DG interface control is tested by opening the switch at
the micro-grid isolating point (refer to Fig. 3). The load is
modeled as a constant power load with no frequency
dependence (NP=NQ=0 and kpf=kqf=0) and micro-grid
operation is initiated at t = 5 seconds. The loads active and
reactive power are 200 kW and 90 kVAR respectively. Figure
5 and Fig. 6 present the DG active power, reactive power,

Fig. 6. Frequency and voltage at the PCC with a constant power load.

B. Effect of Voltage Dependence

In this scenario, the effect of the loads voltage dependence
is analyzed by varying the values of NP and NQ. Similarly,

the loads active and reactive power are adjusted to 200 kW

and 90 kVAR respectively. A micro-grid is initiated at t = 5
seconds. The loads frequency dependence (kpf=kqf=0) is set to
zero. Figure 7 presents the frequency and voltage waveforms
for NP=NQ=2 (constant impedance load), NP=NQ=1
(constant current load) and for NP=NQ=0 (constant power
load). The frequency deviations decrease as the value of NP
and NQ decreases. On the contrary, the voltage deviations
increase as the value of NP and NQ decrease for the presented
cases. The results coincide with the mathematical results
presented in the Appendix.

Fig. 8. Frequency and voltage at the PCC for different values of kpf and kqf
with NP=NQ=2.

Fig. 7. Frequency and voltage at the PCC for different values of NP and NQ.

A second case was simulated where the load is adjusted to

176 kW and 90 kVAR at rated conditions. The loads voltage
dependence parameter is set equal to zero (NP=NQ=0). Figure
9 presents the voltage and frequency deviations for various
values of kpf and kqf. Similarly the frequency deviations
decrease while voltage deviations increase with the increase in
kpf and kqf. Frequency and voltage deviations are dependent on
the load parameters as well as the amount of active and
reactive power mismatch. For constant power loads with no
frequency dependence (C1 and C7 in Appendix), the voltage
deviation is independent on the amount of frequency
deviation.

C. Effect of Frequency Dependence

This scenario focuses on the effect of the loads frequency
dependence on voltage and frequency deviations during
micro-grid operation. The micro-grid operation is analyzed for
different values of kpf and kqf while fixing NP= NQ =2. Figure
8 presents the voltage and frequency waveforms for various
values of kpf and kqf. Similarly, the load was adjusted to
consume 200 kW and 90 kVAR at rated conditions. From Fig.
8, the amount of frequency deviation will decrease as the
value of kpf and kqf increase. On the other hand, the amount of
voltage deviations will increase with the increase in kpf and kqf
(refer to Appendix).

Fig. 9. Frequency and voltage at the PCC for different values of kpf and kqf
with NP=NQ=0.

D. Unstable Operation
In this case, the load was adjusted to consume 176 kW and
90 kVAR at rated conditions. Similarly, the micro-grid
operation was initiated at t = 5 seconds. The loads voltage
dependence parameters was set fixed (NP=NQ=3). Figure 10
presents the frequency and voltage waveforms for different
values of kpf and kqf. For the cases where kpf=kqf=0 and
kpf=kqf=5, the micro-grid was stable and the frequency and
voltage stabilize at values that coincide with those given in
Table II. For the case where kpf and kqf are set equal to -5,
there is no stable micro-grid operating point. The frequency
drifts until it reaches the interface controller limits. Although
case C10, C11, and C12 (refer to Table II) have the same
loading conditions, some case might result in stable operation
while others might result in an unstable operating condition.
Table II shows that, mathematically, there is no solution for
either the frequency or voltage for this specific loading
condition.

mathematical analysis coincides with the simulation results.

VI. APPENDIX
The frequency and voltage at which the micro-grid will
stabilize could be determined from (5) and (6). Table II
presents the values for the frequency and voltage during
micro-grid operation for the loading conditions under study in
this paper.
TABLE II
FREQUENCY AND VOLTAGE FOR DIFFERENT LOADING CONDITIONS
Case
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12

P
(kW)
200
200
200
200
200
200
176
176
176
120
120
120

Q
(kvar)
90
90
90
90
90
90
90
90
90
90
90
90

NP

NQ

kpf

kqf

f (Hz)

V(p.u)

0
1
2
2
2
2
0
0
0
3
3
3

0
1
2
2
2
2
0
0
0
3
3
3

0
0
0
1
3
5
0
2
5
0
5
-5

0
0
0
1
3
5
0
2
5
0
5
-5

60
60.08
60.147
60.085
60.06
60.03
60.12
60.04
60.022
60.52
60.19
No
sol.

0.91
0.917
0.923
0.9177
0.915
0.9129
0.91
0.902
0.9
0.928
0.879
No
sol.

VII. REFERENCES
[1]

Fig. 10. Frequency and voltage at the PCC for different values of kpf and kqf
with NP=NQ=3.

V. CONCLUSIONS
This paper examines the effect of the loads voltage and
frequency dependence on micro-grid operation for inverter
based distributed generation. An exponential load model with
both voltage and frequency dependence was modeled. The DG
interface control was equipped with a P-f and Q-V droop
characteristic. The results and analysis show that the
frequency deviations and voltage deviation during micro-grid
operation are dependent on each other. Although the microgrid load could be within the available DG capacity, unstable
operating conditions could occur as a result of the loads
voltage and frequency dependence. It is important when
designing the DG droop characteristic to take into account
different possible loading conditions and models. The

IEEE Std. 929-2000, IEEE recommended practice for Utility Interface

of Photovoltaic (PV) Systems, April 2000.
[2] UL 1741, Static inverter and charge controllers for use in photovoltaic
systems, Underwriters Laboratories Inc., Northbrook, IL.
[3] IEEE Std. 1547-2003, IEEE Standard for Interconnecting Distributed
Resources with Electric Power Systems, July 2003.
[4] R. H. Lasseter, and P. Paigi, Microgrid: A conceptual solution, IEEE
35th Power Electronics Specialists Conference, vol. 6, pp. 4285 4290,
June 2004.
[5] J. Liang, T. C. Green, G. Weiss, and Q. Zhong, Hybrid control of
multiple inverters in an island-mode distribution system, 2003 IEEE
34th Annual Power Electronics Specialist Conference, 2003. PESC '03,
Vol. 1, pp. 61 66, June 2003.
[6] J. A. Lopes, C. Moreira, and A. Madureira, Defining Control Strategies
for MicroGrids Islanded Operation, IEEE Transcations on Power
Systems, vol. 21, No.2, pp. 916-924, May 2006.
[7] S. Barsali, M. Ceraolo, P. Pelacchi, and D. Poli, Control techniques of
Dispersed Generators to improve the continuity of electricity supply,
IEEE Power Engineering Society Winter Meeting, Vol. 2, pp. 789 - 794,
2002.
[8] F. Katiraei, and M. R. Iravani, Power Managemnet Strategies for a
Microgrid With Multiple Distributed Generation Units, IEEE
TRanscations on Power Systems, vol. 21, no.4, pp. 1821-1831,
Nov.2008.
[9] C. Moreira, F. Resende, and J. Lopes, Using Low Voltage MicroGrids
for Service Restoration, IEEE Transactions on Power Systems, vol. 22,
no.1, pp. 395-403, Feb. 2007.
[10] IEEE Task Force on Load Representation for Dynamic Performance,
Load representation for dynamic performance analysis. IEEE
Transactions on Power Systems, Vol.8, No.2, pp. 472-482, May 1993.

VIII. BIOGRAPHIES
H. H. Zeineldin received the B.Sc. and M.Sc. degrees in electrical
engineering from Cairo University, Egypt, in 1999 and 2002, respectively. In
2006, He received his Ph.D. in electrical and computer engineering from the
University of Waterloo, Ontario, Canada. Dr. Zeineldin worked for Smith and
Andersen Electrical Engineering Inc. where he was involved with projects
involving distribution system design, protection and distributed generation.

6
Recently, he joined the Masdar Institute of Science and Technology in Abu
Dhabi His current interests include power system protection, distributed
generation, and deregulation.
James L. Kirtley Jr. attended the Massachusetts Institute of Technology,
where he earned the Ph. D. degree in Electrical Engineering in 1971. He has
been a member of the faculty of the Department of Electrical Engineering and
Computer Science since 1971 and currently holds the rank of Professor of
Electrical Engineering. He has also worked for General Electric, Large Steam
Turbine Generator Department, as an Electrical Engineer, for SatCon
Technology Corporation as General Manager and as Chief Scientist, and was
Gastdozent at the Swiss Federal Institute of Technology. Dr. Kirtley is a
specialist in electric machinery and electric power systems. He served as
Editor in Chief of the IEEE Transactions on Energy Conversion from 1998 to
2006 and continues to serve as Editor for that journal and as a member of the
Editorial Board of the journal Electric Power Components and Systems. Dr.
Kirtley is a Fellow of IEEE and a member of the United States National
Academy of Engineering. He was awarded the Nikola Tesla prize in 2002 and
the IEEE Third Millenium medal. Dr. Kirtley is a Registered Professional
Engineer in Massachusetts.