Energy and Power Engineering, 2013, 5, 448-453
doi:10.4236/epe.2013.54B086 Published Online July 2013 (http://www.scirp.org/journal/epe)
Research on the Influence Factors and Coordinated
Control Strategies between Unit and Grid for Isolated
Power System
Ge Jin1, Xiaomei Chen2, Yongxin Feng1, Shaoxiang Deng1, Hanting Yan2, Ze x iang Cai2
1Electric Power Research Institute, Guangdong Power Grid, Guangzhou, China
2School of Electric Power, South China University of Technology, Guangzhou, China
Received March, 2013
ABSTRACT
As the existing coordinated control strategies between grid and unit have limitations in isolated power system, this pa-
per introduces new coordinated control strategies which can improve the stability of isolated system operation. This
paper analyzes the power grid side and unit side influence factors on the isolated power system. The dynamic models
which are suitable for islanding operation are applied to simulate and analyze the stability and dynamic characteristics
of the isolated power system under the conditions of different load disturbances and governor parameters. With consid-
ering the differences of frequency characteristics between the interconnected and isolated power system, the adjusting
and optimization methods of under frequency load shedding are proposed to meet the frequency stability requirements
simultaneously in the two cases. Not only proper control strategies of the power plant but the settings of their parame-
ters are suggested to improve the operation stability of the isolated power system. To confirm the correctness and effec-
tiveness of the method mentioned above, the isolated system operation test was conducted under the real power system
condition, and the results show that the proposed coordinated control strategies can greatly improve stability of the iso-
lated power system.
Keywords: Isolated Power System; Coordinated Control Strategies; Under Frequency Load Shedding; Dynamic
Frequency Characteristics; Speed Governor Parameter Settings
1. Introduction
With the cross-regional interconnection of power system
and the continuous expansion of power grid, the connec-
tions between systems at all levels are becoming closer
and closer. And the single fault of power system, extreme
weather or manual operation errors may possibly cause
large area blackout accidents, and even cause the crash of
power system. After that blackout, the power system is
forced to take black start which is a self-help means after
the power system accident and produces huge economic
loss. However, when the large-scale power system fault
occurs, we can take reservations and reasonable meas-
ures for the local power system, and keep isolated system
operation stable. This can not only ensure the power sup-
ply for important load of the city, but also be helpful to
restoration of the whole system which can reduce the
economic loss.
In recent years, many countries around the world have
occurred many largely influenced blackout accidents,
such as the America and Canada blackout accident in
2003. In these accidents, a large number of regional grids
which have the isolated system operation condition fail
to make a full use of the coordinated control ability due
to lack of effective coordinated control measures for the
isolated system operation. And this leads to the unit trip-
ping and the power lose of those areas [1-3].
There are some problems in research of possible risks
grasp and control strategies for practical isolated power
system operation. Firstly the characteristics and relevant
coordinated control strategies of most power plants have
not been considered for isolated system operation [4-7].
Then the stable operation of the power plant depends
more on the major power system than the isolated power
system as the effective power support.
With these problems, this paper analyzes the influence
factors on the isolated system. Dynamic models which
are suitable for the isolated power system are used to
analyze the stability of frequency response under the
different conditions such as load disturbances, speed
governor parameters and logical control conditions. Op-
timization method of the under frequency load shedding
is proposed to meet the frequency stability requirements

G. JIN ET AL. 449
in interconnected and isolated system operation with
considering the differences of frequency characteristics
in the two cases[8-10]. The parameter settings and con-
trol strategies are proposed to improve the isolated op-
erational ability[11-16].
2. Frequency Characteristics of Isolated
Power System and Its Influence Factors
2.1. Frequency Characteristics
For most heavy load in regional power system, the sys-
tem response to the isolated network operation is basi-
cally a consistent frequency transient process which is
often accompanied with voltage changing highly or low-
ly.
When the generated power is shortage in isolated net-
work, the initial transient process is determined by the
response of power plants spinning reserves and low-fre-
quency low-voltage load shedding, often within seconds
reaching the lowest frequency. After that point the sys-
tem frequency response depends on the characteristics of
prime mover and speed governor. With the surplus power
in isolated system, the frequency will rise. Then the
speed governor will work and give a response to reduce
the mechanical power produced by the unit. The ability
to maintain stability instead of losing load of the isolated
network actually depends on the partly load shedding
ability of the unit.
2.2. Influences of the Power Grid
The influence of power system primarily performances
on the operation mode that is set. It is necessary for iso-
lated power grid. So when the operation mode of the
power system is made, all of the possible running situa-
tion that may occur in the isolated power system should
be considered. The automation equipments of power
system including the low-frequency low-voltage load
shedding, and low frequency disconnection device as the
last defense line of the power system are important in
affecting the operation mode of the power system. With
them, the collapse of frequency and voltage can be pre-
vented in the accident and the partly power supply can be
guaranteed. But if automation equipment such as the un-
der frequency load shedding fails to take consideration
on the frequency characteristics both of interconnected
system and isolated system, it will cause frequency drop-
ping dramatically during isolated network runtime, and
the power system will collapse in the end.
2.3. Influences of the Unit
The steam turbine of each unit under different speed de-
viation has different cultivate times in total. It shows the
lifetime of the unit. The low frequency operation of the
steam turbine is more dangerous than the high one. Be-
cause after the frequency rises due to load shedding, it
can be reduced by the governor or OPC, the overspeed
protection control system, to regulate output power.
Figure 1 shows the lifetime of steam turbine under
different frequency deviation. For power plant A, the two
horizontal lines indicate that there is no loss and the tur-
bine can run continuously between the frequency of 49.4
Hz and 50.6 Hz. When the frequency drops to 46.5 Hz,
the turbine can only run for one second, as the lifetime
curve shows at the time of frequency offset. The power
plant B has a lower frequency than 50Hz, because of
quickly action of its frequency governor. When the fre-
quency is lower than 50Hz, the life time of the turbine
will stepped reduce.
When isolated network is forming, some fixed values
of the protection devices may affect the stability of iso-
lated system operation, while current protection of the
unit and its fixed values are setting mainly based on in-
terconnected network operation. So it is necessary to set
reasonable values for the generator protection, which will
not only ensure safe operation of the unit, but also pre-
vent the unstable operation or trip of the units. The fol-
lowing generator protection equipments should be con-
sidered during the isolated system operation: electrical
overspeed trip, mechanical overspeed trip, turbine over-
speed protection control (OPC), low and over voltage
protection of generator, and auxiliary protection settings.
3. Dynamic Models of Isolated Power
System
The influence of the unbalanced power on dynamic fre-
quency characteristic of the power system at the moment
of isolated system formation can be analyzed by the
models in isolated network. Considering the unit in city
grid mainly uses the turbine generator, especially the
reheat type as shown in Figure 2. The domain time con-
stant of this turbine model is the thermal time constant
TR. The constant Km is the total gain to realize the ad-
justment of output power [8].
Figure 1. Turbine lifetime with different frequency offsets

G. JIN ET AL.
450
Typical speed governor model is shown in Figur e 3.
The most important parameter of speed governor is the
differential coefficient. It is marked with R in the feed-
back element of the above figure. And a simplified mod-
el of power system is established based on including its
important dynamic characteristics. By integrating reheat
steam turbine model with speed governor model we get
the reduced order system frequency model, as shown in
the Figure 4.
There are six constants to describe the characteristics
of this model: gain coefficient Km, damping coefficient D,
inertia constant H, thermal time constant TR, power pro-
portion of high pressure cylinder FH, and differential
coefficient R.
The incremental power setting of governor and output
electromagnetic power of generator are the two input
variables of the model. Governor incremental power
value is adjusted according to the unit instruction. The
output setting value needs to meet requirement of the
load. As at the moment of isolated network forming what
is corresponding to the disturbance are system character-
istics of just few seconds, so the governor incremental
power value can be considered to be constant. So elec-
tromagnetic power variation can be seen as a disturbance
power .
d
P
1
1
CH
ST
V
1
1
RH
ST
1
1
CO
ST
I
P
F
L
P
F
H
P
F
Figure 2. Block Diagram of a single reh eat s team turbine.
1
2
1Ts
1Ts

 3
1
T
1
R
 
1
s


Figure 3. Block diagram of a typical speed governor.
H
R
1-F
1+T s
H
F
m
K
D
2
1
Hs

1
R




Figure 4. The reduced order SFR (system frequency re-
sponse) model.
For the situation that load is higher than the active
output power in isolated network, we can make d P

e
P

corresponding to the stepped load change. It is op-
posite when load is less than the active output power.
We take the typical parameters of power system as:
0.95
m
K

, ,, 1.0D4.0H
8.0
R
T

,0.3
H
F

,. 0.05R
We only consider the situation when load is higher
than the active output power, and the load unbalances are
10%, 20%, 30%, 40%, 50% power deficiency. The fre-
quency response curve is shown in the Figure 5.
Take the typical parameters of power system that the
unbalanced load is given 20%, and the differential coef-
ficients are 0.05, 0.06, 0.07, 0.08, 0.09, 0.1. The fre-
quency response curve is shown in the Figure 6. From
the figure, we can see that differential coefficient has no
influence on initial slope of the frequency curve. The
main effects are shown on the recovery time, maximum
frequency deviation and the steady state frequency de-
viation.
4. Under Frequency Load Shedding Scheme
for Isolated Power System
During the isolated network operation time, the adjust
amplitude of generator unit is limited because of the lack
of units. So the influence of load disturbance on the iso-
lated network is very obvious [9]. The frequency charac-
teristics of isolated network are very dependent on the
Figure 5. The Frequency Response of Isolated Power Sys-
tem with Different Power Unbalances.
Main : Gra
p
hs
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
-1.80
-1.60
-1.40
-1.20
-1.00
-0.80
-0.60
-0.40
-0.20
0.00
频率偏差
(Hz)
ω1(R=0.05) ω2(R=0.06) ω3(R=0.07) ω4(R=0.08) ω5(R=0.09) ω5(R=0.1)
Figure 6. The Frequency Response with Different Governor
Differential Coefficient R.

G. JIN ET AL. 451
size of the spinning reserve and frequency regulating
characteristics of the generator.
The practical parameters of isolated system operation
in a certain power system are taken as an example. The
Figure 7 has obviously shown that the generator fre-
quency regulation coefficient reached the maximum val-
ue when the frequency is about 49Hz. Then it decreased
with the gradual depletion of spinning reserve. Therefore,
when under frequency load shedding scheme is adjusted,
it is very necessary to consider the effect of spinning
reserve.
The frequency regulation effect coefficient KS of
power system is determined by both of the load and gen-
erator frequency regulation characteristics together.
Taking the practical power system as an example, we
make assessment of possible isolated system operation
on each regional network. Here we use the regional
power grid A and B to simulate. We assume that it is the
loss of all the 220 kV power supply that causes the iso-
lated system operation. The powers of network A and B
are of 50% and 80% more than power output. Table 1 is
the comparison of the under frequency load shedding
scheme between the existing and the optimizing one in
the two pieces of power system.
Figure 8(a) shows the existing under frequency load
shedding configuration scheme can not effectively re-
strain frequency from sharply falling. Figure 8( b) shows
after adopting optimization load shedding scheme in grid
A the lowest value of frequency is 48Hz after 5 seconds,
and then it returns to 49.6 Hz. Figure 9 shows the fre-
quency curves comparison between the existing and op-
timizing under frequency load shedding in grid B. The
above method is helpful to solve the frequency sharply
dropped problem caused by the possible large distur-
bance occurred when isolated power formed.
The under frequency load shedding is a provincial plan
as a whole and is distributed according to cities. For the
city field of network in which isolated network operation
may appear, the optimization under frequency load shed-
ding scheme is suitable. The deficiency power of the load
can be supplied by other pieces of network.
Figure 7. Generator frequency regulation coefficient varia-
tion curve with the system frequency variation.
Table 1. Comparison between the existing and optimizing
under frequency load shedding scheme in certain pieces of
power system.
Sheaves 1 2 3 4 5
Action Frequency(Hz) 49.048.8 48.6 48.4 48.2
Delay(s) 0.20.2 0.2 0.2 0.2
Existing load shedding
scheme of power grid A 0% 1.5% 0% 0% 10%
Optimization scheme of A4.5% 5.3% 6.1% 6.6%7%
Existing scheme of B 6.8% 0% 23.6% 9.1% 0%
Optimization scheme of B3.8% 4.4% 4.9% 5.4%5.7%
Sheaves 6 7
First
special
round
Second
special
round
Action Frequency(Hz) 48.047.8 49.0 49.0
Delay(s) 0.20.2 15 20
Existing load shedding
scheme of power grid A 0% 1.5% 3% 10%
Optimization scheme of A7.2% 7.3% 3% 3%
Existing scheme of B 12.8% 0% 12.1% 0%
Optimization scheme of B5.9% 6.1% 3% 3%
(a) the present scheme
(b) the optimized scheme
Figure 8. Frequency curve comparison of the present and
optimized under frequency load shedding schemes in Iso-
lated power system A.

G. JIN ET AL.
452
the worst conditions. So this paper mainly does the re-
search of the frequency and power response in the case
of the relatively large imbalance power disturbance when
isolated network forms, and analyzes the corresponding
control strategy.
During isolated system operation, primary frequency
regulation needs to exit from power control in time, and
turns to the valve position control. In order to ensure the
safety of the unit, and the stability of isolated system
operation, OPC can not quit. For the primary frequency
regulation of the unit, the differential coefficient is ex-
pected to set lowly. This can make the power response in
time after the frequency changes, and prevent the sig-
nificant change of frequency. Setting the dead band in a
low level can optimize the dynamic regulating process of
frequency. It is helpful for the stability of frequency to
enlarge the top and bottom frequency regulation limita-
tions, and also for reducing the oscillation. Relatively
fast response of the servo system can increase quickness
of primary frequency regulation of the unit.
(a) present scheme
There is a frequency deviation after primary frequency
regulation. But turbine is not allowed to run with this
deviation of frequency for a long time. We can put in the
frequency regulation control loop for the speed control
during isolated system operation. This is the secondary
frequency regulation. Increasing dead band and limita-
tion of primary frequency control appropriately can de-
crease the number of movements for OPC.
The OPC setting value of the 100 MW unit in one
power plant of the power system is respectively taken for
103% and 106%.The Figure 10 shows the simulation.
Increasing the OPC value appropriately by 2% - 3% can
reduce frequent opening and closing valve caused by
frequency fluctuation.
(b) optimized scheme
Figure 9. Frequency curve comparison of the present and
optimized schemes load shedding in Isolated Network B.
5. Control Strategies of Isolated System
Operation To sum up, we can make suitable isolated system op-
eration strategy by setting the parameters and control
measures of power plant units. The Table 2 takes two
certain local power plants in power system as an example,
and gives advice about parameter settings of the speed
governor.
The unbalanced power becomes the main risk for iso-
lated system operation. Generally the unbalanced power
of the load is most serious at the early period of isolated
power system. If we want the unit to be able to regulate
any kind of isolated power system, it needs to adapt to
(a) Frequency Curve of 103% OPC Setting Value (b) Frequency Curve of 106% OPC Setting Value
Figure 10. The Frequency Curve of 103% and 106% OPC Setting.

G. JIN ET AL. 453
Table 2. The recommendations of parameter value settings of two power plants in power system.
Power Plant Types Parameter Setting Adjustable Differential Coefficient Frequency Modulation Limiter
scope 4%-5% Upper limiter:35%，no lower limiter
A Plant Mechanic hydraulicProposed value 4% 35%
Scope 4%-5% Upper limiter:35%，no lower limiter
B Plant DEH Proposed value 4% ±35%
Power Plant Types Parameter Setting Frequency Modulation Dead Band OPC Value
scope / /
A Plant Mechanic hydraulicProposed value / /
Scope ≤ ±0.034Hz 103%-108%
B Plant DEH Proposed value ±0.017Hz 106%
6. Conclusions
This paper analyses the power grid and unit influence
factors on the isolated system operation. And the system
frequency response model is applied to analyze the spe-
cific influences. By simulating and analyzing the stabil-
ity of regional power system under different load distur-
bances, regulating system parameters and control logics,
parameter settings and control strategies are proposed to
improve the ability of isolated system operation. In the
view of the frequency characteristic difference between
the isolated system operation and networking operation,
we have proposed the optimizing scheme of under fre-
quency load shedding which has considered the fre-
quency characteristic differences. This scheme can meet
the frequency stability requirement both of networking
operation and isolated system operation. Taking the
practical power system as an example, we have con-
ducted the real isolated system operation test. This test
shows the correctness and effectiveness of the proposed
method. The coordinated control strategies improve the
ability of power grid and unit to deal with the accident.
REFERENCES
[1] J. S. Wang, T. T. Deng, J. C. Ran, “Start-up and Opera-
tion Mode of Fossil Plant Units in an Isolated (Little)
Power Grid,” Automation of Electric Power Systems, Vol.
32, No. 21, 2008, pp. 102-106.
[2] H. H. Lin, W. Wang, Y. Y. Wei, “Stable Operation Con-
trol Technologies for Islanded Networks in 220 Kv
Guilin Power Grid,” Southern Power System Technology,
Vol. 2, No. 3, 2008, pp. 49-53.
[3] “Causes of the August 14th Blackout in the United States
and Canada,” US/Canada Power Outage Task Force
2003.
[4] J. Li Jun and P. Luo, “Stability Control Technology for
Isolated Network with 300 MW Thermal Power Unit”
Guangxi Electric Power, No. 1, 2009, pp. 9-13.
[5] H. Huang, G. H. Xu Guanghu and C. Yu, “Operation
Experience of Isolated Network in CSG during Ice Dis-
aster (2008),” Southern Power System Technology, Vol. 2,
No. 5, 2008, pp. 6-9.
[6] F. Pilo, G. Celli, S. Mocci, “Improvement to Reliability
in Active Networks with Intentional Islanding,” IEEE In-
ternational Conference on Electric Utility Deregulation,
Restructuring and Power Technologies, 2004.
doi:10.1109/DRPT.2004.1338023
[7] P. M. Anderson. Power System Protection [M]. 2009.
[8] Z. P. Zhou, X. L. Sun, C. P. Fu, K. Liu, “Insular Power
System Dynamic Characteristic Examination Research,”
Power System Technology, Vol. 32, No. 2, 2008, pp.
78-81.
[9] A. M. Li and Z. X. Cai, “Frequency Dynamics Analysis
and Load Shedding Assessment Based on the Dynamic
Simulation Trajectory of Interconnected Power System,”
Transactions of China Electrotechnical Society, Vol. 24,
No. 9, 2009, pp. 171-177.
[10] J. Wang, X. R. Wang, D. P. Xie, etc, “Investigation of
under Frequency Load Shedding on an Isolated System,”
Power System Protection and Control, Vol. 38, No. 3,
2010, pp. 29-52.
[11] J. Li, “Control Strategy of Single Power Unit with Iso-
lated Operation in Thermal Power Plant,” Electric Power,
Vol. 41, No. 11, 2008, pp. 51-54.
[12] H. W. Lian, Y. Zhao, “Parameter Optimization for Fre-
quency Modulation of a Unit in Isolated Power Grid
Based on Orthogonal Test,” Turbine Technology, Vol. 47,
No. 2, 2005, pp.108-111.
[13] Z. J. Huang, J. Chao, X. Y. Li and P. Kang, “High Fre-
quency Phenomenon Caused by Fault Occurred in South-
ern Part of Guiyang Power Grid and Simulation of Over-
speed Protection Controller,” Power System Technology,
Vol. 31, No. 15, 2007, pp. 26-32.
[14] Z. P. Zhou, “First Frequency Modulation Delimitation for
Electric System Stabilization Analyse,” Henan Electric
Power, No. 03, 2009, pp. 40-45.
[15] “IEEE Working Group on Prime Mover and Energy Sup-
ply Models for System Dynamic Performance Studies.
Dynamic Models for Fossil Fueled Steam Units in Power
System Studies,” IEEE Transactions on Power System,
Vol. 6, No. 2, 1991, pp. 753-761.doi:10.1109/59.76722
[16] F. P. de Mello, Fellow, “Boiler Models for System Dy-
namic Performance Studies,” IEEE Transaction on
Power Systems, Vol. 6, No. 1, 1991, pp. 66-74.
doi:10.1109/59.131048