Energy and Power Engineering, 2013, 5, 869-872
doi:10.4236/epe.2013.54B166 Published Online July 2013 (http://www.scirp.org/journal/epe)
A Research on Emergency DC Power Support
Based on DCOI
Fan Fan, Yingmin Zhang, Xingyuan Li, Wei Ma
School of Electrical Engineering and Information, Key laboratory of Smart Grid Sichuan University, Chengdu, China
Received October, 2012
ABSTRACT
An integrated scheme of emergency DC power support which is based on dominant center of Inertia DCOI is proposed
to improve power system transient stability. The inter-area speed differences equivalent by dominant center of inertia
alleviates the main weakness o f the traditional signal which is made of the freq uencies at both sides and is represen t of
abnormal information of the nonlinear dynamic behavior of the power system based on a model in which two AC sys-
tems are connected by a HVDC Link. Sichuan Power Grid connected with Southwest Power Grid by Deyang-Baoji
HVDC project is used to testify the method. The simulation results show that the control strategy can significantly ad-
vance the transient stab ility of AC/DC system through extended equal area criterion (EEAC).
Keywords: DCOI; Speed Differences; EEAC
1. Introduction
The main appeal of HVDC (High Voltage Direct Current)
with long-distance and large capacity competence for
power transmission is the prominent ability to regulate
reactive and active power. HVDC improves an efficient
and robust approach to transient stability by modulating
DC power into AC grid rapidly, which will remedy power
imbalance between sending-grid and receiving- grid in
order to advance transient angle stability and improve
last-low-voltage level or voltage vibrating conditions [1].
Emergency control is a most significant measure
against large disturbances which are the gravest threat to
the stability to transient security of power systems. Thus,
it becomes necessary to choose the most useful DC
modulation signal from the huge amount of in- formation
for the system stability control services [2]. This paper
identifies and selects the “dangerous generators” to make
the dominant center of inertia while discarding all “harm-
less” ones by appraising the kinetic energy as the index
in the during-fault period respectively. The inter-area
speed differences equivalent by dominant center of iner-
tia can effectively reflect the abnormal information of the
nonlinear dynamic behavior of the power system based
on a model in which two AC systems are connected by a
HVDC link. Accordingly it is as the controller input sig-
nal [3-6].
DC power lift / drop-back can provide certain transfers
of power for the AC system to ensure the stability of the
system by modifying the instructions of the DC system
power to increase or reduce the DC transmission power.
For instance, when there happens a generator loss in the
sending-grid system, DC power drop-b ack can reduce th e
DC power to balance the system's active power shortfall
[7].
In this paper, an emergency DC power support com-
prehensive strategy is devised by combining DC power
lift / drop-back with the dominant center of inertia speed
deviation signal to modify the power set value jointly.
Taking Sichuan power grid for background, it indicates
that the signal deriving from the dominant center of iner-
tia is more efficient than the traditional signal made of
the frequencies at both sides after comparative analysis
of the system characteristics of various operating condi-
tions under typical faults.
2. Mechanism of Edcps Comprhensive
Strategy
In the transient process, the separation of the system is
not decided by the energy of the whole system. The un-
stable situation of the system is caused by a small num-
ber of out-of-step generators seriously deviating from the
system and most of the generators with each other can
still maintain synchronous operation. If all seriously dis-
turbed crew in the system is stabilized, the system is sta-
ble; Otherwise, the system is unstable. Thus, th e stability
of the system is converted to determine the stability of
the severely disturbed generators. During a failure, tran-
sient energy injected into the system is converted into the

F. FAN ET AL.
870
absolute kinetic energy increment of each generator set.
Kinetic and potential energ y convert into each other dur-
ing the process of power angle swing [8,9]. Under the
center of inertia (COI) coordinates each generator rotor
equations of motion are shown as follow:


me COICOI
COI
1
d
dd
d
i
i
ii
iii ii
T
n
i
ii
i
T
t
M
MPPPD
tM
MD
M






 






(1)
where
MT : inertia time constant of system
Mi : inertia time constant of generator

I : power angel of generator
ωi : angular velocity deviation
Pmi : input mechanical power
Pei : output electromagnetic power
PCOI : accelerating power of COI
Transient energy of the system is expressed as follow:
COI
1
2COI
11
d
1
d
2
i
si
i
si
n
i
iimi eii
iT
nn
i
iimi eii
ii T
kp
M
VMPPP
M
M
MPPP
M
VV












 




















(2)
Take the derivative of V as the change rate of the
transient energy:
 
me COI
1
COI COI
11
d d
dd
n
ii
iiii i
iT
nn
i
iiiii
ii
T
M VMPPP
ttM
M
DD
M


 





 




 






 



 





(3)
V Consists of two parts which are the rate of relative
kinetic energy into potential energy and the rate of
potential energ y into relative kinetic energy. If change the
rate of the transient energy to be less than zero or
fluctuate around zero and gradually decreases over time,
the system has a larger stability margin. The kinetic
energy of the relative motion between the center of
inertia of the critical group and the remaining group can
truly reflects out of sync information under dual-machine
equivalents mode:
2
1
2
k
VM

 (4)
Kinetic ene rgy is expres sed by th e speed and according ly
the equivalent of two machine speed deviation can be
used as a feedback signal to realize emergency control.
And the DC power modulation amount generated by the
feedback signal is represented as follow:


d AB
PK K  0

 
(5)
Early in the large disturbance, the generators absorb-
ing kinetic energy deviate from the synchronous speed
and gradually evolve into the relative movement of the
two generator group. The greater the relative kinetic en-
ergy of the two generator groups, the greater the likely-
hood of system instability [10]. After large disturbance,
all generators can be divided into two groups, namely
severely disturbed groups (S groups) and remnant groups
(R groups), and then the whole system can be equaled
into dual-machine-unstable model. For the time-varying
two-machine system, the NRP (not return point) which
decides the stability is the DSP (dynamic saddle point)
where the value of image acceleration power is zero.
During the process of the first swing, first determine the
moment when acceleration kinetic energy of the single
machine for critical group reaches a maximum [11,12].
And then selecting the generator maximum acceleration
kinetic of this moment as the benchmark, participation
factor is defined as the ratio of each generator accelera-
tion kinetic energy with the reference value；For remnant
groups, first calculate each generator deceleration kinetic
energy the moment disturbed trajectory get through the
DSP and take the same critical group generator accelera-
tion kinetic energy maximum value as a reference,
therefore the participation factor is defined as the ratio of
each generator deceleration kinetic energy with the ref-
erence value. To make the formation of the dominant
center of inertia (the Dominant Center of Inertia, DCOI),
selected generators of which participation factors are
greater than 0. 5. According to formula (1), formula (2),
formula (12), the control input signal of the emergency
power support is integrated by speed deviation informa-
tion extracted from the two DCOIs.
ii
iA
A
A
ii
iA
A
A
Ai
iA
M
M
M
M
MM













(6)
j
j
jB
B
B
j
j
jB
B
B
Bj
jB
M
M
M
M
MM













(7)
where:

F. FAN ET AL. 871
Mi : inertia time constant of generator in A group
Mj : inertia time constant of generator in B group
MA : inertia time constant of the equivalent genera-
tor A
MB : inertia time constant of the equivalent genera-
tor B

i : power angel of generator in A group

j : power angel of generator in B group

A : power angel of the equivalent generator A

B : power angel of the equivalent generator B
ωi : angular velocity deviation in A group
ωj : angular velocity deviation in B group
ωA : angular velocity deviation of the equivalent
generator A
ωB : angular velocity deviation of the equivalent
generator B
3. Control Effect
Taking Sichuan power grid in the summer peak load pe-
riod for background, while Debao HVDC system operat-
ing in monopolar mode, this pap er focuses on the impact
of EDCPS strategies on transient stability in Sichuan
power grid. EDCPS strategies are as follows: 1) DC
modulation based on the traditional inertia center (TIC);
2) the speed deviations modulation based on the dominant
center of inertia (DCOI); 3) the comprehensive EDCPS
strategy based on the dominant center of inertia the speed
deviations modulation cooperating with DC power
upgrade / drop back(CEDCPS).
Consider the following fault: the fault-free outage of
generating unit in Ertan Hydropower Station. Investigate
the kinetic energy of the generator of the critical group
and the remaining gr oups respectively. Kinetic energy of
Ertan unit increases the maximum, which is followed by
the order of the Jialing, Jintang factory, Jiangyou,
Pubugou. The detailed information is shown in Table 1.
It shows that Ertan, Jialing, Panzhihua and Jintang plant
units satisfy the conditional and take them to form the
dominant center of inertia of group A. Compare the
effect of the three emergency power support strategies
for improving the system transient stability through the
following simulation [13].
Table 1. The largest accelerating power of generator ruing
the fault period.
Unit name Acceleration power Unit name Acceleration power
Ertan
Jialing 12.282 2 10.782 0 Tianwan Ziyili 1.666 5 1.057 7
Panzhihua 6.647 2 Shiziping 0.981 1
Jintang 6.267 5 Xiaotianhu 0.825 2
Pubugou 4.198 5 Tianlonghu 0.618 8
Fuxi 4.089 1 Weizhou 0.518 8
Baozhou 3.799 0 Xuecheng 0.313 8
Baozhu 2.599 8 C hib usu 0.265 7
Figure 1 shows the degree of oscillation of curve 2 is
stronger than curve1, which is also consistent with the
fact that the kinetic energy of generator units disturbed
seriously is lager than the other s .
The fault results in not on ly the serious system oscilla-
tion but also the power shortage which is up to 320 MW
in Sichuan Power Grid. The power shortage is made up
by droping the DC power. The three DC power control
strategies all can effectively drop the DC power to en-
hance the transient stability of the system. Analyzing
Table 1, Figure 2 and Figure 3 it can be obviously seen
that the improving effect of policy 3 is the best and
Strategy 2 is better than Strategy 1.
Figure 2 tells the trend of the Ertan unit power angle
and Figure 3 tells the trend of the Huangyan - Wanxian
active power by different control strategies in the post-
fault.
The oscillation in system was caused by the serious
fault during the angle-swing-up course. Strategy 2 taken
the DOCI speed deviation signal which can effectively
20151050
0
-0.01
-0.02
-0.03
-0.04
-0.05
-0.06
-0.07
-0.08
t
/
s
2 DCOI
1 TIC
Figure 1. Moudulation SIGNAL Curve s(HZ).
20151050
44
42
40
38
36
34
32
Figure 2. Power angle curves of ertan under different mod-
ulation signal for EDCPS (Degree).

F. FAN ET AL.
872
20151050
2,000
1,500
1,000
500
0
Figure 3. Active power of Huangyan-Wanxian lines (MW).
reflect the inter -area oscillatio n information provides more
damping compared to strategy 1 which is limited by lo cal
information. Strategy 3 with the same ability of providing
damping as strategy 2 makes up the power shrotage to
restore the system to a stable operating point close to the
pre-failure stable equilibrium point by droping DC
power.
4. Conclusions
This paper proposes a wide area information filtering and
integration method ba sed on participation factor which is
defined by relative kinetic energy increase during the
transient pro cess. It provides an effective wa y for screening
and selection of the feedback control signal from the
massive information. Compared with traditionas signal
modulation, the selected signal with higher accuracy can
not only reflect the global features of the system, but also
greatly reduce the amount of computation. After a com-
parative study of the system characteristics of various
EDCPS under typical faults, it indicates that the DOCI
speed deviation as a control signal can significantly en-
hance the inter-area damping. The emergency power
support comprehensive strategy based on the dominate
center of inertia speed deviation modulation cooperating
with DC power increase / back drop further develop and
utilize the potential of DC modulation to maintain system
transient stability. It provides a new way for detection
and effective correction of instability threat caused by
severe disturbance.
5. Acknowledgements
This work is supported by National Natural Science
Foundation of China (No. 51037003) and National High
Technology Research and Development Program of
China (863 Program). At the same time, the authors
ould like to appreciate all research members of the re-
search team for valuable discussion and investigation.
REFERENCES
[1] X.-Y. Li, “High Voltage Direct Current Transmission
System Operation and Control” Beijing: Science Press，
1998.
[2] E. Bristol, “On a New Measure of Interaction for Multi-
variable Process Control,” IEEE Transaction on Automa-
tion Control,Vol. 11, No. 1, 1966, pp. 133-134.
doi:10.1109/TAC.1966.1098266
[3] P. X. Zhang, Y. J. Cao, H. F. Wang, et al., “Application
of Relative Gain Array Method to Analyze Interaction of
Multi-Functional FACTS Controllers,” Proceedings of
the CSEE, 2004, Vol. 24, No. 7, pp. 13-17．
[4] J. Y. Zhang and Y. Z. Sun, “A Vulnerable Cut Sets Based
Sensitivity Analysis Method for Locating Emergency
Control,” Power System Technology, Vol. 31, No. 11,
2007, pp. 1-7.
[5] X. M. Mao, Y. Zhang and L. Guan, “A Novel WAMS
Based HVDC Modulation Method,” Automation of Elec-
tric Power Systems, Vol. 31, No. 7, 2007, pp. 45-49.
[6] J. B. He, L. C. Li, H. X. Chen, et al., “Selection of Feed-
back Signal for Power System Damping Controller Based
on Wide Area Measurements,” Automation of Electric
Systems, Vol. 31, No. 9, 2007, pp. 6-10.
[7] W. D. Yang, Y. S. Xue, Y. Jing, et al., “Emergency DC
Power Support to AC Power System in the South China
Power Grid,” Automation of Electric Power Systems，
Vol. 27, No. 7, 2003, pp. 68-72.
[8] Y. Zhang, L. Wehenkel, P. Rousseaux, et al., “SIME:
Ahybrid Approach to Fast Transient Stability Assessme nt
and Contingency Selection,” Electrical Power Energy
Syst, 1997, Vol. 19, No. 3, pp. 195-208.
doi:10.1016/S0142-0615(96)00047-6
[9] Y. S. Xue, “A Critical Comparison of Various Methods
for Transient Stability Assessment-Part Two Barrier Point
and the Observation Point,” Automation of Electric
Power Systems， Vol. 25, No. 11, 2001, pp. 1-7.
[10] M. C. Liao, G. L. Cai and Y. J. Zhang, “Transient Volt-
age Stability of Received Power Grid in AC/DC Hybrid
Power Systems,” Power System Protection and Control,
Vol. 37, No. 10, 2009, pp. 1-4.
[11] H. J. Wu, Y. H. Wang, X. Y. Li, et al., “Study on DC
modulation for AC/DC hybrid transmission system based
on comprehensive performance index,” Power System
Protection and Control, Vol. 38, No. 22, 2010, pp. 68-73.
[12] M. M. Xu, X. Y. Li, Y. H. Wang, et al., “Impact on the
Damping Characteristics of Sichuan Power Grid by De-
bao DC Modulation,” Power System Protection and Con-
trol, Vol. 38, No. 23, 2010, pp. 141-146.
[13] T. S. Xu, B. J. Li, Y. H. Bao, et al., “Optimal Parame-
ter-Setting of Under Frequency and Under Voltage Load
Shedding for Transient Security,” Automation of Electric
Power System, Vol. 27, No. 22, 2003, pp.12-15.
w