Sie sind auf Seite 1von 5

1

Monitoring of Interconnected Power System


Parameters using PMU based WAMS
Yutaka Ota, Takuhei Hashiguchi, and Hiroyuki Ukai,
Mitsuhiro Sonoda, Yasushi Miwa, and Akira Takeuchi

power outlet properly reflects the power oscillations of upper


Abstract— In order to easily monitor the wide-area power power stations with higher voltage levels. An availability of the
system information covering many power producers, suppliers, power outlet phasor is validated by comparing with the
and power companies, authors have been developed a PMU measurements in the actual power station.
(Phasor Measurement Unit) based WAMS (Wide Area
Measurement System). This WAMS momentarily gather
Next, an estimation method of interconnected power system
information about the power oscillation affecting the stability of parameters such as inertia constants, damping coefficients, and
Japanese longitudinal power systems by signal processing of the synchronizing power coefficients from the monitored power
power outlet phasor. This paper introduces the features of the oscillation is proposed in chapter 3. When obviously certain
PMU based WAMS, and an availability of the power outlet supply and demand imbalance such as a generator trip, a load
phasor is validated by comparing the measurements with actual shedding, or a governor test is occurred, a just-in-time
power station. And as one of the applications, an estimation
method of interconnected power system parameters from the
equivalent swing equation model can be obtained by the
monitored power oscillation information is proposed. proposed method. In this paper, the proposed method is applied
to Japanese western 60Hz systems.
Index Terms—Phasor Measurement Unit, Wide Area
Measurement System, Power Oscillation, Interconnected Power II. PMU BASED WAMS
System Parameters
A. Hardware and Software Features
Fig.1 shows the location of PMU and PDC on Japanese
I. INTRODUCTION wide-area power system. Total eleven PMUs have been already

P MU (Phasor Measurement Unit) is in the spotlight as a key


technology of WAMS (Wide Area Measurement System),
and some applications making use of a merit that
installed at power outlets in universities. The employed PMU in
this research is NCT2000 made by Toshiba Corporation [14],

synchrophasor is the snapshot of power system state have been


developed for American, European, Chinese, and many power
grids [1]-[12].
Authors have already installed PMUs at domestic 100V
power outlets in some universities located in every power
companies of Japanese wide-area power systems. The
synchrophasors are momentarily collected to the PDC (Phasor
Data Concentrator) through the internet. Furthermore, the
power oscillation components are extracted from the power
outlet phasors by using the digital filters [13].
In this paper, the outline of the PMU based WAMS is
explained in chapter 2. It is concerned the phasor at 100V

This work was supported in part by Research Foundation for the


Electrotechnology of Chubu (REFEC), and the Ministry of Education, Science,
Sports and Culture, Grant-in-Aid for Scientific Research (A), 18206028,
2006-2009.
Yutaka Ota and Hiroyuki Ukai are with Nagoya Institute of Technology,
Gokiso-cho Showa-ku Nagoya Aichi 466-8555 JAPAN (phone & fax:
+81-52-735-5389; e-mail: yota@nitech.ac.jp).
Takuhei Hashiguchi is with Kyushu University, 744 Motooka Nishi-ku
Fukuoka 819-0395 JAPAN
Mitsuhiro Sonoda, Yasushi Miwa, and Akira Takeuchi are with Chubu
Electric Power Co., Inc. 1 Higashishin-cho Higashi-ku Nagoya Aichi 461-8680
JAPAN. Fig. 1. Location of PMU and PDC on Japanese power system.
2

TABLE I
SPECIFICATIONS OF DIGITAL FILTERS
Filter Type FIR
Window function Hamming Widow
Width of filter 20[s] (600 samplings)
Cutoff frequency of LPF 0.15[Hz]
Width of pass-band of BPF 0.20[Hz]
Center frequency of BPF Adjustable

1.2
1
0.8

Gain
0.6
0.4
0.2
0
Fig. 2. Task schedule for data collection and data processing. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Frequency [Hz]
(b) Low-pass filter
which is oriented to an internet based multifunctional
1.2
measurement terminal. The instantaneous values of power
outlet voltage are measured with precise time-synchronization. 1 Adjustable
The phasor angle is calculated from the instantaneous values by 0.8
Gain

the phasor computation based on DFT (Discrete Fourier 0.6


Transform) in every period of 1/30[s] for 60Hz system or 0.4
1/25[s] for 50Hz system. 0.2
Fig.2 shows the task schedule for data collection from PMUs
0
to PDC and data processing. The phasor angles, effective
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
values, and time stamps during 10-30[min] (respect. Frequency [Hz]
30-50[min], 50-10[min]) are stored to one data file. Then data (a) Band-pass filter
files of eleven PMUs are downloaded in parallel to PDC#1
Fig. 3. Frequency characteristics of digital filters.
through the internet. Since the size of data file is within
600[kB], the download completes in a few minutes. In this
component. Since the frequency band of the local oscillations is
system, the time for download is set at maximum fifteen
relatively near to that of the interarea oscillations, the width of
minutes preparing bad condition of the internet and
pass-band of BPF should be narrow, in this case 0.20[Hz] is
re-download of data files at previous time interval. Because
chosen. The center frequency is adaptively adjustable to the
PDC#1 is busy for data collection, the data files are copied to
oscillatory period of the interarea oscillation. The
PDC#2 just after the data collection within five minutes. And
interconnected power systems parameters are estimated from
next, data processing is executed on PDC#2.
these power oscillations as explained in the next chapter.
Firstly, the frequencies and phasor angle differences of each
Triggered measurements and calculated information caused
area are calculated. There are various frequency components
by disturbances can be automatically logged on PDC#2. For
other than the power oscillations on the phasor angles. The
example, simultaneous change of frequency caused by the
lower frequency component is caused by the frequency
large-scale supply and demand imbalance such as a generator
deviation from the fundamental frequency, which is common to
trip, a load shedding, or a governor test may be detected by the
the interconnected system. This frequency deviation
deviation from the nominal frequency. In this case, the limit of
component is extracted by using LPF (Low-Pass Filter), and is
deviations are set as 59.8[Hz] and 60.2[Hz] for western 60Hz
subtracted from the phasor angle after the compensation of
system, 49.8[Hz] and 50.2[Hz] for eastern 50Hz system,
phase delay caused by LPF. The higher frequency component
49.7[Hz] and 50.3[Hz] for Hokkaido 50Hz system,
is caused by the neighbor load fluctuation. Therefore, this load
respectively. Total processing time on PDC#2 is within ten
fluctuation component is removed by using BPF (Band-Pass
minutes including data visualizing.
Filter). In this way, the power oscillation components are
The power outlet phasors of each PMU are measured every
extracted by using these digital filters. The specifications and
twenty minutes, collected to PDC#2 through PDC#1 in the next
the frequency characteristics of LPF and BPF are shown in
twenty minutes, and analyzed within ten minutes. Therefore,
Table.I and Fig. 3, respectively. The cutoff frequency of LPF is
the cycle time is fifty minutes.
set as 0.15[Hz] not to influence on the power oscillation
3

Hokkaido
Hokuriku 50Hz system

Chugoku Kansai
Tokyo Tohoku
PMU-
Central
Kyushu
Shikoku Chubu Eastern 50Hz systems
PMU-West
Western 60Hz Systems PMU-E1
(500kV) PMU-E2
(500kV)
PMU-East
PMU-E3
(275kV)
G G
PMU-E4
(275kV) G Governor
Chubu Electric Power Company Test

Fig. 4. PMU installation on Japanese western 60Hz systems and on power stations of Chubu electric power company.

B. Validation of Power Outlet Phasor Governor Test


0.02
West
In order to validate the power outlet phasor, PMUs are Central
temporarily installed in the power stations inside of Chubu 0 E3
E2
electric power company as shown in Fig.4. PMU-E1 is in the
Frequency deviation [Hz]

-0.02 E1
500[kV] power station at west end of Chubu, where is the
-0.04
interconnected point to Kansai. PMU-E2 is in the 500[kV]
power station at east end, and PMU-E3 and PMU-E4 are in the -0.06
East
E4

275[kV] generating stations. Actually, the instantaneous


-0.08
voltage of step-down potential transformer is measured for the
phasor computation. On the other hand, PMU-West, -0.1
PMU-Central, and PMU-East are in the 100[V] power outlets -0.12
at the central supply area of Kyushu, Kansai, and Chubu
electric power company, respectively. -0.14
0 1 2 3 4 5 6 7
Fig.5 (a) shows the transient responses of the frequency
Time [s]
deviation during 7[s] after a generator governor test. The (a) Frequency deviation
frequencies gradually fall to 0.12[Hz] according to supply and Governor Test
demand imbalance, and the power outlet phasor properly 0
West
reflects these trend components. During few seconds after the -2
Phasor angle difference [deg]

test operated, the local oscillation appears especially on


-4
PMU-E4 of a generating station. This local oscillation also
appears on PMU-East, because its location is electrically close -6
Central

to PMU-E4. After the local oscillation is damped, the common -8


interarea oscillations appear on all eastern PMUs (PMU-East,
-10
E1-4) against PMU-West. Interaria oscillations between west
E1
and east area are also confirmed on the phasor angle differences -12
E3
from PMU-West in Fig.5 (b). Each phasor angle differences are -14 E4 East
taken offset adjustment as 0[deg] at 0[s]. E2
-16
These results indicate 100V power outlet phasor properly
0 1 2 3 4 5 6 7
reflects the oscillatory characteristics of upper power system
Time [s]
including the power outlet. We can use the power outlet
(b) Phasor angle difference from PMU-West
phasors as a representative of power companies for power
oscillation monitoring. Fig. 5. Transient responses of 100V power outlet phasors and 500kV/275kV
power station phasors at generator governor test.
4

III. ESTIMATION OF INTERCONNECTED POWER SYSTEM


PARAMETERS Hokuriku
Six major power companies are interconnected as a
longitudinal structure with a long trunk line in Japanese Chugoku Kansai
western 60Hz systems, and the interarea oscillations between PMU- ΔPtl
west area and east area remarkably occur as shown in chapter 2. Central
Kyushu Chubu
Now, an equivalent swing equation model between Chubu
Shikoku Δδ
electric power company and the other power companies is
PMU-East
assumed to be as shown in Fig.6. In this chapter, an estimation
method of model parameters from the extracted interarea
Δf
Western 60Hz Systems Generator
oscillation is proposed by showing a case study in Japanese
western 60Hz systems. Tie-line power flow by tele-metering is Trip ΔP
also used for estimating the synchronizing power coefficient in
addition to the frequency and phasor angle difference by PMU.
In Fig.6, Ms and Ds are the constants of Chubu, and Mag and K
Dag are the aggregated constants of Kansai to Kyushu. When an Mag Dag Ms Ds
certain power imbalance such as a generator trip, a load
shedding, or a governor test is occurred, the inertia constant of
whole power systems, which is sum of Ms and Mag, is well
known to be given by the power imbalance (ΔP) and the rate of
frequency change at the disturbance onset time (t=0[s]) as Two-machine equivalent swing equation model
following equation [15]-[16].
ΔP Fig. 6. An equivalent swing equation model of Japanese western 60Hz systems.
M s + M ag = (1)
1 Δf
solving these simultaneous equations. And the damping
f 0 Δt t =0
coefficient is divided in ratio of each inertia constants. Tie-line
The frequency deviation (Δf) is shown in Fig.7 (a). The rate of power flow and phasor angle difference are taken offset
frequency change at the disturbance onset time is estimated as adjustment as 0[MW] or 0[deg] at the disturbance occur.
first order coefficient of a 5th order polynomial approximate The eigenvalue (natural frequency and damping rate) and the
curve of the measured frequency [15]. In this case, the power estimated parameters are summarized in Table.II, and step
imbalance is given by loss of generation capacity of the tripped response of estimated model is shown in Fig.7 (e). The model
generator. response is confirmed to express the interarea oscillation after
The tie-line power flow (ΔPtl) and phasor angle difference the disturbance.
(Δδ) are shown in Fig.7 (b) and (c), respectively. The
oscillation components are extracted by the digital filters as IV. CONCLUSION
explained in chapter 2. Once the power oscillation components
In this paper, PMU based WAMS in Japanese interconnected
are extracted, the natural frequency (ωn) and damping rate (ξn)
power system is explained. The 100V power outlets in
are calculated as shown in Fig.7 (c), and the synchronizing
universities are validated to monitor the interarea oscillations of
power coefficient (K) is estimated as slope of a collinear
western 60Hz systems through a test measurement of actual
approximate curve of the power oscillation components of
power stations with extra-high voltage and ultra-high voltage
tie-line power flow and phasor angle difference, shown in Fig.7
level.
(d). Then the inertia constant (Meq) and damping coefficient
Assuming an equivalent swing equation model, the inertia
(Deq) of the equivalent swing equation are calculated by using
constants, damping coefficients, and synchronizing power
following equations.
coefficients are estimated by using the monitored interarea
2
4M eq K − Deq Deq oscillation information in the case of a power imbalance such as
ωn = ,  ξ n = (2)
2M eq 2M eq a generator trip, a load shedding, or a governor test. Inertia
These equations yield, constants of self system and opponent system are effective
M s M ag K
information for frequency control of the interconnected power
M eq = = (3) system. Other many applications are expected for Japanese
M s + M ag ω + ξ n2
2
n
wide-area power system and many other wide-area power
2ξ n K systems, a transient and steady state stability analysis using the
Deq = (4)
ωn2 + ξ n2 just-in-time model, design the control system to suppress the
Since both (1) and (4) express the relationship between the interarea oscillations such as a power system stabilizer,
inertia constant of Chubu (Ms) and that of aggregated power recognition of conditions of self system and opponent system
system (Mag), the inertia constant is split into Ms and Mag by from time-varying of the parameters, and so on.
5

0.02 TABLE II
PARAMETERS OF ESTIMATED MODEL
Frequency deviation [Hz]

0
First order coefficient of
-0.02 5th order polynomial approximate curve Natural frequency (ωn) 2.575 [rad/s]
-0.04 Δf Damping rate (ξn) 0.183 [1/s]
-0.06 Synchronizing power coefficient (K) 9123
Δt t =0
Inertia constant of equivalent model (Meq) 1369 [MWs]
-0.08
Damping coefficient of equivalent model (Deq) 501 [MWs2]
-0.1
-0.12
-1 0 1 2 3 4 5 6 7 8 9 10 11 12
(a) Frequency deviation Time [s] REFERENCES
[1] K.E.Martin, “Phasor measurement Systems in the WECC”, Proc. IEEE
700
PES General Meeting, 2006.
600 [2] PMU-Related Research Projects of the Power Systems Engineering
Active power [MW]

500 Research Center , 2006.


http://www.pserc.org/PMU_Related_Research_Projects_Apr_2006.pdf
400 [3] J.F.Hauer, N.B.Bhatt, K.Shah, and S.Kolluri, “Performance of "WAMS
300 East" in providing dynamic information for the North East blackout of
Oscillation component August 14, 2003”, Proc. IEEE PES General Meeting, 2004.
200
[4] “ Analysis and Damping of Inter-Area Oscillations in the
100 UCTE/CENTREL Power System”, Cigre Session 2000, WG38-113,
0 2000.
-1 0 1 2 3 4 5 6 7 8 9 10 11 12 http://www.ucteipsups.org/Pdf/Download/englisch/CIGRE38_113.pdf
Time [s] [5] O.Samuelsson, M.Hemmingsson, A.H.Nielsen, K.O.H.Pedersen, and
(b) Tie-line power flow J.Rasmussen, “Monitoring of Power System Events at Transmission and
Distribution Level”, IEEE Transactions on Power Systems, Vol.21, No.2,
2 pp.1007-1008, 2006.
Phasor angle difference [deg]

1.8 A3 [6] J.Bertsch, C.Carnal, P.Korba, M.Ingram, L.Broski, and W.Sattinger,


1.6 ξ n = ln “Experiences and Benefits of Systems for Wide Area Monitoring”, Proc.
1.4 A4 Western Power Delivery Automation Conference, 2004.
1.2 A3 A4 [7] W.Sattinger, J.Bertsch, and P.Reinhardt, “Operational Experience with
1
0.8 Wide Area Monitoring Systems”, Cigre Session 2006, 2006.
0.6 [8] J.Rasmussen and P.Jorgensen, “Synchronized Phasor Measurements of a
0.4
ωn Power System Event in Eastern Denmark”, IEEE Transactions on Power
0.2 Systems, Vol.21, No.1, pp.278-284, 2006.
Oscillation component [9] X.Xie, Y.Xin, J.Xiao, J.Wu, and Y.Han, “WAMS Applications in
0
-1 0 1 2 3 4 5 6 7 8 9 10 11 12 Chinese Power Systems”, IEEE Power & Energy Magazine, Vol.4, No.1,
Time [s] pp.54-63, 2006.
(c) Phasor angle difference [10] T.J.Wang, C.W.Liu, and Y.H.Liu, “A PMU based Special Protection
Scheme : a Case Study of Taiwan Power System”, Electrical Power and
100
Energy Systems 27, pp.215-223, 2005.
Slope of collinear approximate curve
[11] M.L.Scala, M.D.Benedictis, S.Bruno, A.Grobovoy, N.Bondareva,
Active power [MW]

50 D.Denisova, A.J.Germond, and R.Cherkaoui, “Development of


K Applications in WAMS and WACS : an international cooperation
0 experience”, Proc. IEEE PES General Meeting, 2006.
[12] I.C.Decker, D.Dotta, M.N.Agostini, S.L.Zimath, and A.S.Silva,
“Performance of a Synchronized Phasor Measurments System in the
-50
Brazilian Power System”, Proc. IEEE PES General Meeting,2006.
[13] Y.Ota, H.Ukai, K.Nakamura, Y.Mitani, M.Hojo, and O.Saeki, “PMU
-100 based Power Oscillation Detection System and its Application to Japanese
-0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 Longitudinal Power System”, Proc. 15th Power System Computation
Phasor angle difference [deg] Conference, 2005.
http://www.montefiore.ulg.ac.be/services/stochastic/pscc05/papers/fp514.pdf
(d) P-δ Curve of oscillation component [14] R.Tsukui, P.Beaumont, T.Tanaka, and K.Sekiguchi, “Internet-based
0.6 Protection and Control”, IEEE Computer Application in Power, pp.14-17,
0.5 Model step response 2001.
Phasor angle difference [deg]

0.4
0.3 [15] T.Inoue, H.Taniguchi, Y.Ikeguchi, and K.Yoshida, “Estimation of Power
0.2 System Inertia Constant and Capacity of Spinning-reserve Support
0.1 Generators using Measured Frequency Transients”, IEEE Transactions on
0.0 Power Systems, Vol.12, No.1, 1997.
-0.1 [16] D.P.Chassin, Z.Huang, M.K.Donnelly, C.Hassler, E.Ramirez, and C.Ray,
-0.2
-0.3 “Estimation of WECC System Inertia Using Observed Frequency
-0.4 Transients”, IEEE Transactions on Power Systems, Vol.20, No.2,
-0.5 Measured oscillation component pp.1190-1192, 2005.
-0.6
-1 0 1 2 3 4 5 6 7 8 9 10 11 12
Time [s]
(e) Measured oscillation component and model step response

Fig. 7. Measurements and modeling based on equivalent swing equation.

Das könnte Ihnen auch gefallen