View
218
Download
0
Category
Preview:
Citation preview
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 1/32
Simulation of Wide Area Frequency
Measurements fromPhasor Measurement Units (PMUs)
or
Frequency Disturbance Recorders (FDRs)
Ghadir Radman and Nick Hodges
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 2/32
Outline
• Motivation for the Study• What is Meant by Power System Frequency at Certain
Bus/Location?• Frequency Measured by PMUs/FDRs• Simulation of Frequency Measured by PMUs/FDRs
• Proposed Methods for Simulation of Frequency Measured byPMUs/FDRs
• Some Comments: – Frequency Wave Propagation – Bus Frequency Calculated by PSS/E
– Equivalent Inertial Center Frequency• Five-Bus Example System• Other Test Systems• Discussion
2
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 3/32
Motivation for the Study
• It is anticipated that PMUs and FDRs will be integrated with powersystem for real time control: – Power System Stabilizer (PSS) – Automatic Load Frequency Control (ALFC)
• It would be desired to simulate the system performance before theactual implementation: – In actual implementation we use the measurements from FDRs
and/or PMUs – What should we use for those measurements in simulation
phase?
• We propose a method of estimating the system frequency atcertain bus / location that simulates the measured frequency byPMUs / FDRs
3
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 4/32
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 5/32
What is Meant by Power System Frequency at
Certain Bus/Location? (Cont.)
• The voltage at a generic bus is contributed by all
generators in the power system. Thus this voltage
involves the frequencies of all generators
• Question:
– What is meant by power system frequency at certain
location/bus?
– Which frequency is measured by PMUs / FDRs?
5
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 6/32
What is Meant by Power System Frequency at
Certain Bus/Location? (Cont.)
• Answer:
– In Steady State all generators have the same frequency
( f ss Hz). The ith machine may be considered as a voltagesource of:
– Thus, the voltage at all points of the power system will
have the same frequency ( f ss) and this is the frequencymeasured by PMUs and/or FDRs.
( )issii
t f E t e δ π += 2cos)(
6
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 7/32
What is Meant by Power System Frequency at
Certain Bus/Location? (Cont.)
• Upon a disturbance the power system enters into a dynamicstate and different generators will run with differentfrequencies that vary with respect to time.
In this case, a generator may be considered as a voltagesource of:
where, E i , f i , and δi are slow varying functions of time. Thesetime functions may be found by solving the DifferentialAlgebraic Equations (DAE) representing the power system.
( )iiii
t f E t e δ π += 2cos)(
7
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 8/32
What is Meant by Power System Frequency at
Certain Bus/Location? (Cont.)
At a given time t, for a small interval or window of time
(Δt = n cycles), E i , f i, and δi may be assumed constants.
During this interval, the power system can be represented as:
Transmission
Network
SERIES
DEVICES
[ ]Y-matrix
LO A
D
LO A
D
LO A
D
.
.
.
. . . .
.
.
.
SHUN
T
D
EVI
CE
.
SHUN
T
DEVI
C
E
. . . . . . jXE
I G
8
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 9/32
Now we have a circuit with several voltage sources of differentfrequencies. The actual voltage at Bus i, using superposition theorembecomes:
where, represents the voltage at Bus i due to generator j .
∑=
=++=G
G
N
j ji N ii
act
i t vt vt vt v
1,,1,
)()()()(
What is Meant by Power System Frequency at
Certain Bus/Location? (Cont.)
( ) ( )GGG N i N N iii
t f V t f V ,,1,11,
2cos2cos θ π θ π ++++=
( )∑=
+=G
N
j ji j ji
t f V 1
,, 2cos θ π
9
)(, t v
ji
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 10/32
What is Meant by Power System Frequency at
Certain Bus/Location? (Cont.)
The equation shows that the voltage at Bus i is a multi-frequency voltage. These frequencies are all close to 60 Hz.
Now, even though the voltage at Bus i involves many
frequencies, we approximate this voltage to an estimatedsingle-frequency voltage (this approximation is acceptable fordynamic studies of power systems) defined as follows:
The frequency f iest represents the frequency of the system at
this location / bus and equals the frequency measured byPMUs / FDRs.
( )est
i
est
i
est
i
est
i t f V t v θ π += 2cos)(
10
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 11/32
Frequency Measured by PMUs/FDRs
• PMUs and FDRs implicitly assume that the estimatedvoltage signal is equal to the actual signal:
or
It should be noted that, the above equality is notpossible and it is only an approximation
∑=
==G N
j ji
act
i
est
i t vt vt v
1, )()()(
( ) ( )∑=
+=+G N
j ji j ji
est
i
est
i
est
i t f V t f V
1,,
2cos2cos θ π θ π
11
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 12/32
Frequency Measured by PMUs/FDRs (Cont.)
The FDRs/PMUs do not know the individual terms in the right
hand side of the equation; they only have access to thesampled value of the total (or actual) signal, vi
act (t), whichthey use to calculate f i
est .
The measurements for a certain interval (or window) of time,(Δt = n cycles) are used to find the value of f i
est for the entireduration of Δt .
12
∑=
==G N
j ji
act
i
est
i t vt vt v
1, )()()(
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 13/32
Simulation of Frequency Measured by
PMUs/FDRs
Simulation of the PMUs/FDRs measurements is alsobased on the same equation as used by PMUs/FDRs:
Unlike the real world measurements, in simulationdomain we know every term in the right hand side of theequation using which we try to find the best value of theestimated frequency, f i
est .
∑=
==G N
j ji
act
i
est
i t vt vt v
1, )()()(
or
( ) ( )∑=
+=+G N
j ji j ji
est
i
est
i
est
i t f V t f V
1,,
2cos2cos θ π θ π
13
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 14/32
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 15/32
Proposed Methods for Simulation of Frequency
Measured by PMUs/FDRs
Method 1- Brute Force: – Assuming the actual voltage is identical to the estimated
voltage as follows:
then we have:
– The above equation can be used for several instants of timeduring the interval of interest. The average will taken andchosen to be the estimated frequency
−
= − est
iest
i
act
iest
i
V
t v
t f θ
π
)(cos
2
1 1
15
( )est
i
est
i
est
i
act
i t f V t v θ π += 2cos)(
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 16/32
Proposed Methods for Simulation of Frequency
Measured by PMUs/FDRs (Cont.)
Method 2--Four-Point Total Square Error Minimization: – In this method, f i
est is found by minimizing the sum of thesquare error at four points. These points are t = T/4, t = T/2, t =3T/2, t = 2T , where T = 1/60. Where the Square Error (SE) at agiven time, t k , is found by:
– Thus the Total Square Error (TSE) is given by the following:
– The minimum total square error is found by taking the firstderivate of TSE(t) , setting it equal to zero, and solving for f i
est .
( )2))()( k est ik
act ik t vt vt SE −=
)()()()()( 4321 t SE t SE t SE t SE t TSE +++=
16
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 17/32
Proposed Methods for Simulation of Frequency
Measured by PMUs/FDRs (Cont.)
Method 3- Proportional Voltage-Frequency:
– In this method, f iest is estimated using the
following equation:
– It is assumed that the contribution of each of the
components on the right hand side to f iest isproportional to its magnitude:
∑=
−=
G N
j jest
i
jiest i jiest
i f V
V f
1
,, cos θ θ
17
( ) ( )∑=
+=+G N
j ji j ji
est
i
est
i
est
i t f V t f V
1,,
2cos2cos θ π θ π
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 18/32
Proposed Methods for Simulation of Frequency
Measured by PMUs/FDRs (Cont.)
Comparison of the Methods:
The table shows a comparison of different proposed
methods in terms of Integral of Squared Error (ISE) and
Normalized Error (NE) :
18
Frequency Estimation
Method
Integral of Squared
Error (ISE)Normalized Error
Brute Force 5.946E-16 1.8888E-07
Four-Point Total
Square Error
Minimization
6.2675E-16 1.9392E-07
Proportional Voltage-
Frequency 6.2985E-16 1.944E-07
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 19/32
Some Comments
Frequency Wave Propagation:
– The frequency measurement from PMUs/FDRs has thecharacteristics of a traveling wave which shows theimpacts of a disturbance travels with a certain speed. Thisspeed (although less than the speed of light) is large
enough for the impacts to be considered instantaneous fordynamic behavior studies of power systems.
– In simulation domain, the impact of any disturbance in thesystem is felt by any part of the system instantly. This is
due to the fact that the transmission system is modeled asa static system by its Y-matrix.The magnitude of the impactmay be different for different locations at different times,similar to PMUs/FDRs measurements.
19
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 20/32
Some Comments (Cont.)
Bus Frequency Calculated by PSS/E:
– PSS/E also calculates a frequency for each bus.
– When plotted for the simulations we performed, it
seems that the frequencies of bus-voltages
estimated by PSS/E are unrealistically different
than the frequencies of the generators which areresponsible for those voltages.
20
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 21/32
Some Comments (Cont)
Equivalent Inertial Center Frequency:
The equivalent inertial center frequency of a system is a singlevalue representing (approximately) the frequency of the entire system.
It is calculated by taking a weighted average of machine frequenciesand machine inertias as follows:
The center frequency is only a function of time (not location /bus) and is, thus, the same for all buses in the system considered.
∑=
= G N
ii
ii
c
H
f H f
1
21
c f
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 22/32
Five-Bus Example System
Bus 4
Bus 5
Bus 1 Bus 3
Bus 2
22
A 5-cycle, three-phase short-circuit fault is applied
to Bus 3
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 23/32
Five-Bus Example System (Cont.)
• Bus Voltage Magnitude from Simulation
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 50
0.2
0.4
0.6
0.8
1
1.2
Time (s)
B u s
V o l t a g e M a g n i t u d e s ( p . u . )
V1
V2
V3
V4
V5
23
Five-Bus Example System (Cont.)
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 24/32
Five-Bus Example System (Cont.)
• Bus Voltage Angle from Simulation
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
-20
-15
-10
-5
0
5
10
15
Time (s)
B u s
V o l t a g e
A n g l e s
( d e g )
th1
th2
th3
th4
th5
24
Five-Bus Example System (Cont.)
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 25/32
Five-Bus Example System (Cont.)
• Frequency Simulation for Bus 1
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 559.95
60
60.05
60.1
60.15
Time (s)
F r e q u e n c y
( H z )
Machine Freq 1
Machine Freq 2
Bus Freq 1/PSSE
Bus Freq 1/Proposed Method
Center Frequency
25
Five-Bus Example System (Cont.)
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 26/32
Five-Bus Example System (Cont.)
• Frequency Simulation for Bus 2
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 559.95
60
60.05
60.1
60.15
60.2
Time (s)
F r e q u e n c y
( H z )
Machine Freq 1
Machine Freq 2
Bus Freq 2/PSSE
Bus Freq 2/Proposed Method
Center Frequency
26
Five-Bus Example System (Cont.)
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 27/32
Five-Bus Example System (Cont.)
• Frequency Simulation for Bus 3
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
60
60.2
60.4
60.6
60.8
61
61.2
Time (s)
F r e q u e n c y ( H z )
Machine Freq 1
Machine Freq 2
Bus Freq 3/PSSE
Bus Freq 3/Proposed Method
Center Frequency
27
Five-Bus Example System (Cont.)
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 28/32
Five-Bus Example System (Cont.)
• Frequency Simulation for Bus 4
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 559.95
60
60.05
60.1
60.15
60.2
Time (s)
F r e q u e n c y
( H z )
Machine Freq 1
Machine Freq 2
Bus Freq 4/PSSE
Bus Freq 4/Proposed Method
Center Frequency
28
Five-Bus Example System (Cont.)
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 29/32
Five-Bus Example System (Cont.)
• Frequency Simulation for Bus 5
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
59.95
60
60.05
60.1
60.15
60.2
Time (s)
F r e q u e n c y
( H z )
Machine Freq 1
Machine Freq 2
Bus Freq 5/PSSE
Bus Freq 5/Proposed Method
Center Frequency
29
Five-Bus Example System (Cont.)
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 30/32
30
Other Test Systems
IEEE 57-bus System: A 5-cycle, three-phase
short-circuit fault is applied to Bus 18
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 31/32
Other Test Systems
31
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
59.8
59.9
60
60.1
60.2
60.3
Machine Freq 1
Machine Freq 6
Machine Freq 9
Bus Freq 18/PSSE
Bus Freq 18/Proposed Method
Center Freq
Frequency Simulation for Bus 18
8/9/2019 Radman Ttu Simulation Wide Area Frequency 20090604
http://slidepdf.com/reader/full/radman-ttu-simulation-wide-area-frequency-20090604 32/32
Discussion
?
32
Recommended