Académique Documents
Professionnel Documents
Culture Documents
"Time
-
-
Varying Power Quality Issues:
Varying Power Quality Issues:
An Overview of Advanced Modelling /
An Overview of Advanced Modelling /
Simulation and Control Approaches,
Simulation and Control Approaches,
Applications and Solutions"
Applications and Solutions"
Dr. Paulo Ribeiro Dr. Paulo Ribeiro
Dr. Siddharth Suryanarayanan, Dr. Siddharth Suryanarayanan, Jinglin Jinglin Xu Xu, Dr. Michael Steurer, , Dr. Michael Steurer,
Dr. Dr. Sanjeev Sanjeev Srivastava Srivastava, Dr. David , Dr. David Cartes Cartes, and Dr. Steve Woodruff , and Dr. Steve Woodruff
Department of Engineering, Calvin College, Grand Rapids, MI 4954 Department of Engineering, Calvin College, Grand Rapids, MI 49546. 6.
Center Center for Advanced Power Systems, Florida State University, Tallahass for Advanced Power Systems, Florida State University, Tallahassee, FL 32310. ee, FL 32310.
May 21
rst
to 25
th
2006
FLORIANPOLIS (SC) BRASIL
i
Vision (Strategy)
Vision (Strategy)
Rules
Rules
(Physical, Economics, and Business Imperatives and (Physical, Economics, and Business Imperatives and
Constraints) Constraints)
Road Map
Road Map
(How Things Should Work to Bring Us to the Destination) (How Things Should Work to Bring Us to the Destination)
Man Has Three Sources of Information
(To Guide His Path)
Summary
Summary
-
-
Outline
Outline
Harmonic Distortion on
Harmonic Distortion on
the Transmission Grid
the Transmission Grid
Caused by:
Caused by:
z z
TV Receivers)
TV Receivers)
Harmonic Distortion
Harmonic Distortion
Measurements
Measurements
-
-
230kV
230kV
Substation
Substation
-
-
During World
During World
Cup Soccer Game
Cup Soccer Game
Fortaleza, Brazil, June 1986
Fortaleza, Brazil, June 1986
Brazil 2 X 0 Ireland
Brazil 2 X 0 Ireland
Measurement
Point
The Bigger Picture
The Bigger Picture
(World Cup Soccer and Harmonic Distortion)
(World Cup Soccer and Harmonic Distortion)
Summary of the Measurements
Summary of the Measurements
5
5
th th
Harmonic Voltage
Harmonic Voltage
230kV Bus
230kV Bus
Match Time
The Bigger Picture
The Bigger Picture
Voltage at the 230kV
Voltage at the 230kV
-
-
Detail of 13.2kV Residential Feeder
Detail of 13.2kV Residential Feeder
Disconnection (before the beginning of the match)
Disconnection (before the beginning of the match)
Clearly indicates the origin the
harmonic source / contribution
to the 5
th
harmonic distortion
on the 230kV bus.
Led to Flexibility and
Higher Tolerance
The Bigger Picture
The Bigger Picture
Distorted Time
Distorted Time
-
-
Varying Waveforms
Varying Waveforms
is the Norm
is the Norm
138kV bus voltage THD
0
20
40
60
80
100
120
PQ Disturbance
Power Outage
Cost of Power Outages and PQ Disturbances by Business Sector Cost of Power Outages and PQ Disturbances by Business Sector
Total Annual Cost of Power
Outages and PQ Disturbances
by Business Sector
Cost of:
$14.3
$6.2
$34.9
$66.6-135.6
Source: Primen Study: The Cost of Power Disturbances to Industrial & Digital Economy Companies
Digital
Economy
Continuous
Process
Mfg.
Fabrication
& Essential
Services
Other US
Industry
$Billion
TOTAL
$119 - $188 Billion
40% GDP 60% GDP
The Biggest Picture
The Biggest Picture
The Detail (Road Map) Picture
Combining Spectral and Probabilistic Analysis
How can time-varying harmonic distortion be understood and analyzed from a
more mathematically precise perspective?
Based on some previous mathematical derivations and analysis it has been
observed:
1 - Similarity between spectral analysis and probability distribution functions
2 - The concept of generalized frequency
3 - The concept of evolutionary spectrum
10 100
0
5
10
10
0
ft
j
100 10 j
Continuous Spectrum of a Non-Periodic
Waveform
30 20 10 0 10 20 30
0
5
10
10
0
f
29.883 29.883
lower upper
int
Probability Distribution Function of a
Non-Stationary Process
10 100
0
5
10
10
0
ft
j
100 10 j
The Generalized Concept of Frequency
Suppose that X(t) is a deterministic function which has the form of a
damped sine wave below and illustrated below.
X t ( ) A e
t
2
2
cos
0
t +
( )
40
40
v
i
2 0 i
128
100 50 0 50 100 150
0
5
6
0
ft
j
ft
j
128 128 j
FFT of X(t)
If we represent X(t) as a sum of sine and cosine functions with constant
amplitudes, we need to include components at all frequencies. However, we can
equally well describe X(t) by saying that it consists of two "frequency"
components, each having a time varying amplitude of
A e
t
2
Generalized Frequency
Generalized Frequency
If we have a non
If we have a non
-
-
periodic function X(t)
periodic function X(t)
whose Fourier transform has an absolute
whose Fourier transform has an absolute
maximum at a point w0, we may define w0
maximum at a point w0, we may define w0
as "the frequency" of this function, the
as "the frequency" of this function, the
argument being that locally X(t) behaves
argument being that locally X(t) behaves
like a sine wave with conventional
like a sine wave with conventional
frequency w0, modulated by a "smoothly
frequency w0, modulated by a "smoothly
varying" amplitude frequency.
varying" amplitude frequency.
The Concept of Evolutionary Spectrum
The Concept of Evolutionary Spectrum
Whereas the spectrum of a stationary process describes the
Whereas the spectrum of a stationary process describes the
power
power
-
-
frequency distribution for a whole process (over all
frequency distribution for a whole process (over all
time), the evolutionary spectrum is time dependent and
time), the evolutionary spectrum is time dependent and
describes the local power
describes the local power
-
-
frequency distribution at each
frequency distribution at each
instant of time.
instant of time.
The theory of evolutionary spectra is the only one which can
The theory of evolutionary spectra is the only one which can
preserve the physical interpretation for non
preserve the physical interpretation for non
-
-
stationary
stationary
processes.
processes.
The evolutionary spectrum is a continuously changing
The evolutionary spectrum is a continuously changing
spectrum or in other words, a time
spectrum or in other words, a time
-
-
dependent spectrum.
dependent spectrum.
It is not practical to estimate the spectrum at every instant of
It is not practical to estimate the spectrum at every instant of
time. But if we assume that the spectrum is changing
time. But if we assume that the spectrum is changing
smoothly over time then, by using estimates which involve
smoothly over time then, by using estimates which involve
only local functions of the data, we may attempt to estimate
only local functions of the data, we may attempt to estimate
some form of
some form of
average
average
Applied to time
Applied to time
-
-
varying harmonic distortion
varying harmonic distortion
seems a useful approach, and may help the
seems a useful approach, and may help the
power quality engineer to better understand the
power quality engineer to better understand the
nature of such variations and properly utilize
nature of such variations and properly utilize
analytical tools to predict their behavior.
analytical tools to predict their behavior.
Metrics quantified for 3 cases
Metrics quantified for 3 cases
1) 1)
No additional current harmonic injections on DC
No additional current harmonic injections on DC
-
-
ZEDS
ZEDS
2) 2)
Current harmonic injections corresponding to relaxed V
Current harmonic injections corresponding to relaxed V
THD THD
3) 3)
Time
Time
-
-
varying current harmonic injections and load
varying current harmonic injections and load
corresponding to maximum design criterion
corresponding to maximum design criterion
Case study 1
Case study 1
0.1628
0.1628
C
C
Additional rise in temp
Additional rise in temp
1.21 %
1.21 %
V
V
THD THD
(Phase C)
(Phase C)
0 %
0 %
|
|
7 7
|
|
98.39 %
98.39 %
Life expectancy
Life expectancy
0 %
0 %
|
|
5 5
|
|
0.12868
0.12868
R
R
System conditions
System conditions
No additional current harmonic injections
l
p
ph
h
k
V
V
h
(
(
|
.
|
\
|
=
1
2
1
l
p
ph
H
h
k
i h
V
V
h
K T
(
(
|
.
|
\
|
=
=
1
2
1
) (
1
2
2 2
T T T
T
K
E
e
t
t
+
|
.
|
\
|
=
Harmonic factor
Additional rise in temperature (%)
Life expectancy
Case study 2
Case study 2
2.15
2.15
C
C
Additional rise in temp
Additional rise in temp
80.78 %
80.78 %
Life expectancy
Life expectancy
8.35 %
8.35 %
V
V
THD THD
(Phase C)
(Phase C)
7.5 %
7.5 %
|
|
7 7
|
|
9 %
9 %
|
|
5 5
|
|
0.12868
0.12868
R
R
System conditions
System conditions
Current harmonic injections corresponding to
relaxed V
THD
(9%)
Case study 3
Case study 3
Max Max Min Min System parameter System parameter
21 sec 21 sec 15 sec 15 sec
t t
3 % 3 %
3 % 3 %
0.12382 0.12382
7.5 % 7.5 %
| |
7 7
| |
9 % 9 %
| |
5 5
| |
0.12868 0.12868 R R
Time varying current harmonic injections
1.5076 1.5076 C C
Additional rise Additional rise
in temp in temp
6.31 % 6.31 % V V
THD THD
(Phase C) (Phase C)
86.11 % 86.11 % Life expectancy Life expectancy
Future work
Future work
Input Variations
Input Variations
Input Variations
Expanded Simulink Model
Expanded Simulink Model
Expanded Simulink Model
Expanded Simulink Model
Output can be
Output can be
analyzed for each
analyzed for each
component
component
individually or
individually or
collectively by looking
collectively by looking
at the THD (bottom)
at the THD (bottom)
Statistical of the Fuzzy Logic
Statistical of the Fuzzy Logic
Processing Analysis
Processing Analysis