No Slide Title - Clemson University

Download Report

Transcript No Slide Title - Clemson University

1
EFFECT OF HARMONICS ON
DISTRIBUTED GENERATION
Dr. Elham Makram
Department of Electrical and Computer Engineering
Clemson University
Clemson, South Carolina 29634-0915
Clemson University Electric Power Research Association
DG Conference, Clemson SC
March 14, 2002
2
Outline
• Overview of the effect of nonlinear loads on Power System Harmonics
• Effect of converter drive and arc furnace
• Effect of harmonics on power factor
• Effect of capacitor placement on harmonics
• Impact of unbalance on harmonics
• Need for accurate modeling for harmonic analysis
• Comparison of Time domain and Frequency domain methods
• Performance analysis of models for nonlinear loads
• Change in the scenario due to addition of Distributed Generation
• DG devices and corresponding systems
DG Conference
Clemson University
March 13-15, 2002
3
Effect Of converter drive and arc furnace
20
20
18
18
16
14
5th
7th
11th
13th
17th
19th
23rd
25th
12
10
8
6
4
2
Voltage Magnitude (%)
Current Magnitude (%)
16
5
7
11
13
17
19
23
25
14
12
10
8
6
4
2
0
0
TRANS5_49
LINE3
LINE2
LINE1
Bus 12
Line Number
The presence of a 6-pulse converter on
the power system causes harmonic
currents to flow through the system.
Voltage distortion is negligible
DG Conference
Bus 11
Bus Numbers
Arc
furnace
produces
voltage
distortion and current distortion; the
distortion in voltage being much
higher than that in current. As such
the AF by itself is more harmful to
other customer loads than the
converter drive.
Clemson University
March 13-15, 2002
4
Effect Of Harmonics on Power Factor
1
0.95
Power Factor
A nonlinear load at a bus causes the
power factor at that bus and at
buses close to it to drop.
0.9
Linear Load
AF Load
0.85
0.8
0.75
Bus 12
1
Bus 11
Bus 8
Bus 4
Buses
Power Factor
0.9
0.8
0.7
Without capacitor
With capacitor
0.6
0.5
Placement
of
power
factor
improvement capacitors at buses
close to the nonlinear load bus
worsens the power at the nonlinear
load bus, and also at buses close to
it.
0.4
Bus 12
Bus 11
Bus 8
Buses
DG Conference
Bus 4
Clemson University
March 13-15, 2002
5
Effect of Capacitor placement on Harmonics
The presence of capacitor banks increases the current and voltage distortion at
the nonlinear load bus and the buses close to it.
Before capacitor placement
Current
Voltage
Voltage
2000
400
2000
100
1000
200
1000
0
-100
-200
0
0.02 0.04 0.06 0.08
Time (seconds)
400
300
-1000
-2000
0
0.02 0.04 0.06 0.08
Time (seconds)
1500
0
-200
-400
0
0.02 0.04 0.06 0.08
Time (seconds)
250
200
200
100
500
100
5
10 15
20 25
Harmonic Order
DG Conference
0
0
10 15
20 25
Harmonic Order
Clemson University
0
0
-2000
0
0.02 0.04 0.06 0.08
Time (seconds)
1500
500
50
5
-1000
1000
150
1000
0
0
0
Magnitude in Volt
0
Magnitude in Ampere
200
Magnitude in Volt
Magnitude in Ampere
Current
After capacitor placement
5
10 15 20 25
Harmonic Order
0
0
5
10 15
20 25
Harmonic Order
March 13-15, 2002
6
Magnitude in Amperes
Impact of unbalance on harmonics
In addition to characteristic (6n1)
harmonics, unbalance introduces
uncharacteristic (triplen) harmonics
400
200
0
-200
-400
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
Time (seconds)
Magnitude in percent
25
0.97
20
15
10
5
0
0
0.96
5
10
15
Harmonic Order
20
25
Power Factor
0.95
0.94
Nonlinear Model
Linear Model
0.93
0.92
0.91
0.9
0%
5%
10%
15%
Degree of Unbalance
DG Conference
20%
Unbalance adversely affects the
power factor. Comparison is shown
between the results obtained using
a linear model and those obtained
using a nonlinear model for the arc
furnace
25%
Clemson University
March 13-15, 2002
7
Impact of unbalance on harmonics
• Degree of unbalance affects harmonics differently, i.e. with an increase in the
degree of unbalance, the non-characteristic harmonics increase, whereas the
characteristic harmonics show a decrease.
• The symmetrical component decomposition of individual harmonics helps to
give an insight into the behavior of harmonics under varying degrees of
unbalance.
• Unbalance adversely affects the power factor, although if the unbalance is
within practical limits, the drop in the power factor is not very significant.
• Unbalanced conditions render certain models for nonlinear loads, to be
inadequate for use.
DG Conference
Clemson University
March 13-15, 2002
Comparison of Time domain and Frequency domain methods
8
Current in line 3 (% magnitude)
14
12
10
8
Time
Frequency
6
Frequency domain method of harmonic
analysis gives pessimistic results, and is
not equipped to accurately solve for
certain types of harmonic analysis
problems.
4
2
0
5
7
11
13
17
19
23
25
Harmonic Number
25
The DC side load also determines the
amount of harmonics injected into the
system. There is no provision in the
frequency domain method to model this.
Current Magnitude (%)
20
15
DC load
1/4 DC load
1/8 DC load
10
5
0
5
DG Conference
Clemson University
7
11
13
17
19
Harmonic Number
23
25
March 13-15, 2002
9
Performance analysis of models for nonlinear loads
Voltage-Current characteristics for the arc furnace
Using a Nonlinear model
400
200
300
150
200
100
Voltage (V)
Voltage (V)
Using a Linear model
100
0
-100
50
0
-50
-200
-100
-300
-150
-400
-150
-100
-50
0
50
100
150
-200
-150
Current (kA)
-100
-50
0
50
100
Current (kA)
Preserving the nonlinearity of the problem yields a more accurate solution.
DG Conference
Clemson University
March 13-15, 2002
150
10
Original Picture
Traditional
Distribution
System
NONLINEAR LOADS
Residential
Commercial
Industrial
DG Conference
Clemson University
March 13-15, 2002
Change in the scenario due to addition of Distributed Generation
•
System would no longer be radial
•
DG devices would have power conditioning sub-systems, which would inject
harmonics into the distribution system as well
Traditional
Distribution
System
NONLINEAR LOADS
Residential
DISTRIBUTED
GENERATION
SOURCES
Commercial
Industrial
INTERCONNECTION
DEVICES
• This situation would be different from the case of multiple nonlinear loads
DG Conference
Clemson University
March 13-15, 2002
11
DG devices and corresponding systems to be considered
DG device
Interconnection
Type of system
Photovoltaics
Inverters
Residential
Microturbines
Converter – Inverter
pair
Commercial
Wind Energy
Induction generator –
Converter - Inverter
pair
Industrial
DG Conference
Clemson University
March 13-15, 2002
12
Preliminary results: Impact of a PV System
A Typical Residential PV System
DG Conference
Clemson University
March 13-15, 2002
13
14
Modeling of Residential System
Residential system is modeled by considering four typical houses.
Loads considered in each house are:
Linear Loads (assumed):
 Incandescent light
 Refrigerator load
Nonlinear Loads:
 Compact Fluorescent Lights
 Television Set
 Heat pump
DG Conference
Clemson University
March 13-15, 2002
15
Cases Studied
Case 1:
DG as a standalone unit supplying residential loads
Case 2:
Residential system loads fed by the distribution system only
Case 3:
Residential loads fed by PV and the distribution system
DG Conference
Clemson University
March 13-15, 2002
16
Results for Case 1
Voltage and current at the load bus
Magnitude (amperes)
150
100
50
0
-50
-100
-150
4.82
4.84
4.86
4.88
4.9
4.85
4.87
4.89
Time(seconds)
Frequency Spectrum
Frequency Spectrum
25
20
15
10
5
DG Conference
60
40
20
0
-20
-40
-60
4.83
Time(seconds)
30
25
00
Load current waveform
Magnitude (%)
Magnitude (%)
Magnitude (volts)
Load voltage waveform
5
10 15 20 25
Harmonic Order
20
15
10
5
00
Clemson University
5
10 15 20 25
Harmonic Order
March 13-15, 2002
17
Results for Cases 2 & 3
COMPARISON OF LOAD VOLTAGE
Case 2
Case 3
10
9
Magnitude (%)
8
7
12.47/4.16
4.16/0.208
1- LOADS
6
SUBSTATION
3- LOADS
5
4
LINE REACTOR
HOUSE1 HOUSE2 HOUSE3 HOUSE4
3
RESIDENTIAL LOADS
2
TRANSFORMER
DG
1
0
3
5
7
9
11
13
15
17
19
21
23
25
Harmonic order
DG Conference
Clemson University
March 13-15, 2002
18
Results for Cases 2 & 3
COMPARISON OF LOAD CURRENT
Case 2
Case 3
50
45
Magnitude (%)
40
35
12.47/4.16
4.16/0.208
1- LOADS
30
SUBSTATION
3- LOADS
25
20
LINE REACTOR
HOUSE1 HOUSE2 HOUSE3 HOUSE4
15
RESIDENTIAL LOADS
10
TRANSFORMER
DG
5
0
3
5
7
9
11
13
15
17
19
21
23
25
Harmonic order
DG Conference
Clemson University
March 13-15, 2002
19
Conclusions
• Harmonics produced on the load side by the PV system
were not significant compared to the high current
distortion caused due to the household nonlinear loads.
• Harmonic distortion injected into the distribution system
decreased after connecting the PV system
• The harmonics picture may change than the way it
appears now when number of PV systems connected to
the distribution system increase.
DG Conference
Clemson University
March 13-15, 2002
20
Work In Progress
•
Modeling of other types of DG – Micro turbines, Wind Energy
•
Effect of DG on harmonic distortion in case of commercial and
industrial systems
•
Impact of varying degrees of penetration of DG
•
Combined effect of different combinations of DG devices
DG Conference
Clemson University
March 13-15, 2002
21
Future
Power
System
Transmission system
Load 2
Load 5
Load 1
Solar
Energy
Fuel cell
Microturbine
Load 6
Load 3
Load 4
Wind
energy
Distribution system
DG Conference
Clemson University
March 13-15, 2002