High Frequency Distortion in Power Grids due to Electronic Equipment

1
High Frequency Distortion in Power Grids due to
Electronic Equipment Anders Larsson Luleå
University of Technology
2
Outline of the presentation

• Background and motivation of the work
• Waveform distortion
• Low-frequency distortion (harmonics)
• High-frequency distortion
• Results and analysing methods of measurements
• Measurement on fluorescent tube power by a high
  frequency ballast
• Measurement on multiple fluorescent lamps
• Long term measurement at different locations
• Conclusions

3
Background and motivation of the work
4

• New technology has changed our lifestyle, we use
  more electronic product in our life than ever
• Electronic equipment is often served by a
  regulated power supply
• New technology has made it possible to build
  more energy efficient power supplies
• Often are this type of power supplies nonlinear
• This new technology has brought new distortions
  phenomena to the power grid
• New measurement technology has made it easier
  and cheaper to measure

5
Some questions brought up in the beginning of the
project

• What types of signals can be found in this
  frequency range?
• How do we observe these types of signals?
• What happens when a large number of equipments
  are connected together?
• How does these signals propagate in the LV net
  work?
• Can these signals lead to a barrier to the
  introduction of other equipments such as PLC,
  home care equipments, alarms, audio equipments
  etc?
• Can high frequency distortion lead to
  deterioration of other equipments?

6
Waveform distortionLow frequency distortion
(harmonics)
7
Current drawn by a 100W incandescent lamp
8

• How do we analyze waveform distortion?
• If the current is not sinusoidal it contains
  other frequencies than the fundamental at 50 or
  60 Hz
• One way to analyze the signal is to use the
  Discreet Fourier Transform (DFT) to transfers the
  signal from the time- to the frequency-domain
• There are two reasons to transform the signal
  to quantify the waveform distortion and to
  determent the propagation of the signal

9
Harmonic content of the current drawn by the
incandescent lamp
10
Current drawn by a Computer
11
Harmonic content of the Current drawn by a
Computer
12
Current drawn by a HF- Fluorescent Light
13
Current drawn by a HF- Fluorescent Light
14
High frequency distortion
15

• Almost all new electronic equipments has SMPS
  that uses switching technology in the frequency
  range from about 20 to 80 kHz
• The product standards covering harmonic set
  limits up to about 2 or 3 kHz
• Radio disturbances standards mainly sets limits
  from 150 kHz and up
• High frequency distortion is in this case defined
  from 2 kHz up to about 1 MHz

16
Sources

• Switch Mode Power Supplies
• HF-ballasts
• Active Power Factor Correction
• Power Line Communications
• Other loads containing power electronics e.g.
  converters, dimmers etc.

17
Measurement equipment used
18
(No Transcript)
19
Fluorescent lamps powered by high frequency
ballast
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
Spectrogram of the filtered voltage in the range
between 2 and 150 kHz with 0.5 ms time
resolution, 50 overlap and 1kHz frequency
separation.
25
Multiple lamps
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30

• Some examples of measurement at different
  locations

31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
Some conclusion

• The lamp adds extra high frequency components
• The high frequency components are often
  synchronized with the fundamental frequency
• One lamp generate high frequency notches which
  repeats synchronized with the power system
  frequency
• Published papers describing zero-crossing
  distortion generated by the APFC circuit reminds
  of high frequency notches
• High frequency notches increases with the number
  of lamps but the increase seems not to be linear.
  The STFT shows that these signals is found in the
  lower frequency range

35

• The DFT gives some information about the content
  of high frequency distortion but the time-domain
  information is lost
• The STFT seems like a suitable analysing method
  but in this case when many of the high frequency
  components are synchronized with the fundamental
  it is impossible to get an good frequency
  resolution in the lower frequency range
• There are large deviations between different
  locations and quite surprisingly the highest
  amplitudes were found at the resident
• There is an change of in amplitude of the high
  frequency distortion over time. Some frequencies
  seems to be attenuated by loads coming on while
  other frequencies is generated by the loads

36
Thanks for your attention