1. Characteristics & Issues for Compact Fluorescent Lamps (AS/NZS EL041 Meeting) Neville R. Watson, University of Canterbury 12 March 2008

2. Outline 1.Background 2.Types of CFLs 3.Direct Harmonic Penetration Study 4.Detailed Transient Simulations 5.Other Studies 6.Standards 7. Conclusions

3. Missing Components Fuse on Input Filter Capacitors missing

4. Same brand September 2007 No Fuse No place for Filter Capacitors Does have PTC

5. PTC missing

7. 3 rd Pro. Project 2008

8. Power-factor

9. Types of CFL Ballasts

10. Types of CFL No Power-Factor Control Passive Power-Factor Control Valley-Filling (or equivalent) Active Power-Factor Control

11. Eclipse 20W CFL

12. OSRAM DuluxStar 20W (2007)

13. E-Lite CFL (July 2007)

14. Eco-Bulb (2007)

20. What is possible for a CFL CFL for North American Market

21. What is possible for a CFL

22. Active Power-Factor Control Basic, no filtering Basic, with filteringValley-fill or Equivalent

23. Active Power-Factor Control Basic, no filtering Basic, with filteringValley-fill or Equivalent

24. Notice: form bands based on circuit type

26. Analysis Methods

27. Frequency Domain Analysis Direct Harmonic Penetration Study

28. CFL Characteristics

30. Test System 28,800 (15 4 10 8 6) customers modelled

31. Breakdown of Losses into Branches Branch No. Description 9House Loads 8Service Mains 7LV Feeders 6300 kVA Transformers 511 kV Feeders 433/11 kV Transformers 333 kV Feeders 233/220 kV Transformers 1220kV System

32. Breakdown of Harmonic Losses into Frequencies

33. This is like comparing apples and oranges

37. Voltage Total Harmonic Distortion

38. 21 st Harmonic (1050 Hz) Distortion Level

39. Detailed Transient Simulations

40. Effect of voltage distortion on current waveform

41. Harmonics Very peaky current waveform rich in harmonics

43. 21 st Harmonic Current

45. 4. Detailed Transient Simulations

46. Fixed Injection Model

47. PSCAD/EMTDC Study SCRLoading 0.1500 + j0.0750 pu0.3000 + j0.1500 pu0.4500 + j0.2250 pu System Strength R=0.0075 L=8.0214e-5 Run 1 Load Point 1 = 5398.7378 Run 5 Load Point 1 = 2699.3689 Run 9 Load Point 1 = 1799.579 R=0.075 L=8.0214e-4 Run 2 Load Point 1 = 539.8738 Run 6 Load Point 1 = 269.9369 Run 10 Load Point 1 = 179.9579 R=0.75 L=8.0214e-3 Run 3 Load Point 1 = 53.9874 Run 7 Load Point 1 = 26.9937 Run 11 Load Point 1 = 17.9958 R=7.5 L=8.0214e-3 Run 4 Load Point 1 = 5.3987 Run 8 Load Point 1 = 2.6994 Run 12 Load Point 1 = 1.7996

49. Note magnification of injection for CFLs along feeder

51. Magnification of Voltage along feeder

52. Weaker system so oscillations at lower frequency

54. Measurements made on a Domestic Supply Oscillations

58. Increasing Decreasing For each harmonic there is a system strength and loading where a maxima occurs

59. Increasing Decreasing

60. Increasing Decreasing For each harmonic there is a system strength and loading where a maxima occurs

61. Increasing Decreasing

62. Other Studies

63. 5. Other studies Study 1: D.J. Pileggi, T.J. Gentile, A.E. Emanual,… et al, The Effect of Modern Compact Fluorescent Lights On Voltage Distortion, IEEE Trans. Of Power Delivery, Vol. 8, No. 4, Oct. 1993, pp 2038-1042 Study 2: F.V. Topalis, Efficiency Of Energy Saving Lamps And Harmonic Distortion In Distribution Systems, IEEE Trans. of Power Delivery, Vol. 8, No. 4, Oct. 1993, pp 2038-1042 Study 3: T.-M. Zhou, X.-Y. Zhu, Y.-L. He, W. Cheng and J. Schlejen, Preliminary Investigation to the Effect of Harmonic Distortion by CFL on Quality of Electric Power Systems, (Published?) Study 4: N. Gothelf, Power Quality Effects of CFLs– A Field Study, RIGHT LIGHT 4, 1997 VOLUME 2, pp. 77-81

64. Study 1 Electronically ballasted fluorescent lights with highly distorted current may jeopardize the reliability of the distribution system and the quality of the electric power delivered. D.J. Pileggi, T.J. Gentile, A.E. Emanual,… et al, The Effect of Modern Compact Fluorescent Lights On Voltage Distortion, IEEE Trans. Of Power Delivery, Vol. 8, No. 4, Oct. 1993, pp 2038-1042

65. Study 2 These conclusions permit the formulation of the opinion that the extensive future use of the energy efficient lamps must be associated with simple and low cost filtering and power factor correction techniques F.V. Topalis, Efficiency Of Energy Saving Lamps And Harmonic Distortion In Distribution Systems, IEEE Trans. of Power Delivery, Vol. 8, No. 4, Oct. 1993, pp 2038-1042

66. Study 3 T.-M. Zhou, X.-Y. Zhu, Y.-L. He, W. Cheng and J. Schlejen, Preliminary Investigation to the Effect of Harmonic Distortion by CFL on Quality of Electric Power Systems Four Cases: Single Lamp Home Lab retrofit Field Experiment

67. Study 3 Our experiments show that for CFL THD according to the new IEC 1000 proposal, in extremely high home applications (5 CFLs per home), the contribution to the V-THD is negligible, compared to TV and much less than computer.

68. Home Experiment ? ? THD I PC 120% (NZ 66.6%) CFL 101-103% (NZ 120%) TV 90% (NZ 120%)

69. Lab. Retrofit Test

70. V-THD variation 1.5-2%. Experiment repeated 8 times

71. Study 4 The test was divided in two phases: Phase 1: Measurements in a one-family house. The measurements were taken first without CFLs and then after installing five CFLs. Phase 2: Measurements in a residential district. The measurements were taken in a residential district consisting of 17 houses at existing load and then after installing of three and six CFLs respectively in each house.

72. Study 4 The study shows that replacement of incandescent lamps with CFLs is beneficial both for users and for utilities. The main advantages of CFLs are: - reduced energy consumption - long lifetime - released capacity of the distribution system High harmonic distortion is the main reason that utilities hesitate to advocate increased use of CFLs. They focus mainly on the high relative current distortion. It is true that for CFLs, the relative current distortion expressed in percent of the fundamental may exceed 100%. However, since fundamental current is very low …, the values of harmonic currents are very low too.

73. Study 4 The results indicate, that the harmonic generated by the CFLs in residential districts have only a minor effect on power quality of the supply network. N. Gothelf, Power Quality Effects of CFLs– A Field Study, RIGHT LIGHT 4, 1997 VOLUME 2, pp. 77-81

74. This current waveform is significantly better than most we see in sold in New Zealand

75. Some of the Brands Tested Basix Canopower Connection Dura Lamp Eclipse Ecobulb Elite Everyhome GE Kempthorne LuxTec Marexim Mirabella Nelson Lamps No Frills Osram Panasonic Philips Results Signature Range SmartLamp Toshiba Wotan

76. Other Appliances

80. Standard washing 1.9 Amps THD I = 132% Spinning and Pumping 7.9 Amps THD I = 34%

81. Appliances Tested TV VCR DVD Stereo Clock/radio Home entertainment systems PCs Monitors Printers Scanners Microwave ovens Mills Halogen lights Fluorescent Lamps Fridge/freezers Freezers Washing machines Dryers Plug packs

82. Heat-pump tests 2008

83. 6. Conclusions The wide-spread use of CFLs can be expected to reduce power quality. The weaker networks exhibit less correlation in the harmonic current injection than stronger networks. Resonances between CFL and ac network magnifies some frequency components (even for a relatively strong ac network)

84. Acknowledgements Tas Scott & Stephen Hirsch, Orion N.Z. Ltd Vinod Kumar, Whisper Tech Limited Joseph Lawrence, EPE Centre Manager Lance Frater Ken Smart Geoff Neville, Enermet N.Z. Ltd.

85. The End Any Questions?

86. Valley Fill Circuit (Cct. 2)

87. Improved Valley Fill Circuit

88. Active Power-Factor Control