1. Laboratory and field measurements of the power factor and the harmonic emission from energy-efficient lamps EEDAL, 26 May 2011 Sarah Rönnberg, Mats Wahlberg, Math Bollen Luleå University of Technology, Sweden Presented by Peter Bennich The Swedish Energy Agency

2. Background Replacement of incandescent lamps by energy- efficient lighting Most likely candidates: CFL and LED Disadvantage: heavily distorted waveform Question: is this a problem ?

3. Waveform of an incandescent lamp Current is a nice sine wave, just like the voltage No problems are expected in the grid

4. Compact fluorescent lamp Current is heavily distorted Similar for most LED lamps Concern by network operators that this could become a problem

5. Mixed load Current of one lamp is known to be heavily distorted. But its current is small Unknown is how much these lamps contribute to the total distortion of the current taken by a domestic or commercial customer Two experiments were conducted to find out more about this.

6. Experiment no. I Build a house –Full-scale electric model –Normal electrical equipment –30 incandescent lamps (60 Watt) Replace lamps by CFL and LED –Measure harmonic currents from the total installation

7. 7 The Experiment Full-scale electrical model of a house –Computer, television, microwave, dishwasher, induction stove, heat pump, refrigerator Four test cases: –Past: Only incandescent –Present: Incandescent and CFL –Future: LED and CFL –Far Future: Only LED

8. 108-minute load-switching pattern

9. Active power versus time

10. Reactive power

11. 11 ITHD in Ampere

12. Current harmonic 3

13. Current harmonic 5

14. 14 Harmonic spectrum (99%)

15. Interharmonic spectrum

16. Total power factor

17. Power factor versus active power

18. Experiment no II Choose a hotel –76 rooms Replace lamps by CFL and LED –447 x 40-W incandescent E27  7-W LED –116 x 40-W incandescent E14  8-W CFL Measure harmonic currents from the total installation All lamps had 0.6 power factor

19. 19 Harmonic spectrum (99%)

20. 20 Conclusion from the experiments Minor increase in emission with a domestic customer. No change in emission with the hotel. Consumption and peak current decrease Power-factor alone is not sufficient to quantify emission. Current distortion should be expressed in ampere and not in percent

21. Some more discussion Network operation is a natural monopoly with a regulated income. Replacement by energy-saving lamps reduces consumption and income for the network operators. The increased distortion may require investments in the network; the network operator does not want to carry the risk and pushes for strict emission limits. Long term solutions are needed –Regulation to consider changes in load behaviour –Income to become independent of consumption

22. Some final statements Low-power-factor lamps are NOT less efficient than high-power-factor lamps. High-power-factor lamps are more expensive, create more electronic waste and are likely to have a shorter lifetime and be more prone to light flicker. The use of high-power-factor lamps will reduce the risk for the network operator.

23. Extra: Definitions The apparent power, S, and active power, P, can always be measured as: Measure the momentaneous power And integrate over one or more periods : This gives the Power factor PF = P/S ≠ cos φ unless sinusoidal

24. Extra: An example on losses A household with 3 * 7 incadescent lamps (3 phases) of 60 W replaces these with 11 W CFL, thus providing roughly the same light flux. The Active power is reduced from 1260 W (1.8 A per phase, sinusoidal) to 231 W (I rms ≈ 0.45 A per phase: fundamental component I 0 ≈ 0.33 A, sum of all harmonic currents I H ≈ ca 0.3 A) The losses in the grid, P loss = R * I 2 which means that the losses are reduced by roughly I after 2 / I before 2 ≈ 6 % In all: The power is reduced by ca 1000 W and the losses in the grid is reduced by 94 %!