In the recent years I've noted that many people in Israel, thinks that fluorescent lamps consumes more energy during turning on (Starting), and so they mustn't be turned on often.
Is this claim true or false?

Switch Start and Perfect Start do consume more power in the moment of starting (momentary power), Rapid Start and most HF systems don't



However, lets calculate something....



Consider a 36w fluorescet tube on Switch Start. Lets assume that starting the tube takes 5 seconds of flashing and preheating

During those starting flashes, lets assume that the times that the tube is lit sum up to 2 sec, the time it is preheating sum up to 2 sec, and time it is off sum up to 1 sec.

 - 2 seconds that the lamp is lighting (for short flashes) equal the energy consumption of a lit lamp for 2 seconds. With typical Eco ballast the normal power of the lamp is about 42w i think

 - 2 seconds of preheating may take more energy. The preheating lamp takes about 10V on each cathode at 860mA. Together that sums up to about 17w energy used in lamp cathodes. Ballast loses are I^2 R = 860mA^2 * 50 ohm = 37w energy lost in ballast, 37+17 = 54w

 - 1 second that te starter is glowing, lets give it 10w just so nobody says we forgot it, i doubt it take more than 10w anyway....

Amount of energy used in starting :
2*42 + 2*54 + 1*10 = 202 w*sec

Amount of energy used in normal work :
5*42 = 210w*sec

We found out that our example lamp during 5 sec starting, took LESS energy than when it is working for 5 sec



Lets consider a harder starting - 1 second glow, 5 seconds preheating:
1*10 + 5*54 = 280w*sec

Amount of energy used in normal work :
6*42 = 252w*sec

STILL the lamp in starting takes less energy than in working



Lets consider a lamp on a VERY worn out Perfect Start ballast, that takes 10 (!!!!)seconds to heat, then it strikes
10*54 = 540w*sec

Amount of energy used in normal work :
10*42 = 420w*sec (oh, and im assuming that the ancient Perfect Start R1 is as efficient as the modern Eltam Eco)

AT LAST, the lamp during starting takes more energy than when lighting

540w*sec / 42w = 13 sec
So, the 10 SECONDS long starting of the lamp ON A WORN ANCIENT BALLAST takes as much power as..... 13 seconds of normal work. Yes, exactly this : In 3 extra seconds of normal work it offsets the energy use of starting....

Which cost.... 3sec * 42w / 1000 (w in kw) / 3600(sec in hr) * 0.6(NIS 1KWh) = 0.000021 NIS = 0.000005 $



I call BS

This used to be a common misconception here too. It's probably why the lights (especially the kitchen one) are still always left on at my rellies' gaff. The kitchen and extension lights are both Thorn fittings with electronic gear for 2 6' 70W T8 or 75W T12 lamps.


BG

I think the myth came form a comparison originally intended to educate users about fluorescent sensitivity to switching cycles: In order to be somehow understandable, the author make an comparison, when the extra cycle cost on the lamp life the same as the electricity would, if the lamp would be kept on without switching off.
As this "information" spread among people not really understanding what it is about, the "extra cycle cost the same on lamp life as the electricity when kept ON for xxx minutes" got distorted into the "The lamp draw for starting the same energy as over xxx minutes of burning time" and viola, we have the myth...

This myth is indeed quite persistent and I would guess it spread everywhere together with the fluorescents...

The claim about fluorescent lamp energy consumption during starting, appeared in Israel only recently.

Not at all. I heard of it for years. What happened is probably that the myth reached somebody in the mass media (on a TV show etc) and burst in publicity from there

Good question - what type of fixture? Pre-heat, IS, RS etc? Pre-heat might do, has to heat the cathodes first - low volts high current...I guess

Good question - what type of fixture? Pre-heat, IS, RS etc? Pre-heat might do, has to heat the cathodes first - low volts high current...I guess

Did this really belong to this thread?

If yes, then a response:
The preheating is about a normal operating current at fraction voltage (~15V instead of 70..100V), so fraction of the normal power. And it take few seconds only per start...

Well, I don't see why not he did make a valid point there... and still you really don't have to be rude about it...

Well, I don't see why not he did make a valid point there... and still you really don't have to be rude about it...

If it sounded rude, I do apologize for that, it was by far not the intention (English is not my native language an I still do not have the correct feel for how certain phrases sound)...
I got an impression, than the post belonged to another topic and was here only by mistake...

But then I thought someone not knowing how the lamps do start may really argue in that way, so I added the response...

See my estimates up above, preheating does take more than the lamp power (mostly due to higher I2R ballast losses), but marginally. Unless it is a hard start with a lot of preheating, the difference is still offset by the time when the starter is glowing,so all in all it takes less power than lit lamp

in an extreme case with very hard start (10 sec) on a thermal starter ballast, it would take about as much power as 13 sec of normal work.....

@ Medved

No -  that's ok  I didn't think you were being rude. We are world-wide friends and there will always be tricks with translations in native languages.

I have a tendency to reply to the OP without taking account of the conversations of the thread. Please accept my apologies, I should have copied the OP quotation first. It is me who should apologize for the confusion

cheers
Blake

See my estimates up above, preheating does take more than the lamp power (mostly due to higher I2R ballast losses), but marginally. Unless it is a hard start with a lot of preheating, the difference is still offset by the time when the starter is glowing,so all in all it takes less power than lit lamp

in an extreme case with very hard start (10 sec) on a thermal starter ballast, it would take about as much power as 13 sec of normal work.....

But you assume the ballast deliver twice the normal operating current, in my experience that would be quite unusual, all ballasts I've met are at about 1.4..1.6x the operating current (and if I remember well, such figure is in the standard requirements for the "ENEC" certificate).
So in case of a regular 45Ohm ENEC-certified ballast, the total consumption during preheat is about 1.6*0.43*10*2 (filaments) + (0.43*1.6)^2*45 (ballast wire resistance losses) = ~35W, while the normal power input is about the 45W (not counting the resistivity rise with the temperature), so still the consumption during preheat is quite lower than during normal operation.

But anyway, neither the 35, nor the 54W are so far from the normal 45W to be of any significant difference...