Fluorescent switching vs lifespan

Status
Not open for further replies.

Don S.

Member
Can anyone tell me if modern fluorescent tubes and or electronic ballasts are adversely impacted by excessive switching? The issue is, will motion sensors on individual fixtures lead to reduced tube or ballast life expectancy?
 
Can anyone tell me if modern fluorescent tubes and or electronic ballasts are adversely impacted by excessive switching? The issue is, will motion sensors on individual fixtures lead to reduced tube or ballast life expectancy?

The issue mainly has to do with the incandescent filaments at both ends (thus two pins) that provide the initial heat up for the gases to ionize and then the gas takes over as conductor and the flow through the filament is reduced. The startup inrush that is most damaging to the filaments and they eventually break up like normal incandescent bulbs. So, more switching; shorter life. I've heard regular fluorescent ballast having a life expectance of 50,000 hours, but not heard of any switching restrictions.

I wonder if the above is true for dimmed fluorescent. Anyone?
 

iwire

Moderator
Staff member
Location
Massachusetts
The issue is, will motion sensors on individual fixtures lead to reduced tube or ballast life expectancy?

Yes.

But the question should be will that reduced life expectancy outweigh the savings from shutting down lights that are not in use.
 
I occasionally work on fixtures located in public restrooms that are switched from the wall with motion sensors. The lamp life is drastically shorter than the lamps that are in the general retail area.

I have read that a high ballast factor would help extend the life of frequently switched fluorescents. Even though it has the opposite result in a normal use setting. I have never gotten around to trying it. The explanation stated that the higher voltage/current would light the light with less electrons coming off the filaments on the ends of the lamps. In other words it is more tolerant of a lamp with weak filaments and frequent switching weakens the filaments. One of these days I am going to swap a ballast out with one with a ballast factor of one or higher and see if it improves the situation.
 

Mr. Bill

Senior Member
Location
Michigan
I had some chart somewhere showing lamp life vs switching frequency for different ballast types. Can't find it. Current lamp life expectancy about (20k hours for T8) is based on 3 hours on and 1 hour off. As the 'on' duration gets shorter so does the lamp life. It was a very linear relation ship between switching frequency and lamp life.

Generally Instant and rapids start ballasts have a shorter lamp life than with Programmed start ballasts. It's a very significant difference with very frequent switching. So I generally specifiy programmed start ballasts when occ sensors are used. Cost seemed to be about $3-5 more per ballast. The down side is I've noticed instant start ballasts usually use 1-2 watts less than Programmed start.

I haven't seen any information showing a ballast life expectancy difference for switching.
 

LarryFine

Master Electrician Electric Contractor Richmond VA
Location
Henrico County, VA
Occupation
Electrical Contractor
Since electronic-ballasted fluorescents have both wires from each socket tied together, the end filaments are merely electrodes, and don't have any current flow from pin to pin. Wouldn't that eliminate the filament-life from the equation?
 

Don S.

Member
Programmed Ballasts

Programmed Ballasts

I have frequently seen pros and cons of different ballast types discussed, but never an electrical explanation as to why one is better for a specific duty. I would like to know what is different about a programmed ballast, making it more suitable for motion sensor duty.
 

Mr. Bill

Senior Member
Location
Michigan
The cathode (electrode) is typically the failure point for fluorescent lamps. They are coated with an emissive material to emit electrons for starting an arc. You know that black stuff on the inside end of the lamp when it's old. That's where it comes from. The "harder" the start, the more material that's thrown off getting the lamp to start.

The main difference is what voltage is provided at the start of the lamp. Instant start ballasts slam the lamp with voltage to start the lamp NOW. Programmed start ballasts will actually pre-heat the cathode before starting current flow. That's why there is a delay of a couple seconds after flipping the switch and getting light. I've heard from many that don't like the delay. Rapid start ballasts have become less common.

Maybe another way to think of it is like running your car in the winter. Do you let the engine warm up a bit before driving or do your just jump in and floor it.

Try this link and be sure to look at the figure links for voltage and current vs time.
http://www.unvlt.com/literature/programmed.html

Occupancy sensors are expected to turn the lights on and off several times a day. Without the sensor it's usually on once in the morning and off in the evening. As switching on and off becomes more frequent the programmed start ballasts provides a much longer lamp life vs instant start. That's the main reason why it's considered more suitable for occ sensor duty.
 

mivey

Senior Member
Lamp Life:
R=rated life
h=hours of operation per start
A=actual life
A=R*1.71*(1-EXP(-1*(h/3.89)^0.505))


add example:
20,000 hour lamp, $1.50, 34 watts, $0.10/kWh electric rate.

with a 3.1833 minute on off cycle, the lamp hours are reduced to 3,694 with 6,777 lamp cycles, saving $1.22 in electric costs offset by $1.22 in reduced lamp life.
 
Last edited:

mivey

Senior Member
Here is the standard fluorescent:
20khLampLife.jpg
 

Mr. Bill

Senior Member
Location
Michigan
add example:
20,000 hour lamp, $1.50, 34 watts, $0.10/kWh electric rate.

with a 3.1833 minute on off cycle, the lamp hours are reduced to 3,694 with 6,777 lamp cycles, saving $1.22 in electric costs offset by $1.22 in reduced lamp life.
How exactly are you getting $1.22 savings in electric costs?
 

mivey

Senior Member
How exactly are you getting $1.22 savings in electric costs?
Well, it came with no money-back guarantee, but I looked back on what I scratched out. Whether or not it makes sense today is another story, but it sounded good last night :D.

Here is what I actually did:

Started with the IEEE formula for lamp life (I used a 3 hour run/start):
A1 = R*1.71*(1-EXP(-1*(h1/3.89)^0.505))

the revised life (shorter run time per start):
A2 = R*1.71*(1-EXP(-1*(h2/3.89)^0.505))

# of lamp cycles for the reduced time:
#cycles = A1/(h1-h2) = Lamp life / run time reduction

what I called the "Electric savings":
h2*watts/1000*$/kWh*#cycles =
Electric cost per reduced time cycle * #cycles for the reduced time

Value of lost lamp life:
Lamp cost / original life hours * (original life-revised life)

Feel free to add corrections as I certainly don't feel like going through it right now. I would like to eventually see a good formula for electricity savings vs loss of lamp life so have at it.
 

mivey

Senior Member
I had some chart somewhere showing lamp life vs switching frequency ... It was a very linear relation ship between switching frequency and lamp life.
There is a lamp life graph here (figures 4 & 5):
http://www.lrc.rpi.edu/programs/futures/LF-LampLife/

and in the range 3 hours to 20 hours, the graph looks like what I have in post #10. According to the equation, it is not so linear outside that range. I'm not sure what range the formula is supposed to cover but it is supposed to be from "Economics of Switching Fluorescent Lamps" IEEE Transactions on Industry Applications Vol 24, No 3, May/June 1988
 

a.bisnath

Senior Member
type of ballast?

type of ballast?

rapid start ballasttranformer,break circuit inductive kick type 2 pin socket type starter,electronic ballast all will have different effects on life span all have different principles of operation in starting I give the opinion the electronic ballast is the best
 

Mr. Bill

Senior Member
Location
Michigan
Here is what I actually did:

# of lamp cycles for the reduced time:
#cycles = A1/(h1-h2) = Lamp life / run time reduction

what I called the "Electric savings":
h2*watts/1000*$/kWh*#cycles =
Electric cost per reduced time cycle * #cycles for the reduced time
I don't think the number of cycles has any direct affect on the total energy use. I wouldn't suggest using an equation like this to determine energy savings with occupancy sensors.

I found a study done on energy savings with occupancy sensors for various room types. Look at page 17. Costs are based on $0.08/kWh.
http://www.lrc.rpi.edu/resources/pdf/dorene2.pdf
Annual energy savings (20-min setting):
Break room: $35
Classroom: $260
Conf room: $46
Private office: $27
Restroom: $76

I think this more than covers the cost of the sensor and reduced lamp life after a few years. The bottom of the next page also shows expected lamp life.
 

mivey

Senior Member
gar,

Several curves/tables for slimline hours/start & life hours:

3 : 12,000
6 : 15,000
12 : 18,000
18 : 20,000
24: 22,500

3 : 12,000
8 : 19,000
12 : 23,000
16 : 24,500
20: 25,500
24: 26,000

(line segments instead of smooth curve)
3 : 12,000
6 : 14,000
12 : 18,000
24: 22,500
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
090422-0629 EST

mivey:

Interesting. Much lower than your standard fluorescent curve.

This would imply that proper preheat of the emitting electrode has a substantial impact on the oxide coating life.

What has been done in the last 60 years to have greatly increased the life of the heater cathode combination in CRTs in comparison to radio tubes of the 1940s?

.
 
Status
Not open for further replies.
Top