Do you think LEDs are going to replace fluorescent general lighting?

[Handlobraesing]

Because fluorescent lighting is used predominantly in the commercial and institutional field, replacing fluorescent implies LEDs taking over lighting in this field.

Not likely, not anytime soon at least.

Cree announced an achievement of 131 lumens per watt from an LED in lab, however the result is not based on a real life conditions (inside the fixture at 120F and does not factor in output decay.

Since we're discussing the leading edge LED, its only fair to compare against the state of art fluorescent.

Fluorescent technology:
GE Ultramax ballast + GE high efficiency lamp (lumens per watt is the efficiency of the driver/ballast and lamp combined)

• 98 LPW for a three 4' lamp 9500 lumen setup(1)
• 100 LPW for a three 4' lamp 7200 lumen setup(1)

Five year ballast + three year lamps, total system warranty (2)
I don't know the exact terms of this warranty, but it's not unusual for this kind of system warranty to include labor, meaning that if lamps burn out during the warranty, manufacturer pay for the new bulb/ballast and labor.

• 90% lumen maintenance at 16,000 hours *(3)
• ~88% lumen maintenance at 24,000 hours (end of life)*(3)
• 24,000 hours average life (3) usually means when 50% of lamps burn out.

OK, so a vanilla three lamp fixture with the aforementioned internals probably wouldn't cost more than $100/fixture. Just how much do you expect a 7200 or a 9500 lumen LED fixture to cost up-front?

To be competitive with fluorescent technology, the LED system must offer significantly higher useful lumens per watt, defined as lumens coming from the ceiling per watt drawn from the grid.


My own specifiers:
If the LED lamp modules are designated as lifetime, they should be covered by no less than a five year warranty, including re-lamping labor cost. Furthermore, the lumen decay curve must be comparable or superior to the current generation premium fluorescent. The modules must maintain no less than 88% of initial efficacy over a 100,000 hour service life at operating temperature likely to be encountered inside the fixture. Lamp modules failing or falling below 80% within a five year period from installation must be covered by warranty. If the modules would not maintain >80% output throughout the useful life of the fixture, then the modules will not be considered "lifetime". If an auxiliary cooling is necessary, the power consumption of these devices will be penalized against the system efficacy. The total cost of ownership (capital cost, labor&parts, and power) over a five year time frame should offer a justifiable amount of competitive advange over the time proven fluorescent technology.

Do you think LEDs will replace fluorescent anytime soon? "Yes" means you feel that LEDs will be available and offered such that the above specifiers are met.

[winny]

I doubt it. The industry is conservative. The end-users are conservative. They already have fixtures for fluorescent lamps so they will continue to buy lamps for them. Just look what happened when T5 tubes came. Did everyone throw out their T8s? No, they still use them and will continue to do so until the fixture/ballast breaks. When it does break, they want something that looks the same and works the same.

Your example is a bit to conservative/moderate too . A 50 khour ballast will last 11.4 years if they are in operation for 12 hours per day. Philips TL-D Xtreme have a service life of 75 khours at one cycle per day so that's 17.1 years with 12 hours per day and 365 days/year.

Taking color rendering into account too, LEDs have a bit more work to do as well...

[jtr1962]

It's a virtual certainty LEDs will replace fluorescents but not any time soon. First off, many businesses invested heavily in the last few years replacing their old T12 fixtures with T8 ones. Since the life cycle for new fixtures is on the order of 15 to 20 years then it's likely LED fixtures will replace the T8s, but not before their life cycle is up. LED may start to be used in new installations within the next 5 years but some problems need to be worked out. Second, we need 131 lm/W LEDs in actual production. Third, they need to be cost effective on a total life cycle basis with T8 or T5 fluorescents. Fourth, they need to improve the color rendering and color temp variation. A CCT of 6027K with a CRI probably in the high 70s just doesn't cut it for general lighting. Rather, you need to get the CRI at least into the mid 80s, and have LEDs binned into the common fluorescent lamp color temperatures of 3000K, 3500K, 4100K, 5000K, and 6500K with variances of less than 100K. Better yet would be RGB fixtures allowing the user to adjust color temp to their tastes but RGB still has a way to go compared to blue plus YAG phosphor in terms of efficiency. As for ballasts, it is quite easy and cheap to make LED ballasts which outform fluorescent ones efficiency-wise while being cheaper to manufacture.

On the bright side, I feel LEDs will virtually obsolete household incandescent lamps within 5 years as the total cost of ownership of incandescent lighting is somewhat higher than state of the art fluorescent, and will continue to climb along with energy costs. Also, LEDs don't have the disadvantages which have prevented wider CFL adoption, most notably lack of ability to be easily dimmed and shorter life from frequent starts.

[karlthev]

Seems like a great spot to pose a question as to where I might be able to buy under the cabinet LED "fixtures" (if that is the correct term) for my kitchen. I've got those dang halogen lights now which get hot as the dickens and are a bear to change. They also have actually "browned" (not quite blackened) the plastic fixtures in which they are housed under the cabinetry, giving me a real sense of security (sarcasm) as you might well understand. Anyway, ideas of where to get strips, fixtures, etc for LEDs to install under my kitchen cabinets?

Karl

[Handlobraesing]

Originally Posted by jtr1962

It's a virtual certainty LEDs will replace fluorescents but not any time soon. First off, many businesses invested heavily in the last few years replacing their old T12 fixtures with T8 ones. Since the life cycle for new fixtures is on the order of 15 to 20 years then it's likely LED fixtures will replace the T8s, but not before their life cycle is up. LED may start to be used in new installations within the next 5 years but some problems need to be worked out. Second, we need 131 lm/W LEDs in actual production. Third, they need to be cost effective on a total life cycle basis with T8 or T5 fluorescents. Fourth, they need to improve the color rendering and color temp variation. A CCT of 6027K with a CRI probably in the high 70s just doesn't cut it for general lighting. Rather, you need to get the CRI at least into the mid 80s, and have LEDs binned into the common fluorescent lamp color temperatures of 3000K, 3500K, 4100K, 5000K, and 6500K with variances of less than 100K. Better yet would be RGB fixtures allowing the user to adjust color temp to their tastes but RGB still has a way to go compared to blue plus YAG phosphor in terms of efficiency. As for ballasts, it is quite easy and cheap to make LED ballasts which outform fluorescent ones efficiency-wise while being cheaper to manufacture.

On the bright side, I feel LEDs will virtually obsolete household incandescent lamps within 5 years as the total cost of ownership of incandescent lighting is somewhat higher than state of the art fluorescent, and will continue to climb along with energy costs. Also, LEDs don't have the disadvantages which have prevented wider CFL adoption, most notably lack of ability to be easily dimmed and shorter life from frequent starts.

There is a new ballast type called the programmed rapid start type that heats the lamp ends thoroughly before it allows it to strike and it greatly improves the lamp life where it is cycled frequently. I believe you get around 100,000 cycles. This is the type specified on motion detector controlled installations, but unfortunately, this type takes about a whole second to fire up and in the case of a compact type fluorescent, it takes five to ten minutes to reach full output.

On dimming, CFLs are just as easily dimmed as LEDs using a specially designed ballast.

Many consumers found CFLs too expensive in the earlier stage and this is why the market is flooded with low purchase cost that often goes bad about as quickly as a regular light bulb.

Commercial grade ballasts are quite expensive ($15-20 each in bulk, $50-60 ea in bulk for dimming type) but they are usually built far better. I've had quite a few cheapie CFLs die over a few year time span from the integral ballast frying randomly, but I've yet to have a single 4 foot T8 or the ballast go bad. I like it a lot as they're basically maintenance free.

Could I ask why you say LED drivers/ballasts would be cheaper than a fluorescent one? If it's going to be a commercial grade, the power supply/ballast MUST have a power factor correction and yield a power factor of 0.95 or better.

[NewBie]

May 30, 2006 - LED Lighting Fixtures Inc. (LLF) today announced
another record in solid state fixture performance at 80 lumens per watt from its most recent recessed downlight. This product was tested by an independent lab as providing 600 lumens of warm white light at a color temperature of 3100 Kelvin, with a color rendering index (CRI) of 92. Total power supplied to the fixture from a residential voltage AC power source was 7.5 watts. The company used LEDs supplied by Cree, Inc. (NASDAQ: CREE). These results reflect a ten percent increase in efficiency from the company’s previous record of 73 lumens per watt announced on April 24, 2006.
http://www.ledlightingfixtures.com/pr_053006.pdf


Remember, this is the finished fixture efficiency, and with the fluorescents, bulb efficiencies, get lower, once you include the diffuser, fixture, and ballast. Also, remember high CRI (color rendering index) bulbs have lower lm/W numbers. Once you include the top end ballasts, fixtures, and diffusers, you would likely have to reduce your fluorescent efficiency numbers by 20-30%. Toss in a high CRI bulb, and your lumens are down even more.


See also:
The company used CSA International's Atlanta Lighting Laboratory to conduct the product measurement. CSA International is accredited by the Occupational Health and Safety Administration (OSHA) as a Nationally Recognized Testing Laboratory (NRTL).

Not bad, the numbers were done by an independent testing lab.

[brickbat]

Not any time soon, but improvements in fluorescent lighting are small and incremental, while LEDs are imrpoving relatively fast.

Interesting, this: 131 lm/W from Cree

http://www.cree.com/press/press_deta...=1150834953712

[Sub_Umbra]

I thought white LEDs were fluorescent.

[jtr1962]

Originally Posted by Handlobraesing

There is a new ballast type called the programmed rapid start type that heats the lamp ends thoroughly before it allows it to strike and it greatly improves the lamp life where it is cycled frequently. I believe you get around 100,000 cycles. This is the type specified on motion detector controlled installations, but unfortunately, this type takes about a whole second to fire up and in the case of a compact type fluorescent, it takes five to ten minutes to reach full output.

I'm aware of this. Unfortunately, most commodity CFLs don't use PRS ballasts.

On dimming, CFLs are just as easily dimmed as LEDs using a specially designed ballast.

Again true, but unfortunately this isn't universal even though a dimmable ballast probably costs only pennies more than a standard one.

Many consumers found CFLs too expensive in the earlier stage and this is why the market is flooded with low purchase cost that often goes bad about as quickly as a regular light bulb.

It's a sad but true fact that the majority of consumers don't look past the initial purchase price, thus forcing CFL makers to essentially sell garbage in order to make a sale. While on the subject, I feel it would be better if we pushed for wider adoption of linear T8 and T5 in residences rather than CFLs. About the only advantage of CFLs is that you don't need to change your fixture. Long-term their operating costs and replacement costs are far higher than linear tubes. I'm of the opinion that new residences should be required to have linear 4 foot T8 or T5 fixtures in every room rather than standard screw-base sockets. While getting people to change fixtures to adopt linear fluorescents might be a hard sell, if the fixture is already there when you buy the house or move into the apartment you'll use it.

Could I ask why you say LED drivers/ballasts would be cheaper than a fluorescent one? If it's going to be a commercial grade, the power supply/ballast MUST have a power factor correction and yield a power factor of 0.95 or better.

An LED driver is basically a simple constant current circuit. You only need a small inductor plus a few other parts versus the larger, more expensive transformers you need even in electronic ballasts. As for PFC, another small inductor and chip takes care of that. You can probably design a decent LED ballast with PFC which fits in a box the size of book of matches and costs a few dollars. As much as I love fluorescent lights, their driving requirements are far more complex than LEDs, especially if there's dimming involved. Add in the fact that per lumen LEDs use far less raw materials and it's easy to see that eventually they will be way cheaper to manufacture than either incandescents or fluorescents. My guess is in ten years time you'll be able to buy a dimmable 10 watt, 2000 lumen LED complete with integral heatsink and driver electronics to operate directly off 120VAC for under $1. Sure, we have a way to go yet, but seeing how much other electronics have evolved over the last decade I don't feel this is much of a stretch.

In closing, yes, I'm a big fan of fluorescent lighting but I feel the handwriting in on the wall and the days of "gas and glass" are numbered. Outside of 3 chandeliers which will hopefully be retrofitted with LED bulb replacements within a few years, we use linear tubes just about everywhere. Over the last few years we've been gradually changing over to T8s. I anticipate that by the time these fixtures run their life cycle in 20 or so years LED will be a more viable replacement.

[Handlobraesing]

Originally Posted by jtr1962

I'm aware of this. Unfortunately, most commodity CFLs don't use PRS ballasts.

So, the consumer would likely gasp at the price of an LED drop in. If $5.99 ea is excessive for a CFL, $19.95 wouldn't even get a look.

Again true, but unfortunately this isn't universal even though a dimmable ballast probably costs only pennies more than a standard one.

Supply and demand plays a big part...

An LED driver is basically a simple constant current circuit. You only need a small inductor plus a few other parts versus the larger, more expensive transformers you need even in electronic ballasts. As for PFC, another small inductor and chip takes care of that. You can probably design a decent LED ballast with PFC which fits in a box the size of book of matches and costs a few dollars. As much as I love fluorescent lights, their driving requirements are far more complex than LEDs, especially if there's dimming involved. Add in the fact that per lumen LEDs use far less raw materials and it's easy to see that eventually they will be way cheaper to manufacture than either incandescents or fluorescents. My guess is in ten years time you'll be able to buy a dimmable 10 watt, 2000 lumen LED complete with integral heatsink and driver electronics to operate directly off 120VAC for under $1. Sure, we have a way to go yet, but seeing how much other electronics have evolved over the last decade I don't feel this is much of a stretch.

Fluorescent ballast is a constant current power supply too once the lamps are started. The LC resonance creates ~40KHz AC. Advance Transformer makes LED driver and they're not cheap and the level of compexity is about the same. You still need a way of stepping down 277V or 120V AC to working voltage efficiently.

In closing, yes, I'm a big fan of fluorescent lighting but I feel the handwriting in on the wall and the days of "gas and glass" are numbered. Outside of 3 chandeliers which will hopefully be retrofitted with LED bulb replacements within a few years, we use linear tubes just about everywhere. Over the last few years we've been gradually changing over to T8s. I anticipate that by the time these fixtures run their life cycle in 20 or so years LED will be a more viable replacement.

We'll see I suppose. There's been talk of alternative fuel for decades, but it hasn't made much of a change in fossil fuel engine usages.

[James S]

If it's so easy to make dimmable CF bulbs, then why do even the very expensive ones I've experimented with suck so badly? They have a very small range of variation and in my limited experience dont last as long. Probably for the "keep the price down and use crap parts inside" reasons.

Also CF bulbs look horrible in fancy chandeliers. I have a couple here that I like and simply will not put CF bulbs in, but I'd put sparkly LED bulbs in them if they dimmed as nicely as regular bulbs do.

If LED bulbs can be made with a full range dimmable ballast in them, that will help to get them into some niche markets, and from there hopefully to lower prices and serious usage.

I have replaced all my garage and work light lamps with T8 bulbs and electronic ballasts and I LOVE them for that. But I have no interest in them for bedroom lighting. The 2 color choices easily available for replacement bulbs simply dont cut it.

[winny]

Originally Posted by James S

If it's so easy to make dimmable CF bulbs, then why do even the very expensive ones I've experimented with suck so badly? They have a very small range of variation and in my limited experience dont last as long. Probably for the "keep the price down and use crap parts inside" reasons.

Drop me a PM with your address and I'll send you one which doesn't IMHO. That is if you run on 230V...

Originally Posted by James S

I have replaced all my garage and work light lamps with T8 bulbs and electronic ballasts and I LOVE them for that. But I have no interest in them for bedroom lighting. The 2 color choices easily available for replacement bulbs simply dont cut it.

How hard is it to come by what you need? At least here you can get 827, 830, 840, 850, 865, 930, 940 and 950 but you need to order all but the two first so a couple of days of waiting...