LED manufacturer fast and loose with data

[Mr. Bill]

Our local USGBC chapter is having a LEED for schools meeting in which a local LED manufacturer will be talking about the savings Schools can gain by using LEDs as a replacement for fluorescent. I am concerned that this LED manufacturer is significantly overstating the performance of his product and playing off of the public?s excitement of LEDs to make a quick profit. Schools need every penny they can right now so I hate seeing them wasting it on snake oil.

I read the description about the meeting on the group's website and a big red flag popped up for me when it describes LED?s as using 95% of its energy for light and only 5% for heat. There is no way this is physically possible. The actual value for LED light energy is a range from 1.5-15%. Manufacturer quality is very dependant on this. Look at this luminous efficacy table on Wikipedia.
http://en.wikipedia.org/wiki/Luminous_efficiency

So I looked into the manufacturer?s website which is very sparse on real data. The first thing I saw was the extremely high lumens/watt performance. Some higher then the best laboratory tested LED?s.
5775 lum / 38W = 152 lum/W
2925 lum / 15W = 195 lum/W
1900 lum / 8W = 237 lum/W
4425 lum / 18W = 245 lum/W
http://www.cleanlightgreenlight.com/CLGLMediaKit.pdf

My job is building lighting design and it still took me a couple of hours to sort thru how they?re fudging the numbers and I still don?t see everything that they?re doing. What hope does a general consumer have?

They?re using ?effective lumens? instead of lumens in their data. Without any description of what effective lumens means or why it?s more appropriate then lumens. The lumen is actually an SI unit which has been used for about 100 years for all sources of light. Looking on another LED website and it appears they?re adding 46% to lumens to create effective lumens. Just because of the color of the light, I think. This would mean the 245 lum/W value above is only 168 lum/W if we are using SI units. My best guess as to what the difference between a lumen and effective lumen is would be the difference between photopic and scotopic vision. This is the difference in perception for the human eye in daytime vs nighttime conditions. Unless the LEDs are being used in small quantities at night there?s no reason to use scotopic, or effective lumen values. Unless to inflate their numbers.
http://www.ideallights.co.uk/corporate/downloads/commercialgradeledT8.pdf (2MB file)

Still 168 lum/W is really, really high performance. And this is where I?m stuck with lack of data. There could be heat losses, driver losses, design losses. I even question where and how the lumens were measured. Has this ever been tested in an independent light lab? An article in LED magazine gives us some idea of what?s going on. Neal Hunter, CEO of LED Lighting Fixtures warns that ?there?s a gap between data and reality?.
http://www.ledsmagazine.com/features/4/4/9

Even their labor savings claim isn?t true. T8 lamps last 20k hours. LED?s last 40k-100k hours. Best case a T8 lamp would have to be replaced 5 times before the LED lamp reached its end of life. These T8 lamp replacements fit into a regular T8 fluorescent lamp socket but the electronic ballast has to be disconnected from the circuit for the LED replacement to work. I?m sure opening the fixture and disconnecting the ballast will take more time then replacing a burned out T8 lamp 5 times.

I also question if these are such wonderful products why do the manufacturers sell directly to customers rather then to lighting designers like myself who have to scratch for each watt of savings on LEED projects. Lighting is a simple yet confusing subject. And not a whole lot of regulation which allows companies to say just about anything, especially for new products. Do they even have IES files so accurate computer lighting models can be make to know how many fixtures are needed in each room?

LED?s are exciting and new. Great for accent lighting, task lighting, cold environments. But I don't feel they?re ready for general building lighting to replace fluorescent. I fear that this presentation will just be a big sales pitch about how perfect LED lighting is. No light source is perfect. Everything has pros and cons. With the advancement of LED lighting maybe in 5-10 years it will be closer to the perfection currently marketed. And maybe better marketing rules so the public knows what they?re buying. There?s even a new Energy Star rating for LED lighting now.

I?m tempted to attend this meeting but I may not have time. Also I can?t say I?m looking forward to being the naysayer in the corner of the room. People like the thought of LED lighting more then they like some engineer in a crowd.

Has anyone else seen this type of marketing or have any additional knowledge about how they're inflating their numbers?

 

[drbond24]

When Benjamin Franklin debated someone, he didn't argue his point so much as he made his opponent argue against himself. Franklin would mostly ask questions, the answers to which were damaging to his opponent.

I recommend that rather than sitting in the corner of the room and being the naysayer, you simply stand up and ask the guy to explain the difference between lumens and effective lumens and why they use effective lumens. Ask him to explain anything else you're worried about while you're at it. In the process of answering your questions he'll have to do your job for you. If he won't (or can't) answer them, he won't seem like such the knight on a white horse anyway.

 

[ptrip]

I would also be very wary of what he's offering.

I've looked into the retrofit tubes for a large office project ... it became a wash cost-wise in my calculations ... over a 10 year life span.

I love new technology. I have been called the "gadget girl". I am extremely excited about the new LED technology ... but there's such a gap in quality from one manufaturer to another on the actual LEDs. This is one of those industries where "you get what you pay for" ... hopefully!

I've had lighting reps bring me garage fixtures, which are being tested by our local university, recessed can lights and show me cut fixtures of exterior area lighting. None of them have mentioned a 2x2 or 2x4 fixture ... or even the replacement tubes ... which would be my main interest, honestly.

My gut response on the recessed downlights was that they would make good accent lights. The lights you would put in front of elevator doors ... where there's still a 2x4 in the vestibule ... or the lights you would put in an alcove or in a display cabinet. Not something you would use for area lighting in a room yet.

The technology just isn't quite there yet. It's oh-so-close. But it's just not there.

The cost is also still prohibitive. The parking light fixtures (which I'm die-ing to try out ... the cut-off is gorgeous from a LEED standpoint) are quadruple the cost of a standard MH lamp. The salesman pointed out that you wouldn't have the ongoing maintenance cost and you wouldn't need as hefty a pole (since the fixture is quite a bit lighter) ... but still.

The quoted distributor cost of the recessed can lights was around $75!

There's no way my cash-strapped schools can afford any of that!

I have one high school and four elementary schools on my design docket for the summer ... all requiring minimum LEED Silver certification ... I (most likely) will not be specifying LEDs (excepting the std exit signs) for any of them.

I did like how most of the projects that company was advertising included: stop signals, accent lighting, freezer/refrigerator (display) and building/facade lighting ... all proven uses for LED at this point in time.

If you can, go ... be the informed nay-sayer. I'd join you if I could!

 

[bbaumer]

Don't let the made-in-China fact and the false advertising scare you. Jump on board with both feet!

Seriously, we are about to test some of these, in small quantity. I suspect a lot of mis-truths will present themselves but expect the product will eventually get to the point where quality and cost make it feasible. That is a few years away though in my estimation.

 

[petersonra]

Why not just try a few and see for yourself.

I am more mad at the CFL lies about how much light they produce. if you say your bulb is equivilant to a 60W incandecant bulb, it should be. Not that it produces the same amount of light, but in a far less useful spectrum.

 

[weressl]

Don't let the made-in-China fact and the false advertising scare you. Jump on board with both feet!

Seriously, we are about to test some of these, in small quantity. I suspect a lot of mis-truths will present themselves but expect the product will eventually get to the point where quality and cost make it feasible. That is a few years away though in my estimation.

I have 4 of these.

3 of them seems to work OK, but I have 2 spots died in about 3 months on one unit with about 2-3 hours of daily use.

The 4th one is another story. First I've lost the entire center row. then lost multiple points in each remaining row and as of 3 days ago I have lost the second row with only one row remaining that has few missing spots.

So while the 80,000-100,000 hours of longevity sounds attractive, it only works if the manufacturing control is assured in how it is put together.

I also have serious trapidations over what will happen to the clear acrylic tube - instead of glass - that houses the assembly. I expect it to deteriorate with age and cloud and crackle(sopider) with age. Yet to be seen.

As far as light output? Probably 1/3 of the equivalent length of the fluorescent. Just eyeballing it though, did not put a meter to it.

Don't do it! Yet......:smile:

 

[ptrip]

I have 4 of these.



As far as light output? Probably 1/3 of the equivalent length of the fluorescent. Just eyeballing it though, did not put a meter to it.

Don't do it! Yet......:smile:

Thanks for sharing your experience!

I need to see if I can con one of my lighting reps into finding me those LED recessed panels to try out in my finished basement!

 

[gndrod]

LED Hype vs. Design

LED Hype vs. Design

I read the description about the meeting on the group's website and a big red flag popped up for me when it describes LED’s as using 95% of its energy for light and only 5% for heat. There is no way this is physically possible.
***Might be possible with substrate Doping technology such as Quantum Dots.

The actual value for LED light energy is a range from 1.5-15%. ***The LED output by itself this is correct.

So I looked into the manufacturer’s website which is very sparse on real data. The first thing I saw was the extremely high lumens/watt performance. Some higher then the best laboratory tested LED’s.
5775 lum / 38W = 152 lum/W
2925 lum / 15W = 195 lum/W
1900 lum / 8W = 237 lum/W
4425 lum / 18W = 245 lum/W
*** These seem very aggressive claims without detailed photometric data.

They’re using “effective lumens” instead of lumens in their data. Without any description of what effective lumens means or why it’s more appropriate then lumens.
*** I agree that lumens description lacks performance characteristics of the Source SSL as there is a big difference when considering Point and Zonal comparisons of the chip being used.

The lumen is actually an SI unit which has been used for about 100 years for all sources of light. Looking on another LED website and it appears they’re adding 46% to lumens to create effective lumens.
***Would this be due to using different surface illuminance reflectivity efficiencies of a particular housing design? CREE has a good example of Light-Source efficacy comparisons between CFL and CREE's XR-E LED XLamp. www.cree.com/xlamp

Just because of the color of the light, I think. This would mean the 245 lum/W value above is only 168 lum/W if we are using SI units.
*** Spectral wavelength does seem to have efficiency differences in output lumens designing to 3000 Kelvin vs 6500 K ranges.

My best guess as to what the difference between a lumen and effective lumen is would be the difference between photopic and scotopic vision. This is the difference in perception for the human eye in daytime vs nighttime conditions. Unless the LEDs are being used in small quantities at night there’s no reason to use scotopic, or effective lumen values. Unless to inflate their numbers.
*** This is hard to relate due to differences in individual eye acuity. Task lighting (scotopic) using concentrated beam widths [40 deg] can be very effective as compared to lesser efficient Mood or daylight (photopic) LED Source defused [110 deg] lighting design.

Still 168 lum/W is really, really high performance. And this is where I’m stuck with lack of data. There could be heat losses, driver losses, design losses. I even question where and how the lumens were measured. Has this ever been tested in an independent light lab? An article in LED magazine gives us some idea of what’s going on. Neal Hunter, CEO of LED Lighting Fixtures warns that “there’s a gap between data and reality”.
*** Truer words were never been spoken.

These T8 lamp replacements fit into a regular T8 fluorescent lamp socket but the electronic ballast has to be disconnected from the circuit for the LED replacement to work. I’m sure opening the fixture and disconnecting the ballast will take more time then replacing a burned out T8 lamp 5 times.
***This is where energy management must start with new construction design.

I also question if these are such wonderful products why do the manufacturers sell directly to customers rather then to lighting designers like myself who have to scratch for each watt of savings on LEED projects. Lighting is a simple yet confusing subject. And not a whole lot of regulation which allows companies to say just about anything, especially for new products. Do they even have IES files so accurate computer lighting models can be make to know how many fixtures are needed in each room?
*** Models do exist but must be correlated with Mfrs and their photometric charts partnered to the Source lamping or SSL's.

LED’s are exciting and new. Great for accent lighting, task lighting, cold environments. But I don't feel they’re ready for general building lighting to replace fluorescent. I fear that this presentation will just be a big sales pitch about how perfect LED lighting is. No light source is perfect. Everything has pros and cons. With the advancement of LED lighting maybe in 5-10 years it will be closer to the perfection currently marketed. And maybe better marketing rules so the public knows what they’re buying. There’s even a new Energy Star rating for LED lighting now.

I’m tempted to attend this meeting but I may not have time. Also I can’t say I’m looking forward to being the naysayer in the corner of the room. People like the thought of LED lighting more then they like some engineer in a crowd.

***I would make the time. Like Lazlo, I purchase new LED Luminaires and see how they work in the real environment. Manufacturer's are aggressive in the commercial arena. That kind of leaves the cost and quality of LED lighting availability somewhat to be desired for the average consumer currently.

I hope this helps. rbj

 

[LarryFine]

Mr. Rod, if you don't mind, the way you intersperse multiple quotes and responses in a single post is to interrupt the post you're quoting with a [/quote], insert your response, and restart the next section you want to quote with a [quote].

The beginning quote with the quoted party's name and the final end quote are already there. What using [/quote] and [quote] does is makes your words appear outside the blue-background area, and you don't have to use a string of *'s.

 

[pbeasley]

In general, I would agree that a good dose of pessimism is a good thing with this information.

I will add that the LED industry in general is having a problem with the concept of light output measured in lumens. The lumen measurement takes into account light energy at multiple bandwidths. Since the output of most "white light" LED's are in an extremely narrow band, they are at a disadvantage, some would even add unfair tothat, when using traditional measurement methods. There is a project underway to create a more accurate light output measurement. Hopefully that will be finalized soon. There is a lot of interest in this project, especially as more and more of the traditional lamp manufacturers start to produce LED lamp.

In the mean time, buyer beware and remember everything you have learned during the birth of the CFL... It's going to be the same issues all over again.

 

[Mr. Bill]

I hope this helps. rbj

Thanks for the response.

This is hard to relate due to differences in individual eye acuity. Task lighting (scotopic) using concentrated beam widths [40 deg] can be very effective as compared to lesser efficient Mood or daylight (photopic) LED Source defused [110 deg] lighting design.

I disagree that task lighting is scotopic. When you're walking at night using only starlight then scotopic vision is used. Anything over a few footcandles would be photopic. Desks are usually 30-50 fc which is well within the photopic range. Adding task lighting would only increase this further. Nighttime driving or lit parking lots would be a mixture of the two, or mesopic.
http://www.visual-3d.com/Education/LightingLessons/Documents/PhotopicScotopiclumens_4 _2_.pdf

I was just thinking what if hybrid car manufacturer's had started advertising cars that got 100 effective miles/gallon. 'Well, it's a totally different engine design so we felt miles had to be measured differently.' I see what these LED manufacturers are doing as the same thing. It's just less noticable since the lamps don't have odometers and we're not paying for it's energy individually each week.

 

[Mr. Bill]

When Benjamin Franklin debated someone, he didn't argue his point so much as he made his opponent argue against himself. Franklin would mostly ask questions, the answers to which were damaging to his opponent.

I recommend that rather than sitting in the corner of the room and being the naysayer, you simply stand up and ask the guy to explain the difference between lumens and effective lumens and why they use effective lumens. Ask him to explain anything else you're worried about while you're at it. In the process of answering your questions he'll have to do your job for you. If he won't (or can't) answer them, he won't seem like such the knight on a white horse anyway.

The meeting is in 2 weeks. If I go I'll try this technique. Already started a list of questions. Also contacted the meeting organizer about my conserns. She seemed to appreciate my concerns and said it would be good if I could attend.

I have 4 of these.

The 4th one is another story. First I've lost the entire center row. then lost multiple points in each remaining row and as of 3 days ago I have lost the second row with only one row remaining that has few missing spots.

Don't do it! Yet......:smile:

It's nice to hear from someone with first hand experience of a similar product.

 

[nakulak]

I have a question for you to ask if you don't mind - "while we are all aware of the mercury and other evironmentally hazardous products in other types of bulbs, what dangerous chemicals can we look forward to seeing in these led products ?"

 

[winnie]

I believe that gnd rod made a good point about a possible source for 'effective lumen'; if you have a T-8 tube that directs all of its light downward, then a larger fraction of its lumen output will make it out of (some) fixtures the if you have a T-8 tube that doesn't direct its output.

Still, I agree that the concept of 'effective lumen' is marketing speak intended to hide the less than stellar reality.

I believe that the wikipedia article is behind the times in terms of LED efficiency; the best commercially available white LEDs are in the 100-110 lumen per watt range. Since the spectral efficacy of white LEDs is in the 300-350 lumen per watt range (meaning that if you have 1 watt of _photons_, you have perhaps 300-350 lumens of visible light), 100 lumen per watt suggests an efficiency of perhaps 25-30%. Still nowhere near 95%!

At a local supermarket, they installed a new freezer section that uses LED lighting. The color looks nicer than the florescent lights, but is arguably dimmer (I've not put a meter to it). LEDs probably make sense in this application since they get more efficient when cold.

-Jon

 

[Mr. Bill]

My question list:

- LED technology is advancing rapidly, how do I as a consumer differentiate cheap, older technology with the most efficient current technology?

- The white LED life expectancy is 50,000 hours, what is the LED fixture life expectancy?

- Are these lumen values a calculated total or were these fixtures measured for total lumen output?

- What is the difference between lumens and effective lumens? Isn?t this a bit like someone selling a car that?s 100 effective miles/gallon?

- I?ve heard that old computers and phones can be dangerous to the environment. How should the electronics in these lamps be safely disposed of?

 

[drbond24]

My question list:

- LED technology is advancing rapidly, how do I as a consumer differentiate cheap, older technology with the most efficient current technology?

- The white LED life expectancy is 50,000 hours, what is the LED fixture life expectancy?

- Are these lumen values a calculated total or were these fixtures measured for total lumen output?

- What is the difference between lumens and effective lumens? Isn?t this a bit like someone selling a car that?s 100 effective miles/gallon?

- I?ve heard that old computers and phones can be dangerous to the environment. How should the electronics in these lamps be safely disposed of?

FWIW, I approve. I especially like the last two.

Be sure and let us know how it goes.

 

[Mr. Bill]

There's actually a 36 page US Dept. of Energy report issued last month on LED lamp replacements for fluorescent.
http://www1.eere.energy.gov/buildings/ssl/benchmark.html

Conclusions
CALiPER testing of currently available LED replacements for 4-ft T12 and T8 fluorescent lamps indicates a significant performance gap between the two technologies. Light-emitting diode linear replacement lamps are marketed as one-for-one drop-in retrofits for general fluorescent applications; however, their comparatively low light output could result in unacceptably low illumination levels in retrofit applications. To maintain existing light levels, it would be necessary to install additional LED replacement lamps, thereby diminishing potential energy savings and requiring additional fixtures or fixture modifications.

 

[legato]

One thing that I always ask LED reps that stop by our office with new LED products is the actual power consumption vs. what the LED uses...

Its easy to say an LED uses 1W of power and provides X amount of light...But what they fail to mention (unless you ask them the question) is how much the driver actually draws, are they overdriving the LED, and how easy is it to replace LED components that fail.

 

[mxslick]

I have 4 of these.

3 of them seems to work OK, but I have 2 spots died in about 3 months on one unit with about 2-3 hours of daily use.

The 4th one is another story. First I've lost the entire center row. then lost multiple points in each remaining row and as of 3 days ago I have lost the second row with only one row remaining that has few missing spots.

Where did you get those from? How much did they cost?

 

[bbaumer]

Just sat through a presentation by a local distributor on lamps manufactured in Taiwan by a company called All-Line International.

These guys are offering a 50,000 hour guarantee on the lamps. Right now they are 120V only but claim they will have 277V available this summer.

Also not over-stating output lumens, I don't think anyway. 1200 lumens/lamp. They are claiming the perceived output is equal to or greater than a F32T8.

They gave us 4 sample tubes to experiment with.

More to come..........