DoE Announces 'L Prize' For Solid-State Lighting

erikaaboe notes that the US Department of Energy has announced a competition to develop efficient solid-state lighting technology. The "L Prize" program will allocate as much as $20 million in cash prizes for innovations to replace the common light bulb. Further details are available at the L Prize website. From the press release: "Lighting products meeting the competition requirements would consume just 17% of the energy used by most incandescent lamps in use today. The plan also includes a rigorous evaluation process, including testing of proposed products by independent laboratories (conducted through DOE's CALiPER test program), as well as field evaluations by DOE and utility partners to assess products in real world conditions. Four major California utilities ... have signed a Memorandum of Understanding with DOE, agreeing to work cooperatively to promote high-efficiency solid-state lighting technologies."

Comments

Re:Solid-state?

[Awptimus Prime]

Oh, I see. They want to avoid utilizing a vacuum. This doesn't seem to matter either, as long as someone comes up with a way to do it with greater efficiency.

It'll take a lot of research and effort to figure out how to make a better LED with only (up to) $20m in rewards.

Re:Solid-state?

[David Gerard]

Fluorescents work on a gas being turned into a plasma, so wouldn't qualify. LEDs are solid-state, but are presently very expensive as lightbulbs. Incandescents are fragile, but might be "solid state", but fail on the power requirement.

Re:Sooo.....

[Anonymous Coward]

I'd scoff, but I'm going to give you the benefit of the doubt and assume you're trolling brilliantly. What follows is clearly for the benefit of others. IF they pay the $20 M out, the savings to the economy, in just the US, could be measured in major fractions of a TRILLION. Not that Fermilab isn't very worth while, even critical, but it's more of a high risk, high reward extremely long term investment. You know, like investigating the photoelectric effect.

Significantly bright LEDs are very expensive

[MichaelCrawford]

I expect they'll get there eventually, but they're not practical for regular home or office lighting yet.

They work great for flashlights, and the headlight and taillight on my bike use LEDs.

But I researched LED lights a couple months ago, and found that a "60 watt replacement" LED light was expected to cost well over a hundred dollars, and at that time was still in development, and not yet available.

I finally settled for a couple twisty bulbs, but I'm not too happy about it because they contain mercury.

I'm also not too happy that the mercury warning on the package just advised me to dispose of them "according to local laws". As if it would be OK to let the mercury into the groundwater if there wasn't a law specifically against doing so!

Re:Significantly bright LEDs are very expensive

[Larsrc]

Check out dealextreme.com, they have them available. No, not down to CF prices yet, but significantly more efficient.

Re:Significantly bright LEDs are very expensive

[Yvanhoe]

But I researched LED lights a couple months ago, and found that a "60 watt replacement" LED light was expected to cost well over a hundred dollars, and at that time was still in development, and not yet available. Errr, I see spot bulbs composed of a dozen of lights, claiming to output as much light as a 60W incandescent bulb, for 3 euros at my local store... Is there some factor I am missing ?

Re:Significantly bright LEDs are very expensive

[jabuzz]

Nice try but there has been now mercury in watch batteries for over a decade.

Re:Significantly bright LEDs are very expensive

[TheRaven64]

CFL's aren't worth the energy they save Here, a 60W incandescent costs around 50p and lasts about six months. A 20W CF costs about £3-5. I don't know how long they last because I haven't yet had to replace any of the ones I bought four years ago, but we'll say four years for the sake of argument. Over four years, the capital cost of incandescents is £4, while the cost of CFs is £3-5. If you run it for two hours a day, then you are saving 58kWhs, which is about £6-10 (depending on a lot of factors) at the moment. The TCO of a CF is therefore somewhere between £5-11 lower than incandescents over four years. Scale that up by the number of bulbs in the average house and it's a significant saving.

Re:Significantly bright LEDs are very expensive

[halcyon1234]

I finally settled for a couple twisty bulbs, but I'm not too happy about it because they contain mercury

Oh, please. There's more mercury in your watch battery than in the CFL. And it's not like its posing any actual danger to you. The mercury isn't released into the air when the CFL is broken. If it does break, you can clean it up with a vacuum and a pair of rubber gloves, just like a non-CFL. No need to call in a hazmat team.

Unless you're going to get down on the floor and lick up the broken CFL, it doesn't pose much a risk to you.

Environmentally, the tiny bit of mercury is far outweighed by the massive energy usage and packaging savings the CFL gives you over its life time. The CFL can either be disposed of with your paints and batteries (ie: safely), or some places even have CFL recycling programs to reclaim the materials.

Take a gander at the Natural Resources Canada FAQ about CFLs. It includes a link to a health study about the effects of CFL's mercury.

Re:Significantly bright LEDs are very expensive

[Smidge204]

An LED bulb to replace a 30W incandescent is $45.

Just about all LED lamps available now are made from discreet white LEDs that cost about 50 cents each in bulk. Factor in PCB, power supply and assembly and they're still stupidly expensive.

What they need is a purpose-build manufacturing method to form dozens of LED junctions as a single unit specifically as a bulk light source.
=Smidge=

Re:Significantly bright LEDs are very expensive

[Yvanhoe]

Well, I still have on my kitchen table a spot that has about 15 LEDs (it DOES look like LEDs), cost me 3 euros, haven't tested it yet (the shape is a bit weird and doesn't fit in any of my lamps). It is 16 euros if you want to build one from parts at an expensive retailer. I think that the Chinese factory easily cut down that price by 3 or 4.

Re:Significantly bright LEDs are very expensive

[Anonymous Coward]

There's still dozens of sources of lead that are insanely higher than CFLs. Tire weights are still lead. As are most bullets. Hell, a chip of paint from an old house has more lead than a CFL. Are your computer boards lead free? How about the alarm clock you toss in the trash? Most small electronics still have lead solder and contain far far more heavy metals than CFLs and I haven't even started talking about cadmium and other commonly used heavy metals or coal burning.

Decades? Not really

[JustShootThemAll]

The bright white leds that are currently used are not all that stable. Light output decreases with use because the phosphor coating degrades. Remember that white leds are actually UV-leds that need a phosphor coating not unlike fluorescent tubes.

It takes about 1000 hours for the led to reach 50% light output. The time from 100% to about 85% is measured in single digit hours!

So, no, light fixtures that last for decades are right out. With current technology, that is.

Re:Decades? Not really

[minimum]

Bullshit. This applies to very cheap LED's bought from China (or similar). Leading manufacturers like Cree, Lumileds and the rest claim 75% of lumen maintenance after 50'000 hours.

Re:Why not fluorescents?

[SlashWombat]

Your car headlights are supposed to illuminate the curb. There is also the matter of parallax as you get closer to the light/Bicycle, you can determine the distance.

If you have trouble with this, I suggest you either get your eyes tested, or stop smoking whatever it is.

Alternatively, try riding a bicycle (even during the day) for an hour or so a day, for a couple of weeks. You will discover that it doesn't matter how visible you are, ignorant arseholes in cars will actively try to run you down anyway.

Re:Prizes probably help little

[jank1887]

Actually, it's not the X-Prize driving these. It was the DARPA grand challenge (the one with the autonomous vehicles). Off the top of my head, I don't recall which pre-dates which, but the success (and notoriety) of the DARPA prize program has led the powers at be to give authority for additional prize programs in other areas. DoD is currently sponsoring a Wearable Power prize program, and I'm not at all surprised to see DOE get in on the deal as well.

Re:Err , LEDs?

[Smidge204]

Spectrum is one legit problem with LEDs. In general it's difficult to get full spectrum lighting from LEDs - but it IS possible. The problem right now is although LEDs have the best Lumen/Watt efficiency, they have the worst Lumen/Dollar ratio.

Regarding CFLs... I was at the hardware store getting stuff to fix a lamp and decided to put down $5 for a pair of 23W CFLs (7000 lumen/100W equivalent). I have to say that, having owned one of the very early CFL types several years ago and being very disappointed with it, I was VERY surprised at these new ones. Instant-on brightness was equal to the 100W incandescent it replaced, and it actually got BRIGHTER after a minute or so. The light has a slight tint to it - not quite as "yellow" as sunlight but not white/blueish like the 4' tubes in most offices.

All I can say is give it a try. Made a believer out of me.
=Smidge=

Re:Why not fluorescents?

[Smidge204]

This hasn't been true since they stopped using magnetic ballasts.

=Smidge=

Re:Lasers

[hakr89]

The problem with using lasers is that they produce light at only one wavelength. This is part of why they're so efficient, they make only one kind of light. White light is a blend of light from throughout the spectrum. You aren't going to be able to make white light with a single laser, you will need at least three (Red, Green, and Blue). To make more realistic (whiter) light, you would want each laser to transmit more of its area of the spectrum, reducing its efficiency and making it less of a laser as the goal of a laser is to emit coherent light. This is why lasers aren't suitable for this kind of lighting.

See: http://en.wikipedia.org/wiki/Laser

BTW,

[WindBourne]

the HIV cure will not be forthcoming for quite sometime. The reason is that the vast majority of the RD money now goes into drug companies. These folks have NO desire to solve this. For example, look at the history of ulcers. At one time, malox was the #1 drug sold here. Every drug company said that it was genetic. Then 2 regular docs solved it. It was a simple bacterial infection. In fact, it was a common bacteria. But none of the companies who had been paid MILLIONS (back then a lot of money) could solve that (or could they).
Now, we have a NEW genetic disease up and rising: Psoriasis. Evey drug company is screaming that it is genetic. And they have all sorts of new drugs that treat the symptoms. But everybody seems to ignore several little things about:

1. UV lights (i.e. time outside) will clear it up almost better than the drugs do.
2. the rate of rise is MUCH higher than a genetic disease would allow. The fact that it hits ppl who have never had it in either family is ignored.

This is the next ulcer. And while HIV does have a lot more basic research going on, it is obvious that the majority of the research is devoted to solving the symptom issues and not the issue of the bug itself.

Re:Err , LEDs?

[Bryansix]

They exist but they are super rare. My Dad works for So Cal Edison and they ran a program where they sold lamps with dimmable CFL's for like $5 for the whole package. I've only seen them in stores once.

Re:Solid-state?

[bplipschitz]

Wouldn't it be assumed all modern light bulbs are 'solid-state' and will continue to be?

Hollow state [think vacuum tube] actually. Solid state is like an LED.

Mod this down--this isn't insightul, it's just a question.

Re:Lasers

[Anonymous Coward]

The only problem with this is not a technical one, the FDA regulates lasers, and any product that outputs more than 5mw has to have an FDA variance. Diode lasers you see are usually red green and now blue, the green and the blue are usually DPSS where the frequencies are doubled using the lines of a YAG laser KTP crystal and it used to be lithium triborate as a frequency doubler. Thus 1064 nm -> 532nm (green) and 808nm -> 404nm (blue ray), etc. The problem is that red diode lasers are hard to find at over 200mw unless you use a gas laser (krypton/HeNe). Combining Ar/Kr can make a white light laser, but that is not solid state, but makes a hell of a laser show with a PCAOM and a few good scanners.

Re:Lasers

[Ungrounded Lightning]

Right now, diode lasers are among the most efficient (if not the most efficient) light emitters available. I'm guessing the winner, if there is one, will involve a laser or three plus diffusers/despeckling to get general lighting.

"Despeckling" means spreading the frequency of the light. The speckle comes from interference patters from the monochromatic light from the laser bouncing off surface textures. Broadband sources have speckle in each frequency, too. But the speckle from a swath of minutely different frequencies averages out to a non-speckled reflection.

So combining a laser with nanoparticles to efficiently swap the energy around between frequencies until it's smeared out into a plesant white light should do the job. B-)