Re: \"White-light LEDs May Make Light Bulbs Obsolet
I agree with the general premise of the article. While LEDs have drawbacks preventing them from capturing the general lighting market at the present time, by 2012 I think all of the major problems will have been solved. Here are the main problems as I see them, and why I think they will be solved:
cost: This is the big problem right now, and even if all the other problems were solved the present cost per lumen of around $0.10 compares very unfavorably with the cost per lumen of 4-foot fluorescent tubes ($0.001), or even CFLs ($0.002 to $0.008). However, the good news is that the raw materials for LEDs are dirt cheap, and once manufacturing processes are streamlined and they are made in larger quantities they might very well be far cheaper than any current light source. Look for the cost per lumen to become competitive with other light sources within a decade at most.
efficiency: the best production white LEDs at around 35 to 40 lm/W are only about twice as efficient as light bulbs. Typical Luxeons seem to run around 25 to 30 lm/W, or only 25 to 50% better than the efficiency of a typical 100 W incandescent lamp (~17 lm/W). This compares very poorly with CFLs at 60 to 70 lm/W, or the best fluorescent tubes at ~100 lm/W. The good news is that next year we will have production LEDs of around 60 lm/W efficiency from Nichia and probably Lumileds, and red LEDs have already surpassed 100 lm/W in the lab. I see no good reason why we won't have production LEDs with efficiencies of 100 to 150 lm/W by 2012 given the current rates of progress, and I feel we'll eventually approach wall-plug efficiencies of 80 to 90%, meaning 200+ lm/W for white light, and efficiencies ranging from maybe 50 to over 500 lm/W for colored LEDs, depending upon the color.
color consistency and color rendering:current white LEDs already have acceptable color rendering and consistency for many lighting applications, although they are not quite there yet for more critical retail and home lighting needs. Certain technologies on the horizon, such as the UV + RGB phosphor LED, promise to address these problems within a decade. Indeed, I see no good reason why we won't have LEDs by 2012 which have CRIs approaching 100, and are available in any of the more commonly used color temps. Furthermore, LEDs may finally enable the ultimate light source-namely one within an equal energy distribution of all frequencies across the visible spectrum. Thus far this has not been attainable with any other lighting technology.
As for certain advantages LEDs already possess-we have robustness, ability to make colored light with unsurpassed efficiency, and very long lifetimes. These advantages will certainly improve with time, and while LEDs in their current form may not necessarily be the light of the future, I feel that some sort of quantum-based light source is what will be primarily used in the future. I don't think incandescent bulbs will even be made in ten years time outside of for niche markets (much like tube amplifiers) because alternative sources will be cheaper, last longer, be more efficient, and have superior light emitting properties. Indeed, given the vast superiority of fluorescent and metal halide in terms of operating costs and light quality for nearly all general lighting applications, I fail to see why incandescent bulbs are still made in large quantities any more even in 2004. By 2012 I suspect incandescent bulbs will have faded into the dust bins of history along with vinyl records, magnetic core computer memories, and tube stereos. I think it will take somewhat longer for fluorescent and other discharge lights to disappear given their present efficiencies and the huge installed base in many businesses, but I would say by 2025 or 2030 they will largely be gone as well given the normal 20 to 25 year cycle for upgrading lighting fixtures.