With all the big-name replacement lamp folks like GE, Osram-Sylvania and Philips all fired up about LED technology being the only solution for the future, it was interesting to see a recent note from the National Lighting Bureau that summarized the May 2013 DOE report entitled The Adoption of Light-Emitting Diodes in Common Lighting Applications (available on the DOE website, or you can reach it through NLB's site here).

Nine application areas were investigated, broken into three groups:

• Indoor lamps (A-type, directional, MR16, and decorative);
• Indoor luminaires (downlight, troffer and other common fluorescent fixtures, and high-bay); and
• Outdoor luminaires (streetlight and parking-lot/garage).

Some of the NLB summarized 'major findings' in the report include:

• In 2012, about 49 million installations relied on LED technology, with LED A-type lamps leading the way, being used in just over four of every ten LED installations. Nonetheless, LED A-type lamps account for less than 1% of installed A-type lamps. That's not the case when it comes to MR16 lamps, however; about one of every ten installed uses LED technology.
• The LEDs installed in 2012 generated source-energy savings of some 71 trillion British thermal units (tBtu), which translates into energy-cost savings of about $675 million.
• Annual source energy savings could approach almost 3.9 quadrillion Btu (quads), if all nine applications used LEDs exclusively.

We want it, we need it, it's good for everybody. Got it. Interestingly, in the same note, the NLB also pointed to their take on the status of adoption in certain areas, especially with regard to the troffer and other fluorescent (tube) fixture space. NLB is an industry organization, but it seems to be made up of a good cross section of members/sponsors that have their fingers deep into the LED pie. The current Chair of NLB, John Lewis of Visioneering Corp, who also serves as the IES representative to the NLB, offered this take:

"We're still a long way from 100% reliance on LEDs and, frankly, I doubt we'll ever see anything even close to that given the technology under development and the significant progress still being made within the traditional lighting technologies, resulting in ever more efficiency and longer life. I don't know how many people are aware of it, but two of the Bureau's sponsors - GE and OSRAM SYLVANIA - both produce fluorescent lamps with rated lives as high as 84,000 hours, which is actually longer than the rated life of many LEDs. Although the traditional technologies cannot do what LED technology can - the ability to change color, for example - the fact is that some of these features have little value in many of the applications where LEDs and traditional technologies compete head-to-head, on a bottom-line basis. But, unquestionably, LEDs are gaining ground in the bottom-line department, as efficiencies improve and prices decline."

84,000 hours! That's the LED zone, man. Back off... "Yeah," I thought, "but what about with more start up cycles and what about the lumen maintenance? Fluorescents don't always do so well at that." So off to the data sheets (GE Ultra Energy Saving Ecolux 100 lm/W T8 here, Osram Octron Ecologic 3 T8 here). Holy jiminy... The GE product lifetime only decreases from 84,000 hours to 80,000 hours when you change from 12 hours/start-cycle to 3 hours/start-cycle. That's not helpful. Neither is the 90-95% lumen maintenance. The only bright spots, if you could call them that, are that the lamps still contain mercury (ah-ha!) and the 80,000 hours is the point on the survival curve where 50% of the lamps can be expected to be dead. Pretty much all are alive at 40,000 hours, which is the kick off point for the accelerating death trend. The measure of the mercury is an interesting one to those not living by the LEED credit handbook, where it is 'credited' with 22 picograms Hg per mean lumen hour. Makes sense as a lifetime measure of the mercury that needs to be dealt with, but when it comes to disposal time, it's about the grams on that day that are filling up the recycling machine (or entering the waste stream for the bad lamp tossers), and zero is better than "some". I know, I know, we're grasping at straws here, but we need some hope.

So the incumbents haven't tossed in the towel, nor have the heavily entrenched LED-loving manufacturers that are still selling about 99.9% of the fluorescent lamp numbers they did prior to the emergence of LED lighting. Yes, it's that low a penetration rate, withe the DOE estimating LED penetration in 2012 at less than 0.1%. Still a ways to go.

So where is the incentive to swap that troffer out for an LED version, or to retrofit an LED replacement lamp into that 4-foot tube socket? Fortunately, there are still T12's out there, and LEDs can cite their key advantages a little more loudly in that space, especially as we have seen LED troffer prices come down below the $200 mark, which puts it right in the space of architectural fluorescents that can do "something" other than just on/off (aka dimming). When it comes to the high-efficiency T8's LEDs are also able to bring a little efficacy advantage to the party. If I did the math right, at the same lumen output over 80,000 hours, an LED troffer at 120 lm/W will save its owner about $51 in energy compared to the 25W 100 lm/W fluorescent (17% savings @$.15/KWh). That one's a little more dicey, as it pretty much requires a re-planning of the space and relocation of the luminaires to get the foot-candle parity unless the LED luminaire just happens to be exactly the same total output as the fluorescent fixture.

More importantly, of course, is that while the incumbents seem to still be able to make progress, LEDs should make more progress faster. Semiconductor technologies always seem to follow a steady path towards more and cheaper, and LEDs haven't defied this so far. They use tiny amounts of phosphor compared to the tubes, so a few more rare earth price spikes will hurt the incumbents a lot more than the LEDs. They are also able to produce a more natural, and controllable, light. We have the advantage, but need to recognize that incremental progress and time will be significant factors. And there is plenty of pie available with close to 1 billion troffers installed in the US that we'll be able to get eventually... sooner if we're creative.