01 Jul 2011
With many manufacturers now able to turn out LED chips with excellent efficacy, improving the $/lumen cost of high-quality white light is now the key priority.
As the LED market goes from strength to strength – near-doubling in value in 2010 – and as semiconductor-based light bulbs increasingly appear on the shelves of homeware stores, it’s tempting to feel that the era of solid-state lighting is well under way.
But while it’s true that the past few years have seen a remarkable improvement in the efficacy of white LED chips, packages and lamps, measured in terms of the number of lumens emitted per watt of electrical power consumed, general illumination applications remain at a very early stage of development.
Most estimates for the general lighting sector put LED penetration at less than 5% currently. And although designers are keen to exploit the possibilities of this new illumination technology, and consumers are understandably attracted to the idea of cutting their energy bills, there is still plenty of development to be done before the LED becomes dominant – even though it is now accepted to be a question of “when”, and not “if”.
According to Mark van den Berg, director of marketing for Europe, the Middle East and Africa at LED manufacturer Philips Lumileds, the key questions now surround the quality of LED lighting and systems, and their cost. “The [chip] efficiency is now good enough,” he told delegates at the euroLED 2011 event held in Coventry, UK, earlier this month. “[But] we have to deliver more lumens per dollar to drive market adoption.”
Van den Berg also quantified the reduction of LED package costs required for widespread adoption, quoting the target of a 20-fold improvement that has recently been established by the US Department of Energy’s (DOE) latest multi-year program plan (MYPP) for solid-state lighting development. It is targeting an LED package cost of just $1 per kilolumen by 2020, with luminaires at only $5 per kilolumen (quoted at around $50 in the DOE report, these have declined to $40 or less in recent months).
As of 2010, the state-of-the-art LED package price ranged from $10 per kilolumen upwards, but because of the combination of blue light and conversion phosphors used to generate the best efficacies, that figure only applies to the cool-white light that both designers and consumers dislike.
The DOE’s 2010 estimate for warm-white LED packages is between $20 and $25 per kilolumen – although recent improvements have nudged that figure down to $18. But even so, the near 20-fold improvement cited by van den Berg holds true, and the DOE has set an interim target of $7.50 per kilolumen by 2015.
Although general lighting applications certainly gained strong market traction in high-end applications in 2010, it was TV backlights that drove the bulk of the 93% market growth measured by analysts at Strategies Unlimited. And while the tremendous volumes being produced by the likes of Samsung for TV applications is impressive, this and other maturing markets, such as automotive lighting, differ from general illumination in some crucial respects, said van den Berg.
In a mass-produced TV or a car, the need is for a single type of LED that fits the requirements of the application and can be produced in huge volumes. In other words, a commodity item with no real subtlety – millions of TVs or cars can use pretty much the same device. But in general lighting, the requirements for different types of applications – whether task lights, residential, office or streetlights – can vary wildly, making the LED source much less of a commodity item.
The precise nature of the application also radically impacts the business case for installing LED lighting – van den Berg cites the Alpine tunnels of Italy as an obvious candidate for an LED solution, because the lights need to be on constantly and so the payback period for even an expensive retrofit is rapid. In contrast, he says that office lighting has proved to be a much tougher market so far, partly because the incumbent T-5 fluorescent light is such an effective – and cheap – product.
Requirements in the retail sector are again different, with the emphasis on light and color quality, getting the right CRI for each application, and where the primary competition is with halogen lamps rather than fluorescents. Retailers also want to see their products “sparkle” under lights – and can be highly skeptical of the ability of LEDs to do this. And to retrofit domestic lightbulbs, without incentives van den Berg sees $10 for a replacement bulb as the target required for mass adoption – compared with the typical $40 price today.
For outdoor lighting such as streetlamps, adoption is beginning to take off around the world. But van den Berg still thinks that the payback period is usually too long, and needs to come down to less than three years through efficacy and cost improvements to really spark off widespread deployment. That means reaching a system efficacy of more than 120 lm/W with the kind of warm white light that is preferable in city center locations, alongside improved reliability and color stability.
So it is evident that the “general lighting market” is in reality a combination of many different markets, each requiring an emphasis on different qualities. Cost, of course, is common to all, but some applications are much more ready for the LED revolution than others.
One company enjoying a measure of success in the retail sector is Bridgelux. The California-based chip maker supplied the sports/fashion chain Intersport with LEDs to light a store in Austria for a December 2010 refit. According to Tom van den Bussche, Bridgelux director of marketing for Europe, the versatility of the new lighting technology means that it can be used to guide shoppers through a store more effectively – a classic case of “value-add”. Intersport appears convinced, with van den Bussche saying that five stores are now being retrofitted in similar fashion, with the potential to continue with hundreds more across Europe.
Other tangible advantages in the retail sector result from the absence of UV wavelengths in LED illumination. UV light from conventional halogen or fluorescent lamps make textiles fade with time – and when items of clothing are placed under strong lighting in a fashion store they can degrade quickly. Similarly, illuminated food will stay fresher for longer without being subject to the heat produced by conventional lamps.
Cree’s VP of field sales Bob Pollock agrees with van den Berg regarding the difference between a TV-quality LED and something suitable for installing in the home. For that, he says, “we need a different class of LED”. Pollock estimates that, in 2010, the cost of a cool-white LED package had dropped to only 10% of the 2005 price – with a further 33% drop expected this year. But that is only for a lamp with a color-correlated temperature of 6000 K.
For the kind of warm-white lamp that consumers might actually want to buy, Cree’s original CR6 downlight – considered a replacement for a traditional 60 W bulb and featuring a combination of some 42 red and white LEDs to deliver the same kind of favorable “hue” – consumed only 12 W to produce 650 lumens. But it cost even wholesalers more than $100. That was in 2007.
Illustrating the rapid pace of change in the industry, Cree’s CR6 replacement, released in 2010, used only 8 LED chips to deliver the same output while consuming just 10.5 W. But even though that lamp still costs in the region of $50, it can be bought much more cheaply – for example in certain US States, utility companies subsidize the cost to just $20, close to the “tipping point” of $10 suggested by van den Berg.
Where Pollock sees a major market for Cree’s LED-based lamps is in the replacement of the hugely popular MR-16 halogens (Pollock suggests a global market of more than a billion units per year, mostly 50 W lamps with an efficacy of only 15 lm/W and a short lifetime of only 800 hours). But until Cree launched its MT-G product earlier this year, there was no LED-based replacement that could approach the light quality of the MR-16 – and Pollock expects this product to be a winner with consumers.
The Samsung threat
There is a strong subtext to these messages from the traditional top-tier players. They know that that Samsung is coming after them. And after Samsung will come the emerging Chinese contenders as suppliers in the region consolidate.
Investor sentiment certainly seems jittery, with Cree’s stock price at a 52-week low, warning signals over Philips’ lighting business, and questions arising over the timing of Osram’s initial public offering.
Meanwhile Samsung, collaborating with Acuity Brands, has established an operational center in the US, and publicly announced its intentions to go after the US lighting market. Osram reacted with a patent infringement lawsuit; Samsung’s counter-claim the inevitable response.
While that and other battles rumble on, the focus for all is going to be on lowering costs while improving warm-white quality. Samsung will have the edge on cost for now, and the others on quality. But, given the volumes that Samsung produces, for how long? The solid-state lighting revolution isn’t just coming – it’s already begun. But there are plenty of battles yet to be won - in the laboratory, in manufacturing, and in the courtroom.
About the Author
Mike Hatcher is the Editor in Chief of optics.org