Text-Alternative Version: Solid-State Lighting Early Lessons Learned Webinar

Linda Sandahl: Welcome, ladies and gentlemen. I'm Linda Sandahl with the Pacific Northwest National Laboratory, and I'd like to welcome you to today's webcast, Solid-State Lighting: Early Lessons Learned on the Way to Market, brought to you by the US Department of Energy's Solid-State Lighting Program. I'm happy to introduce our speaker today, Kelly Gordon of Pacific Northwest National Laboratory. Kelly has been with PNNL since 2000, engaged in energy efficiency and emerging technologies issues with an emphasis on lighting technologies. She currently serves as deputy program manager for DOE's SSL Market Development Support Program and is also the task lead for the L Prize competition. So Kelly, please begin.

Kelly Gordon: OK, thank you very much, Linda, and thanks to everyone who has dialed in and connected in today. We really appreciate you taking an hour out to join us to discuss early lessons learned in the solid-state lighting market. So what we're going to do in the next hour is basically I will go through the presentation, providing a brief background on why we did this study that has recently been published and how we did it, and summarizing the 12 lessons. Now, any of these lessons, I'm sure you'll see, could be a presentation unto themselves, so we're going to hit the high points on these. For the full story, I would urge you to go and download the report and check through, especially those that are of particular interest to you. I will try to mention the implications of these lessons going forward, as well, as I go through them. And then, as Linda said, please submit your questions as they occur to you, and we will try to answer as many as time allows.

Now I'd like to start by kind of zooming up to the big picture here. Why does DOE pay attention to solid-state lighting? Why do we invest public dollars in this area? It's really to try and achieve the potential energy savings of this new lighting technology, and the projections indicate that this could be a significant energy resource. We're talking about 46% savings over the 2010 baseline projection by 2030. That amount of energy puts it on par with wind energy, with PV, solar energy as an energy resource. This is the energy saved that can be used elsewhere in the economy while still getting good lighting service.

So how do we get to that goal? There's a couple of factors at play here. One certainly is the technology performance itself, of course. The LED packages – how efficacious are they, how much light are they putting out for every watt of electricity that goes in? And as you know, this has been increasing very steadily, very rapidly over the past decade, and we expect that to continue. DOE has pretty high projections and goals for the efficacy of LED packages, and this is published and updated every year in the Multi-Year Program Plan.

And right along with that is the pricing – pricing needs to continue to come down to a level that will make it very competitive with any other light source. We do see the prices trending in the right direction. This chart is from a report that we published at the end of 2013, looking at both pricing and efficacy trends for various categories of LED products. This chart addresses pricing specifically and kind of boils it down to LED lamps, replacement lamps, and LED luminaires. And we see that in both cases, the prices will be going down. All indications are that prices will continue to go down. We estimated out to 2017, so for LED lamps going down more than 50% by 2017, compared to 2013. For luminaires a little less, but about 30% in that same time frame. So that's certainly another part of the story here. You've got to get performance up, you've got to get prices down.

But there's more to it, as well. And I think we learned that from the CFL experience. I'm going to be a little melodramatic here and put this picture up. And certainly the CFL experience was a little more slow motion than what's indicated here. But over many years, we saw that that technology had some bumps, had some problems as it came into the market. There were performance issues, there were issues with expectations not met, and so we really want to learn from that. And in fact, in 2006, DOE took a pretty careful look at what had been learned from that experience of introducing compact fluorescent lamps. We talked to manufacturers, to energy efficiency programs, to retailers, others who had been involved in the market and drew out some lessons from that.

For the key lesson, that we really kept in mind were these – coordination and collaboration is key. So, the different players in the market that have a stake in the success of an energy efficient technology. They really should coordinate and collaborate. That's very important. Establish standards in product testing – this is very important for being able to evaluate the performance of a product compared to the claim for that product. Will it really last as long as they say it will? Does it really put out the light that they say it does? In order to do that, you've got to have some standardized way of looking at it – that just underpins everything and the trust in the market. Introduce new lighting technology first in applications where benefits are clearly established – now that seems very common sense, but sometimes that's not the case. It needs to be somewhere where it meets the lighting need, and where it makes sense economically. And then respond to the market and resolve problems and issues quickly. If there are performance issues or disappointments in products, that needs to be addressed fast, or you very quickly lose that credibility.

So keeping that all in mind, many efforts started, both by DOE and by other organizations, right around 2007, as LED products were starting to come into the market. So if we then fast forward up to 2013, DOE thought this would be a good time to pause and think about "well, what have we learned so far?" There are now, when you look at what happened between 2007 and 2013, there are now LED products available for almost every kind of general illumination application. There's still a few where they don't quite have the output, but those numbers are really going up – the output that you can get out of LED luminaires and replacement lamps. And so we see just an explosion of new products and applications. So we wanted to look at what have we learned so far, the lessons that we compiled are really from the perspective of the DOE program – from the CALIPER testing that we do, the Lighting Facts program, GATEWAY demos, et cetera.

And what all of those activities provide to DOE in the program is extensive interaction with market players – so manufacturers, utilities, and energy efficiency programs, lighting designers, and specifiers, retailers, building owners, municipalities – it really gives us an opportunity to have constant interaction and feedback from the market. And so that's the perspective that we took. It wasn't a matter of doing interviews and looking back 15 or so years. It was really we're in the thick of it now, and we're getting that feedback constantly.

So this is the cover of the report. This is what you'll see if you go out there to download it. And the approach we took to the lessons was to have kind of a brief template for each lesson that we wanted to draw attention to. So it starts with a lesson statement and an introduction – what is it, the significance – why does it matter, why is this important, background – what do you need to know to put this in context, challenges – what ongoing challenges does this lesson present? And importantly, we wanted to keep it to things that are still questions or challenges or obstacles facing the further development of this market. So there's probably more lessons that we could have listed – things that have been addressed and yet we learned something from that, but we decided to keep it to those that are ongoing issues and for which there are implications for the future. So we discussed that in the report, as well.

OK, let's get into the lessons now. We have 12 of them go through, and as I said, I'm going to try to hit the high points and we can address the rest in the questions. So the first one has to do with testing costs. Rigorous testing requirements adopted in the early days of the SSL industry were necessary to counter exaggerated claims of performance by some manufacturers, but they eventually led to unreasonably high testing costs. This is certainly something we have heard from the manufacturers loud and clear. Testing costs have really accelerated, and there are actions being taken to help to streamline that, help to ameliorate that issue. But I would ask you to cast your mind back to 2007 and to some of the products – you know, products seemed to be coming out of the woodwork at that time. All different levels of quality and claims and lots of different things were claimed at that time, and it was really before we had the test procedures in place to properly evaluate them.

So that early action to say hey, look we need to have evidence, we need to have some common basis on which these can be compared, was critically important early in the market. And the feedback that we've gotten from other regions of the world, for example, in Europe, has indicated that they really wish that they had had that back then, because they're really still dealing with a lot of quality problems in the LED market. And so this is maybe a painful process to go through, but I think it's going to have long-term benefits. We do see many of the qualification programs now accepting family grouping of products, for example, so that it limits the number of tests that have to be performed on variations of products. I think that helps. There's more that can be done in sharing information and streamlining the qualification processes.

The next lesson has to do with lifetime. Despite the promise of long life, there's no standard way to rate the lifetime and reliability of LED products. And we do have – obviously LM-80 is a very important test procedure that was developed and first published in 2008 that has to do with lumen maintenance of the LED packages themselves. And then TM-21 is a method for extrapolating the LM-80 test results. And that's certainly an important part of this story, but it really is not the whole story. Lumen maintenance does not tell you how long a given LED lamp or luminaire will actually last.

So one of the groups that DOE convenes regularly is the LED Systems Reliability Consortium. Many manufacturers participate in this group specifically to work on these issues of reliability and lifetime. And so we did a kind of informal poll with them last year on what are the most common failure modes really when you look at lamps and luminaires. And by far, it really is in the power supply and the driver component that is more likely to be the failure point, rather than the LED and rather than long-term lumen maintenance. There may be other issues – there are LED failures related to shorts in connections and problems with the board, there can be moisture ingress – there could be various issues, but the lumen depreciation, normal lumen depreciation, is typically not the failure mode for these products, is what we're finding. So there's still much more to be done on characterizing and estimating lifetime and reliability of LED products.

Lesson three has to do with product families. Specifiers prefer complete families of products, but the rapid evolution of LED technology presents a challenge to manufacturers in creating and maintaining complete product lines. This is an example – this is a little snapshot from the Osram Sylvania lamp catalog. This happens to be halogen PAR30s, but you see you can get these in lots of different flavors. They all will look the same, and being a PAR30 lamp, but you can get various light outputs, you can get various beam angles, and that gives you different center beam candlepowers, different intensities, depending on where you need to use this lamp.

And there are many more – this is just three of them from their collection. So designers are accustomed to being able to really fine tune what they need for an application and be able to grab products from the same line so that they have a consistent look, consistent color, et cetera, and in various output and intensities. Now that can be more challenging with LED products where the LEDs are changing so rapidly, and other materials and costs, and other factors in the market that can make it more difficult for the LED product lines to stay as complete and as consistent with one another. I think that's something that's definitely going to improve over time, but it's just something to keep in mind that you may not have as full a product line as you are used to.

Lesson four has to do with color quality. The range of color quality available with LED-based products and the limitations of existing color metrics may confuse users. So this picture is from a recent study that we did on LED PAR38 lamps. These lamps are all nominally 3000 K that you can see even from the photograph – and I know photographs often depict colors differently than we see them, but it's really the difference that you see across them. There can be noticeable differences across lamps that nominally have the same color temperature and there can be variations in other areas, as well. With LED, we have to become more educated on the color issues.

And sometimes it seems like there's just too much information that we have to understand and process as we try to make decisions. There's correlated color temperature and DuV and color rendering and the spectral power distribution. There's so many different terms. What do they all mean? Do I need to understand all of them? Even for the basic consumer going into Home Depot or Lowe's or another big box, what does soft white mean, what does bright light mean? Now that sounds good, I think I want bright light. Daylight, I love daylight. Maybe that's what I want in my home. But what does it really mean? And unfortunately, those terms are not standardized, so one manufacturer's bright white may be different from another manufacturer's. And that's something that we really would encourage that there be some standardization of that, or harmonization of those terms to make it a little bit easier for the consumer, and even for more sophisticated users.

Lesson five continues on the color theme – color stability. The color delivered by some LEDs shifts over time, enough to negatively impact adoption in some applications. Now this is, of course, not everything. Some places you wouldn't notice, but there are other applications that are more sensitive where it can become an issue. The data shown in this graph is actually from some older CALiPER data color testing that we did over 6,000 hours and over 12,000 hours. The main point being here that we did see that some LED products shifted very noticeably. It would be more than seven step MacAdam ellipse, which would be certainly visually noticeable in many applications. Others had very little color shift over that period of time. But for applications such as in a museum or on a white wall, a wall washing application, or even in some retail applications, that can be significant.

So this is an actual case from the Smithsonian American Art Museum. And again, don't pay too much attention to the color as it appears on the screen, the camera sees things differently than our eye, but just to the relative change here. The new lamp on the left was one color, and then in less than 6,000 hours, it had shifted noticeably, which for this particular application, that was too much of a shift. And so that is the limiting factor in the lifetime of that particular lamp. So we have seen that it can happen. I think people are paying attention to it. And again, this is an aspect that will be improving.

Lesson six – flicker. Some LEDs flicker noticeably, which may negatively impact adoption in some applications. And you'll notice that we do phrase the lessons in that way – we're talking about some applications here, not necessarily all, but it has to be something that designers and specifiers and manufacturers are all aware of, that this can be a problem. So we measured the flicker waveform on many different LED products, and it's really all over the map. You can see just three examples there on the bottom there, where the flicker looks very different across those different products. And it really has to dol with how it is designed. It's something that we encourage people to look at when they're considering products. You can use a flicker wheel of various types to help you detect it if you are not somebody who naturally sees it, and it varies greatly across people, how sensitive they are to visible flicker.

But the point is that LEDs can flicker more than other sources. So this provides you kind of a frame of reference. You have flicker index on the vertical axis and percent flicker on the horizontal. And then the area down here in the lower left corner is incandescent and fluorescent. These red boxes are magnetically ballasted fluorescent, which is a technology that kind of had a reputation for flickering, for being noticeable. You could see it out of the corner of your eye if you were in an office with magnetically ballasted lamps. So that kind of provides you a point of comparison. Where does LED fall? Well, LED can be much worse. Some products that we've measured have been much, much worse than that. Some of them have been much better, and well within that frame of reference that we're used to from standard technologies. Again, with so many things with LEDs there can be great variability in the performance, depending on how well it's designed and what trade-offs have been undertaken in the design of the product. There are efforts to better characterize flicker, to provide more of an industry standard approach to flicker, but that is an ongoing process.

Lesson seven – glare. LEDs can cause a glare which may negatively impact adoption in some applications. Any light source can cause glare, right? The issue becomes with LEDs it's a very small point source, and as it gets more efficacious, we're packing more and more light into a very tiny source, and so if it is not shielded properly or reflected properly, it can be a glare source. And glare can be anywhere – just like flicker, it can be anywhere from mildly annoying to, actually, kind of dangerous. If you're there and the light is shining in your eye and you can't see that people are there in front of your car, that can actually be dangerous. So it's really something that should be evaluated in luminaires, very much take into context where observers are going to be, where people in the space are going to be, relative to the luminaire. There are some metrics for glare that are used and probably others needed. The main thing I think is to be aware that it's a consideration, and you should check it out.

Lesson eight has to do with dimming. Achieving high-quality dimming performance with LED lamps is difficult, but improving. Again, we always hear that LEDs are inherently dimmable. Yes, it's a digital technology, certainly they are dimmable with properly designed dimmers. The problems come in in trying to put this new technology onto the existing stock of dimmers that are out there in the market, all of which were designed for incandescent or maybe for fluorescent. So depending on the characteristics of the LED sources and drivers, the characteristics of the dimmer, and the number and type of light sources that you're putting on a circuit, you may encounter a lot of different behaviors from the light. You may get a limited dimming range, a very unpredictable dimming curve – I'm not going to read through all that list, but I think many of you know the issues there. LEDs can dim very well, but it really needs to be matched with an appropriate dimmer. And as we go forward here, there will be more and more good dimmers and dimming systems that are specifically designed for LEDs. So over time, this will turn over. But we have to be realistic about some of the challenges in putting this technology on the existing infrastructure.

Lesson nine – interoperability. Greater interoperability of lighting control components and more sensible specifications for lighting control systems are required to maximize the energy savings delivered by LED-based sources. I think this is one of the most exciting areas when we talk about LED lighting and the service improvement we can get and the energy savings that we can realize from the technology. It really comes down to how do we control it? You can adjust the light levels, you can adjust the color appearance of LED light sources, you can say when it should be on and when it should be off. When we can really fine tune our use of lighting, that is going to contribute so much to comfort and performance of lighting, as well as energy efficiency. But it's not easy. We need more common protocols to be able to make all of the different elements of these systems talk to each other. One example is the ZigBee protocol, which allows for the development of apps – for example, that can control your lights in your home. Lots of different ways to do that and different schemes that could come up, but it's using a common language. And that really helps to allow for innovation, but around a common theme that we can work with on product development.

Lesson ten – serviceability. Lack of LED product serviceability and interchangeability has created more market adoption barriers in certain sectors. The photos that you see here are from the judging of the Next Generation Luminaires Design Competition, which DOE organizes along with the IES and the IALD. And these judges that we have for this contest are mostly lighting designers and specifiers. They like to get very hands on with the luminaires. They want to see how would this thing be serviced? If something goes wrong, how do we get into it? Are there parts that can be interchanged? Is there a light engine that could be changed out if there's a better one in the future?

These are difficult issues, because at this stage of market development and product development for LEDs, everybody is trying to find their niche in this market, and there are a lot of proprietary solutions out there. So you can't just slap in one LED modular light source for another and expect it to work. But there are efforts to help provide some common definition interfaces – the Zhaga effort is one of those. So this is an example of the mechanical interface for one of the Zhaga standards. They also provide for electrical and thermal interfaces, and this may be part of the solution. So that you could have a luminaire that has a light engine that can be, then, taken out and replaced with another down the line. Lots of issues to be worked out with serviceability and interchangeability. I think people are thinking that way, and we'll see progress in the coming years.

Lesson 11 – existing infrastructure. Existing lighting infrastructure limits the full potential of SSL. More effort is needed to open the doors to new lighting systems and form factors. So in all of these product categories that I'm showing, there are LED products now that directly replace incandescent, halogen, HID, fluorescent. And that's fine – we've got millions of light fixtures out there that could benefit, potentially, from a replacement with an LED product. It could save energy, it can improve or match lighting service, but there's also trade-offs with that, because you're trying to fit into a form factor and into a shape and into a box that may limit the full potential of solid-state lighting technology. And it can stymie the innovation of really exploiting those potentials.

So we need to think about what else can it do, and I'm going to draw on a rendering that Fred Maxik from Lighting Science Group shared recently at one of our workshops. So for example with roadway lighting, does it need to go up on a high pole and be shining down on the roadway? With LED, could we potentially see something where it would be more embedded in the roadway itself, still very well illuminating the surface of the road. Another example from GE Lighting, this is a product that's on the market now, simply a different form factor that is made possible by solid-state lighting that can provide better light distribution for some applications. So we think we'll see a lot more of this kind of innovation. It takes a long time for infrastructure to turn over, as you know. We're talking about millions of existing buildings, the whole electricity infrastructure. There's always going to be a role for replacements, but we also want to see continued innovation and thinking outside the boxes.

OK, we made it to lesson 12 – qualification programs. Programs that provide ways to identify quality LED products have helped support market adoption. So the DesignLights Consortium has grown phenomenally in the past few years. That has to do with commercial luminaires, and provides for some performance requirements and an extensive screening process, with the ENERGY STAR program that covers some replacement lamps in residential style fixtures, the Lighting Facts program that DOE runs that is really kind of a truth in advertising type program, and the CALiPER testing program.

All of these types of programs really help consumers, specifiers, others to get good objective information about LED products and be able to compare and evaluate their performance. The Lighting Facts program is now up to over 13,000 listed products – and this one is not a qualification program, it's a listing, so it doesn't have any specific performance requirements. It simply asks for test reports so that the claimed performance can be verified through independent testing, and there's verification testing, as well. And so this is a useful tool for people to use in searching for products and filtering products by performance, and it can also tell you is it DLC listed, is it ENERGY STAR, and other attributes. And we hope to use that as a way to streamline that process, looping back to lesson one about minimizing the testing qualification and costs.

So what has happened so far with LED? Now kind of comparing it to six years out since market introduction, where do we come, and how does that compare to CFLs? Now CFLs after they were on the market for six years had just barely scratched any market penetration – 0.1%. And granted, this was a very different technology – CFLs were only trying to target incandescent lamps, basically A lamps, and it was a much more limited market. LED, it turns out, is going to be much more than that. It's really a replacement for any light source, or most light sources. And after six years, LED had already reached a market share of over 4% – that's luminaires and lamps, not just replacement lamps.

And how does that compare to another technology that's taken off very fast – the smartphone, after six years, was higher than where LEDs have been. It seems like they're everywhere now, so it's hard to believe that it was only at that level, but we definitely see the LED market trending up, and there's a lot more to go. Looking at the energy savings potential, DOE estimated that in 2012, the technology had already saved 71 terra BTUs, and that's just barely scratching the surface of the potential. That's just looking at those nine product categories that we studied in that adoption report. So there is a long way to go and much more work to do to address the challenges and the lessons that we've drawn attention to in this report, and we look very much forward to working with all of you in the marketplace to help address those and realize the potential of this technology. I'm going to stop here.

Linda Sandahl: OK, great. Well, we are out of time. And I would just like to thank everybody for participating in today's webcast brought to you by the US Department of Energy, and thank you to Kelly Gordon for making the presentation.