U.S. Department of Energy - Energy Efficiency and Renewable Energy

Building Technologies Office – Information Resources

Text-Alternative Version: CALiPER: Troffers, Kits, and Tubes Webcast

Below is the text-alternative version of the "CALiPER: Troffers, Kits, and Tubes" webcast, held June 20, 2013.

Linda Sandahl: So we're very happy to have as our speaker today, Naomi Miller of Pacific Northwest National Laboratory. Naomi is an engineer in the Solid-State Lighting Program here at PNNL and she focuses on lighting quality, energy efficiency, sustainability, and the acceleration of LED technology deployment. Previously, Naomi was principal of Naomi Miller Lighting Design in Troy, New York. She's a Fellow with the Illuminating Engineering Society of North America and the International Association of Lighting Designers. She holds a Bachelors of Arts and Design degree from the Massachusetts Institute of Technology and a Master's Degree in Lighting from Rensselaer Polytechnic Institute and for those of you who know Naomi, she has great taste in hats. Okay, Naomi? Go ahead and begin.

Naomi Miller: Thank you, Linda. I wasn't expecting a comment about hats. My goodness. I'm going to be talking about Troffers, Kits, and Tubes and first I need some response from Terry Shoemaker to let me know that everyone's seeing just my left-hand screen. Is that correct?

All right. Great. Okay. Onward I go. We at Pacific Northwest National Laboratories have had directive from the Department of Energy to figure out how to make LEDs more usable in architectural applications so that they can be adopted more quickly and save energy sooner. So in that interest, we took a look at one of the most boring light fixtures on the planet and that is the troffer.

All right. The most boring fixture on the planet is the recessed troffer. You're seeing here one of the most common troffers. It's a 2 x 4 parabolic troffer with an 18 cell louver, very common. This is another kind of troffer that you see often in classrooms or in office spaces. Here's a collection of six troffers of different flavors, starting with the lens troffer at the top, and then proceeding through to a parabolic into a volumetric, or high-efficiency luminaire; some planar and non-planar diffuser luminaires; some that are solid diffusers and some that have perforated metal baskets. In this study, we wanted to take a look at all of those or as many of those types of fixtures as we possibly could.

There are reasons to switch to LEDs. LEDs have lots of advantages and certainly, LEDs may make sense in new projects, but one question that always comes up is "does it make sense at this point to look at LEDs for remodel projects in existing spaces when you are, perhaps, gutting a space and putting in fixtures, reusing the existing troffers, but modifying them?" Or does it make sense to refresh those troffers so that they have a new appearance to them using LEDs? But there are all kinds of questions that come up and appearance is one of them, glare is another, cost is another, reflected glare on the computer screen, color quality, etc. I won't go through all of these right now, but you all understand what the issues are.

So in this interest, we procured 24 pairs of recessed luminaires, 2x2s and 2x4s. As many of them as possible were equipped with 0-10V volt dimming controls, and then we invited 18 lighting designers, lighting engineers, and facility managers to come in, observe these fixtures, score them, and tell us what they thought of them.

Now, we had 24 pairs of fixtures of basically four different flavors. The plan is in the upper right-hand corner there on your screen. That is just a floor plan with 2x2 acoustical tiles. The yellow 2x2s and 2x4s are fluorescent benchmark luminaires the light blue ones are T8 tube replacements using LEDs. The pinkish ones are retrofit kits using LEDs and then the purple ones are dedicated LED troffers. Those are troffers that were never born as fluorescent; they were born as LED troffers.

Now, in the interest of turning this into an Olympic competition, we had some qualifying heats. We had some criteria for selecting these fixtures. They, of course, had to go into an acoustical ceiling mounting. We targeted 4,000K although we know that most of fixtures in the United States are 3,500K. It turned out that when we looked at the options available in LED, there were quite a few more products available in 4,000K than 3,500, so we opted to shoot for 4,000K. we tried to get CRIs above 80 and R9s above zero and you'll see later how well we did on that. We targeted between 2,500 and 6,000 fixture lumens, since that covers the range of what's out there pretty well. We ordered them with dimming drivers wherever they were available. In terms of power quality, we were looking for power factor greater than 0.9, THD less than 20 percent, and all the fixtures that we ended up getting in LED met the power factor requirement. In THD, all but two T8 tube replacement products met that requirement. Two of them had THDs of 37.4 percent and another had 46 percent THD.

We really wanted to test a range of optics and perceived quality levels. So we had a range of products that were high-price, low-price, mid-price. We wanted to see what was out there. And then one of the most important criteria was they needed to be received in ten weeks from the order point. They were ordered about a year ago and they had ten weeks to get to us, so the final date in which we could get these tubes or kits or troffers was the end of July 2012.

Here's an action shot of the athletes and the athletes are not the people there. The athletes are the fixtures in the ceiling. And you can see them all shut off in the upper left. If you turn all of those products on, it was like one great big light sauna. We had over 350 foot candles and it just felt marvelous in there. Great light therapy space but of course, that's unrealistic. So we showed these fixtures to the observers, eight at a time; four pairs of fixtures at a time and you can see that illustrated in the bottom right photo.

Here are the judges hard at work. They weren't always craning their necks looking at the fixtures overhead. We asked them to come in in smaller groups and to observe fixtures as though they were using the office space. So walking around, sitting, using computer screens. We had laptop screens there. We had iPad screens. We had printed material of different kinds, and we also had some moveable walls. So what looks like kind of a chalkboard on wheels there is actually a gypboard wall that the observers could move around so that they could look at light patterns from different kinds of light fixtures on wall surfaces.

They were asked to fill out survey forms, couple of pages worth. I won't go through each question in detail. Now here are our Olympic athletes; 24 different fixtures, and you can see there are a bunch of lens troffers. There are a bunch of volumetric fixtures, parabolic fixtures, diffuser fixtures, all kinds of different flavors of fixtures. You probably recognize some of these. So let's look at the results. We're basically going to carve the results into different heats of the decathlon, here.

And then at the end we know that all of the athletes need to go into a room and pee into a little cup to make sure they're playing fair. So we also had an inspector come in, it was actually Carl Bloomfield from Intertek, came in, inspected all the luminaires to make sure that any fixtures that had been retrofitted still met electrical safety requirements.

So let's talk about the first one. Energy efficiency and luminaire efficacy. This is lumens per watt and you can see the fluorescent benchmark fixtures using 28 watt energy saving lamps and 0-10V dimming ballasts, ranged in efficacy from 57 to 62 lumens per watt. The LED tube retrofits came in a little bit lower than that, but they also went higher than that. So they ranged from 55 to 76. The retrofit kits, where you take out the doorframe, usually, retrofit the guts inside, and usually put in a replacement face on the troffer, those kinds of kits came in between 60 and 77 lumens per watt, but the big winner here was the dedicated LED troffers, which came in between 74 and 107 lumens per watt, so they kind of cleaned the clock of our fluorescent benchmark troffers.

Energy use. Now we can't just look at efficacy, which is lumens per watt, because you can have very high efficacy products that actually draw more wattage than the fluorescent benchmarks. Our fluorescent benchmarks came in between 49 and 83 lumens per watt and they were -- the LED equivalents, in general came in lower, but there was one LED tube retrofit product that came in very close to the wattage of the fluorescent benchmark. So what's important to point out here is that there's a tremendous range in energy power use from all of these fixture types whether you're talking about tube retrofits; whether you're talking about retrofit kits; or whether you're talking about dedicated LED products. So you need to be very aware that these products are not all the same. There's a tremendous range in power draw and light output.

Getting to light output. The fixtures ranged between about 2,500 lumens up to about 50, almost 5,500. And we asked the observers "was this about the amount of light that was correct for an office or a classroom installation?" Was it too much light? Was it too little light? Now, they had a pair of each fixtures to look at, and they had light meters to play with, so they were judging based on the light that was produced by a pair of fixtures in a nine-foot ceiling in this mockup office space.

In general, the fixtures that were delivering over 5,000 lumens were considered delivering too much light for that application. The ones that were at the low end, down below 2,500 lumens – those were considered, in general, okay, perhaps a tad too low. The ones that really came in being identified as being acceptable range of light output ended up falling between 2,500 and about 4,500 lumens. So that's the range that we think we should be looking at for most office and classroom spaces. Of course, it'll depend on your application, your ceiling height, the spacing of the fixtures, etc.

Color quality. Kits and tubes generally produce poorer color characteristics than the dedicated LED troffers or the fluorescent troffers.

The fluorescent troffers were very consistent, about 82 CRI with an R9 value and you probably remember that an R9 value is a measure of how well the red colors are rendered in a color palette. So if you have an R9 of above zero, that's generally considered good. If you have an R9 of above 50, that's considered very good. If it's an R9 of about 75 or above, that's considered excellent. So the fluorescent benchmarks came in in the good range. The dedicated LED troffers ranged from poor CRI and poor R9, you can see the minimum value was 68 and a minus 41 R9, all the way up to a CRI of 91 and an R9 of 58. So we've got kind of miserable performance there, up to very, very good performance in the dedicated LED troffers. There was actually only one dedicated LED troffer that really had a poor color rendering and that brought the averages down lower than you might expect.

The LED tube retrofits, some of those also had low color characteristics, but also good color characteristics – not excellent, but good color characteristics. The LED retrofit kits were very consistent. Those actually were very close in performance to the fluorescent benchmarks in terms of color quality.

Dimming performance. Those that had 0-10V dimming available had about the same dimming performance as the fluorescent benchmarks. They dimmed down to about five percent of light output, some of them a little bit lower. Some of them didn't go down to much lower than say 15 or 18 percent, but that was very similar to the performance of the fluorescent. However, one product, one dedicated LED 2x2 product did not come in with a 0-10V dimming driver. It instead came with its own proprietary stepping driver. And it had 256 steps. That steppiness, as it dropped from high light output to low light output and vice versa, was very much disliked and in fact, that steppiness was interpreted as flicker. So we definitely got feedback from the observers that that stepping dimming was not something that was going to be acceptable in most office or classroom spaces.

When we dimmed these products, we found that about a third of the products flickered noticeably when they were dimmed. Now, most of these are using pulse-width modulation dimming techniques on the drivers, so that means that in order to reduce the perceived output, the driver is turning the light on and off, on and off, on and off, many times a second. And when it goes from 100 percent to zero percent light output, even very rapidly, it was noticeable as flicker to the observers through what's called a phantom array effect, which basically means that you see a stroboscopic effect from any moving objects or if your eye is moving relative to the light fixture, that it is noticeable. You kind of see a stroboscopic effect of that light fixture relative to your eye movement. That can be a problem for some observers and some individuals. Some people don't notice it much at all.

Of the fluorescent troffers, none of them flickered when they were dimmed. Now we know that those were operating at high frequency in the neighborhood of 30, 40, 50,000 hertz so that on/off or oscillation in light output happened too fast for the human visual system and brain to pick up the change. Flicker is not a problem with high frequency electronic dimming ballasts, which is what we use these days. It used to be a problem with magnetic ballasts, but fortunately, we don't have many magnetic ballasts, yet. If we do have magnetically ballasted fluorescent products, they should be changed immediately because they weren't very energy efficient, so it's a good idea to change those out anyway.

The LED products at full light output, there was very little perceivable flicker. So the flicker was really only picked up when the drivers were dimmed and it was not all of the drivers. It was only about a third of the LED drivers that produced perceivable flicker and this was in the kits and the dedicated LED troffers. By and large, the LED tubes were not dimmable so we really didn't get feedback on the LED tube retrofit products in terms of flicker.

Here is an illustration of a flicker wheel. On the right-hand side, you can see that the flicker wheel is producing basically smooth gray circles as the wheel spins. On the left, you can see kind of a checkerboard pattern. When that checkerboard pattern shows up, that means that there is probably flicker happening from the light source that's overhead.

Just as an illustration of what flicker can look like on, well, for lack of better word, an oscilloscope. This is what's happening from a fluorescent high-frequency ballasted dimming product. At 100 percent light output, you can see in the middle of the screen here an illustration of a plot on the Y-axis. You're seeing a zero line up to one. That's light output. So zero means that the light output is fully off. One means that it's fully on and this is about 90 milliseconds, or 1/10th of a second over the horizontal axis.

Now you can see when we dim it down to 75 percent output on the dimmer, you're getting straight light output across the time period so you can see that that straight horizontal line means effectively, we're getting DC output or steady output. So you get steady output at full output, steady output at 75 percent of the dimmer, 50 percent, 25 percent of the dimmer, etc. So it's very consistent performance.

Now let's go to an LED troffer. This one is also very steady performance. You can see a little bit of ripple on the top of the light output there. So even if you look at 75 percent of the dimmer, you see a little bit of light output variation – that's not enough to really pick up. So this particular LED fixture and, specifically the driver – because it's the driver that dictates flicker, not the LEDs themselves – is not producing much perceivable flicker.

Here's a product that is producing a small amount of variation at high output, even when it's on a switch, as opposed to maximum output of the dimmer. This is actually very similar to the kind of waveform that we used to see from magnetically ballasted fluorescent and it's about 20 percent flicker, flicker index of .07 at full output on the dimmer. As it dims down, that sinusoidal waveform continues. It just gets smaller and smaller. The percent flicker at 25 percent output is about 16 percent. The flicker index is .05.

Now, let's look at a different waveform. You can see it 100 percent of the dimmer on this particular LED troffer. It's essentially DC output – continuous, unwavering output, but as soon as you put it on the dimmer and dimmed down to 75 percent or less on that dimmer, do you see the on/off, on/off, on/off behavior? This is going on and off about 250 times per second and this was perceivable to most of our observers. You can see even down at 25 percent, there's an on/off behavior there. And here's another one. This one is exhibiting that on/off behavior using pulse width modulation dimming technique. This one is operating at about 480 hertz so it goes on/off 480 times a second. Now this was perceivable to fewer of our observers, but there were still a few people who picked up the flicker from this.

So why do we care? Flicker is a function of frequency, which means how often it goes from full output to a lesser output; the modulation depth, which is basically how much it's modulating from full output. The DC com opponent which is basically how much steady output you're getting over the course of the cycle, and the duty cycle, which is what percentage of the time during the cycle the light is on. We care because flicker can be a trigger for people who get frequent headaches or migraines. It can also increase autistic behaviors, especially among children. It can be distracting and we know from a couple of studies, including a study at the NRC in Canada and a study at the Lighting Research Center at RPI that flicker can reduce reading speeds, so it can reduce task performance. We don't really know exactly what safe levels are which have been established. We can tell you what very, very safe levels are but some of those are unrealistic, given the kinds of products that we have available for lighting these days. So there are groups working to establish what exactly safe levels of flicker might be for different populations and for different kinds of applications.

Drivers that use pulse width modulation will produce 100 percent flicker and you can eliminate that if you go to very high frequencies. We think that those frequencies need to be above 500 hertz, possibly 800 hertz, possibly 1,000 hertz. That number has not been fully established yet. Drivers that use constant current reduction, that's CCR, may be more acceptable because they don't actually reduce the light output to a zero point during the course of the cycle.

It's important for you to see an LED product before ordering it because we don't have firm flicker tests at this point in time. So getting a flicker wheel to test products or waving a pencil or waving your finger very rapidly underneath the light source can help you see the stroboscopic effect that is a result of flicker.

Next Olympic event here for our decathlon is light patterns produced on the adjacent wall. So we know that funky light patterns can make spaces look cave-like. Basically, all the troffers we looked at were fine except for the three parabolic louver troffers and of these, you can see here three fixtures marked B, D and K. The one on the right is fluorescent. The one on the left and the one in the middle are fluorescent fixtures that have been retrofitted with T8 LED tubes. Notice the LED tubes, because they don't emit light off the backside of the tube or at least the ones we tested did not, they didn't send any light onto the rear reflector of the parabolic troffer. So as a result, the pattern of light that you got out didn't have any bounced reflective light and therefore, the pattern of light on the wall was actually harsher than what you normally get from a fluorescent tube that does send light out the back side. So you can see on the wall, the far wall there, a softer pattern on the wall from the fluorescent fixture than from the LED tubes.

You can also see one more thing that's happening from the LED tubes. Since they don't emit light on the backside, the contrast between the LED tube in the parabolic adjacent to the reflector that you can see through the parabolic louver is greater than it is in the fluorescent. Now, it's neither good nor bad. You just need to be aware that if you're specifying LED tube products or installing LED tube products, that this could easily happen.

Overhead glare and direct glare. Glare is an enigma. It's very frustrating trying to find exactly what factors from luminaires trigger glare response. We tried a lot of different factors connected with or photometric characteristics connected with the luminaires. The one that ended up working best to predict the observers' experience of glare was looking at the maximum spot luminance on the luminaire measured from ten degrees from vertical. So this is really the kind of glare that you would feel from overhead and that might make you feel like you want to wear a baseball cap to get rid of the glare.

When there were spot luminances greater than about 20,000 candelas per square meter as measured with this luminance meter, spot luminance meter, the fixture was rated as uncomfortable. The same thing happened with direct glare, discomfort glare. Also, it was related to the maximum brightness on measured brightness on the luminaire. It was compounded when you had dramatic brightness patterns. You can see in all three of these fixtures, if you're looking at the luminaire, it either looks stripy, or it looks like there are caterpillars crawling across the surface, or it looks really weird and strange. And when it looks really weird and strange, you have these bright/dark patterns that are kind of distracting. That's when the observers responded that it was uncomfortable.

Now let's look at visual appeal. If you looked at basic categories of fixtures, the products that came out most highly rated were the ones that you see on the top of the pedestal there. So they like that appearance best. The one that came in second were the diffuser or prismatic lens fixtures, when it was a partial lens in the whole 2x4 or 2x2 aperture. The ones that came in sort of rated indifferently were the white diffuser products. Whether they were LED or fluorescent didn't matter in any of these issues. The ones that came in least liked were the ones with funky dramatic brightness patterns. And so the two at the bottom where the ones that basically were cut from the team.

Okay, we've looked at the decathlon events. Now let's look at the drug test. What happened when we looked at the retrofitted troffers, fluorescent troffers that were retrofitted with either tube retrofit kits or LED replacement kits. So the biggest problem that we found was that the kits were not labeled properly. It's really important when a fluorescent fixture is changed out, permanently changed out to a different kind of technology, to make sure that there's a transfer of responsibility from the original troffer manufacturer to the kit or the tube manufacturer. Those labels were in many cases missing or sometimes the label was confusing. It listed a lamp that was different from the tube that had actually been installed. It, in many cases, didn't say, "Thou shalt not put in a replacement fluorescent lamp into the socket," because, of course, there could be some safety issues if that were ever done by someone who didn't know the troffer had been relabeled. So that is number one issue. Labels were a big deal.

So manufacturers of kits and tubes need to be extremely cautious about making sure that instructions have those labels in them. Also, some retrofit tubes are wired differently. Some are wired from both ends; some are wired from one end; some are wired from a separate driver that is mounted basically in place of a ballast, but the wiring is different from the original wiring of the fluorescent ballast. So because of those differences, it's really imperative to make sure that the user of this troffer knows that they can't just replace a tube LED product with another tube LED product, because it may not be compatible, electrically.

There is very limited interchangeability among all of these products. The tubes must be NRTL approved and they must comply with UL1598C, which is a standard for retrofit kits, and it must also comply with UL1993. The retrofit LED kits must be NRTL approved and comply with 1598C only, not UL1993.

Another problem that was identified is that not all fluorescent troffers are heat tested during their safety testing, because some fluorescent troffers are temperature exempt and as a consequence, the retrofit product cannot automatically assume that the chassis of the fluorescent troffer is going to be an adequate heat sink and that it will not become a heat problem in the space. So it's awfully important to make sure that any retrofit kits are labeled for specific types of fluorescent troffers. Over half of the tubes and kits that were installed in our installation could've been disqualified by an inspector.

What did we learn? Well, let's start with the LED tubes. They produce some funky luminaire appearance, including some stripy appearance on lenses and some stripy patterns in parabolics and some funky patterns on walls. It's very important to choose the beam angle on these tubes very carefully. We now know that these beam angles can range between about 90 degrees and up to – they're advertised as up to 220 degrees – and that will make a very different pattern of light in whatever luminaire you put them into. These LED tubes can change the distribution of light from the luminaire, particularly when the luminaire relies on the light from the backside of the fluorescent in sculpting the light pattern that's coming out. And you see here a photometric report, a before pattern from the batwing distribution of the parabolic and then you see that after it's been re-lamped, it now is producing basically a cosine type distribution, so that's different from what it was originally supposed to be.

LED tubes can have some unexpected installation problems, especially concerning sockets. Sockets are the weakest point in retrofitting LED products. They can be good. LED tubes can be good if they're well engineered, but they may not save you much energy unless you drop light levels.

Let's look at the LED troffers. So the tubes got the bronze medal. The LED troffer kits are, you're gonna see, get the silver metal. They can be good if well engineered, but again, they may not save you much energy unless you drop light levels, and that's basically because the efficacy is, in some cases, pretty much the same as the efficacy that you get from a fluorescent fixture, but if you're careful with what you choose, you can find kits that are higher efficacy and very good products. Some of them have nationally recognized testing laboratory labeling implications and you must check for compliance with UL standards. Check to see if there's any limitation on which manufacturers' troffers can be used or the depth of the troffer. We did find that some information on the kits did not specify how deep the troffer needed to be and if it wasn't deep enough, that was a problem for installation. Look for 70 lumens per watt or higher on the kit as installed in the replacement. So there's the silver metal.

Let's go to the LED troffers that were born as LED troffers. The efficacy is very high. These puppies showed great promise. Look for LpWs of 90 or higher. Avoid products that have dramatic brightness patterns on the lenses because they can be glaring and distracting. The color range is all over the map, but most of the products were as good or even better than fluorescent. Flicker was a problem with some dimming LED drivers and there's no complete metric at this point in time. Dedicated LED troffers are a really good option for new installations. In fact, if price weren't an issue, it would be kind of a slam-dunk solution. They get the gold medal but it's really important to see it and mock it up before you buy a bunch of them.

Now, let's talk about the next competition. How do we make sure we're getting the better troffers? The Better Building Alliance, which is another Department of Energy program that Linda heads, has produced an LED troffer specification. This shows a target minimum luminaire efficacy of 85 lumens per watt; a warranty period; power quality recommendations; driver efficiency recommendations; minimum luminaire lumens for three different sizes of troffers that you see there; a minimum spacing criteria of one to two in both the parallel and perpendicular planes for the luminaire distribution. Color rendering index greater than 80 and an R9 greater than zero but you now know that you can do far better than that with a lot of these troffers. Lumen maintenance – this is what's recommended with the BBA troffer specification. You can now find products on the market that even beat this.

So we have the Better Buildings Alliance specification available at this website. Here's the conclusion from our 18 observers. The LEDs have not made troffers better, but they have made them more energy efficient. So in places where you need to use troffers for whatever reason, yes they're not the most glamorous product on the planet, but they often have application in offices, cafeterias, corridors, lobbies, classrooms, in places like that, we can go to LEDs and LED replacement tubes and kits, if we're careful, and do a very good job with them.

The full report, which names names on all of these different 24 products, is available at the DOE website. Here's the link, although the easiest thing to do is just go to Google and put in "DOE CALiPER, troffer," and click on the first Google listing and that gets you there. Thank you very much. It's been a pleasure and great fun and I've learned a heck of a lot on this CALiPER troffer study and it was great to try to communicate what we've learned to the audience out there in Internet Land. Questions?

Linda Sandahl: Thank you so much, Naomi. We did get a number of questions and we've got just under about 15 minutes to get through these. The first question is: what is the typical lifetime of an electronic ballast and how often do they typically get replaced in a commercial office?

Naomi Miller: My current understanding is that if you're buying good quality electronic ballasts, the typical lifetime is about 60,000 hours. Now, my information may be old. They may be – vary from there, but that's a good target value and it's worth asking the fluorescent luminaire manufacturers what their listing or expecting from their ballasts. Electronic ballast life can be shortened if they're put in very hot environments, so it is dependent on the temperature of the plenum that they're going into.

Linda Sandahl: Okay, thank you. The next question is: given significant progress that's been made in lumen output improvements and cost reductions for LED products over the past two years, where do you expect LED penetration to go for the troffer market in the 2013 to 2015 time frame?

Naomi Miller: Boy. I have seen some amazing troffer prices. It looks to me that in large quantities, LED troffers may not be significantly more expensive than comparable fluorescent troffers, especially if you're looking at dimming fluorescent troffers. So my guess is fluorescent is going to lose market share pretty dramatically in new applications where you're buying large quantities of troffers at a time.

Linda Sandahl: Okay. The next question is: do you think there will be an LED equivalent to fluorescent custom ballast factors which are used to meet LEED or design goals watts per square foot or lumens per square foot?

Naomi Miller: What a good question. I don't know that I know the answer to that, offhand, but I can tell you that I have seen LED troffers that are available in two light output levels or in some cases, three. So I think it is possible to use the same appearance of LED troffer in different spaces in a building and have, let's say, 2,000 lumens coming out of one 2 x 2 troffer and the identical looking troffer in another room can be producing 4,000 lumens or 4,500 lumens. So can you change out the driver to do that? I think that's a good question to ask manufacturers.

Linda Sandahl: Okay. The next question: is there a scale/Excel available for the replacement of fluorescent tubes on a lumen-per-watt basis, et cetera?

Naomi Miller: A scale. Oh. You know, it's kind of tricky. If you're looking at LED tube offerings, they often, the manufacturers will list the lumen output for the tube and, of course, the wattage draw, but the tube manufacturers don't always tell you what the lumen output is once they are put inside of a troffer. So my preference would be to have tube manufacturers' photometer the most typical troffers. Let's say a two-lamp, three-lamp, or four-lamp lensed, K-12 lens troffer; a two-lamp, three-lamp, four-lamp parabolic, and other variations of fixtures with their LED tubes and give you lumen output and distribution from those luminaires rather than just a lumen value for their lamp itself. Because the lumen value does not necessarily transfer to the lumen output that you're expecting from the troffer. So if, for example, I had a troffer that is – let's say it's a two-lamp fluorescent 2x4, and it's emitting, it has 6,000 lumens from two 3,000 lumen fluorescent lamps. Now, I put in, instead, two 1,500 lumen LED lamps. Am I going to get half the light in that troffer and will I get the same distribution? The answer is there's no way to know unless you have that specific fixture photometered with those specific lamps. It probably is going to produce something higher than 50 percent of the light, depending on the tube beam angle, but what's going to happen to the light output and what's going to happen to the light pattern, it's hard to say. So I would encourage you to talk to manufacturers of the LED tubes and have them provide you the information on what happens to their tubes once they are photometered inside typical troffers.

Linda Sandahl: Okay and can you comment on the observer demographic? Were they lighting savvy, non-lighting people, or a mix?

Naomi Miller: They were all lighting savvy. So they knew what they were looking at. They were primarily lighting designers but we had some lighting engineers. We had a facility engineer from a Portland medical facility. We had one theatrical lighting designer, who was from the University of Florida. So it was an interesting group of people. They had different kinds of backgrounds and we had some really good discussion.

Linda Sandahl: Okay. Were the LED tube retrofits and the LED retrofit kits lumens-per-watt based on the source only, or on the source in the fixture?

Naomi Miller: I'm sorry, would you ask that again?

Linda Sandahl: Sure. Were the LED tube retrofits and the LED retrofit kits lumen-per-watt based on the source only or on the source in the fixture?

Naomi Miller: Ah. It was the LED kits were based, kits and LED troffers, were based on lumen output from the luminaire. And the LED tube replacements were based on the highest lumen output we could find from LED tubes that we could also get within the ten-week time frame.

Linda Sandahl: Okay and how did you measure or calculate lumen-per-watt?

Naomi Miller: We were looking at the lumen output from the luminaires divided by the input watts to the luminaire. This was all not based on tubes but on luminaire efficacy.

Linda Sandahl: Okay and on slide 14 is that the total system efficacy?

Naomi Miller: Yes.

Linda Sandahl: Okay. Regarding the energy use of LED tube retrofits, were the drivers integrated in the tube, i.e., the fluorescent ballast was disconnected or removed or did the LED tube run off of the fluorescent ballast?

Naomi Miller: None of them ran off of the fluorescent ballast. Four out of five of them ran on 120-volt power. The ballast had been disconnected and removed from the fluorescent luminaire. One of the products was a prototype product that had a separate driver so that driver essentially replaced the fluorescent ballast. All of the products mounted using the fluorescent sockets, the existing fluorescent sockets, although in some cases, we had to change the sockets in order to make sure it was either a shunted or a non-shunted fluorescent socket, depending on what the specific lamp needed.

Linda Sandahl: Okay. And how much do you estimate that the fixture in which you put the LED tube impacted the CRI, the average and R9?

Naomi Miller: Practically zero. We did get photometry of the LED products in open air. We compared the listing of the color characteristics that came from the manufacturer and the color characteristics that were measured in open air to the color characteristics that were measured in the luminaire. The change was very, very small, if any.

Linda Sandahl: Okay. And will we have access to all the data, such as what brands had noticeable flicker in dimmer mode?

Naomi Miller: Yes.

Linda Sandahl: That included in the report?

Naomi Miller: That is in the report.

Linda Sandahl: Okay. From my understanding –

Naomi Miller: Now I, I've – Linda, let me add one caveat. We know that manufacturers are becoming more aware of flicker and it, remember that these products were purchased about a year ago. Even the products that are listed as having flicker issues in the report, I think it's fine to contact those manufacturers and say, "Have you changed the driver so that it would eliminate any flicker issues in the future?" Because let's give the manufacturers the benefit of the doubt. They're seeing this data at the same time that you are. Many of them may realize that they need to change their driver design in order to take care of the problem. Go ahead.

Linda Sandahl: Okay. What is PWM on slide 25 stand for?

Naomi Miller: I'm sorry. I should've explained that. PWM means pulse width modulation and that is a technique for dimming an LED by turning it on and off very, very rapidly.

Linda Sandahl: Okay and where can one buy a flicker wheel?

Naomi Miller: Well, to be perfectly honest, the way we got our flicker wheels here was we called up Sylvania because they had manufactured flicker wheels for, back in the '90s, to help you identify whether a fluorescent troffer had a magnetic ballast in it or an electronic ballast and I know that they keep some on hand. So call your local Sylvania representative and see if you can get some.

Linda Sandahl: Okay and we have time for probably one more question. From my understanding, California Title 24 will require dimming controls to be installed in new lighting and renovation when greater than 10 percent of lighting is being upgraded. Do you see a lack of product availability to support the dimming requirement, given the amount of flicker your survey experienced?

Naomi Miller: No, because I think the manufacturers are recognizing the issue and they, we already know that we have two-thirds of the products that did not flicker, so I think that manufacturers are going to be ready to provide dimming troffers without flicker.

Linda Sandahl: Okay, great.

Naomi Miller: By the time that is enforced, let's put it that way.

Linda Sandahl: Okay. Well Naomi, thank you, and thank you everyone else for participating in today's webcast. Watch for the PDF of today's presentation and the video on DOE's SSL website in the next couple of weeks and you will receive an e-mail here soon asking for some feedback and the e-mail will also include instructions for obtaining continuing education credits, in case you were interested in that. So again, thank you all. This concludes our webinar and you may all disconnect.