Text-Alternative Version: Understanding and Evaluating In Situ TMP and LM-80 Reports

Below is the text-alternative version of Day 2 of the three-part TINSSL Webinar Series "Understanding and Evaluating LM-79 & LM-80 Reports for SSL Technology," held August 11, 2010.

Sarah Eckstein: Today's webinar will focus primarily on the understanding and evaluating in situ temperature measurement points and LM-80 reports. Tomorrow's webinar—and, as you notice, the time difference for tomorrow starts at 11:00 a.m. Eastern Standard Time, 8:00 Pacific, and that webinar will focus on tying in what we've went over and covered in days¬†1 and 2 and putting it all together, walking through examples of real test reports.

So I apologize. Several of you that were on the phone yesterday, it'll seem a little repetitive. The next three slides I'll cover. So I just ask those that were on the call yesterday to just stay patient while I go through these quick three slides.

And essentially, I'm gonna just give a quick overview of who NEEP is, what NEEP provides as a technical information network for solid-state lighting, and what the DesignLightsTM Consortium for Qualified Products List is. Jason will be referencing this program in his presentation.

After presenting these three slides, I will introduce Jason Tuenge and Kelly Gordon. So, NEEP is a regional energy efficiency nonprofit that collaborates with policymakers, energy efficiency programs, and industry to accelerate the use of energy efficiency in buildings and in homes.

The Technical Information Network for Solid-State Lighting essentially increases the awareness of solid-state lighting technology, performance and application. NEEP has received a grant from the Department of Energy to deploy information and dissemination related to solid-state lighting.

TINSSL members who make up a broad audience receive updates on solid-state lighting technologies and opportunities to participate in education seminars such as this one.

One of the other activities I wanted to mention that NEEP is heavily involved in is the Design Lights Consortium for Qualified Products List project. NEEP is the project manager cooperating with 26 other utilities and energy efficiency programs throughout the nation and Canada on this particular project, and they work alongside—well, we work alongside our contractor, our third-party contractor, to evaluate and verify manufacturers' product cut sheets comparing that to their LM-80 and LM-79 test reports.

And the main purpose for setting this project up originally, and still is present 'til today, is to build a prescriptive list of commercial LED products and essentially helping those utilities weed out the junk and recognize and award those high-quality, high-energy-efficient LED products that are out there.

So without further ado, I'd like to introduce our main presenter for today's webcast, Jason Tuenge. Jason joined PNNL in 2008 and has ten years of experience in the lighting industry. He provides lighting engineering support for DOE's activities within the Commercial Building Initiative and the marketing-based programs for solid-state lighting.

Jason also manages the technology fact sheets program, producing educational material on topics ranging from solid-state lighting itself to human factors and economics. And I'll introduce Kelly Gordon, who will support the Q&A portion of today's webinar.

Kelly has been with PNNL since 2000 and has more than 20 years of experience in the energy efficiency field with a particular focus on efficient lighting technologies. Kelly currently manages PNNL's activities related to DOE's L Prize competition and has had further involvement in the development of the DOE ENERGY STAR criteria for solid-state lighting luminaires and integral replacement lamps. So, I am handing it over to Jason, and we'll start today's webcast.

Jason Tuenge: Okay. Thanks, Sarah. Can everybody hear me fine?

Sarah Eckstein: Yes, I can.

Jason Tuenge: Okay. Thanks. So over these three days, yesterday, today, and tomorrow, as we've mentioned before, we'll be covering how, basically, you connect LM-79 reports with LM-80 reports. Yesterday looked at LM-79 and how it can be used to cover items like light output, efficacy, CCT, CRI.

Today we'll be looking at LM-80 and showing how it can be used to actually track the lumen depreciation for a source. And we'll also be seeing how ISTMT is used to connect the dots between the two.

LM-80 is like LM-79 in that it's dedicated to LED technology. And its focus is lumen maintenance, but also measures color maintenance as well. It does not provide any mechanism for or any method for estimating a product's useful life or extrapolating data. And if you've gathered data to 6,000 hours, it doesn't tell you what your light output is gonna be at, say, 60,000 hours. It only tells you how to take your measurements up, you know, as the product is operating over time, no extrapolation from that data.

And it's intended strictly for LEDs themselves, so LED packages or modules, not luminaires or integral lamps. Unlike LM-79, LM-80 does appear to mandate content for LM-80 reports. So instead of just recommending items to be included in a report, it actually lists items that must be in a report.

It does stop short of providing a format that does at least list the things that should be in the report. And so, as I mentioned before, the primary goal is to establish the test environment to track lumen depreciation and also to look at color shift over time.

Here we've got a diagram that's probably fancier than anything you're gonna see in the LM-80 report that was produced—this image was produced by CALiPER as part of their long-term testing program. And here you can see—really, the reason for showing this diagram is that you don't necessarily get a smooth shift from one color to another. You know, maybe from kind of a warm white off in the direction of green or something, sometimes like in this sketch that's here. Looking at this one, in particular, you can see this one ventured off in all sorts of directions and backtracked and almost ended up back at the same place it started by the time it was all said and done. That may not have been acceptable to occupants to sit under that kind of unusual effect. Probably undesirable.

Oh, I need to get this—there we go. So LM-80 requires three—we've got some strange formatting going on here. Sorry about that. Requires three measurements at different temperatures, so three different test conditions where you're controlling for the temperature that the LED's operating at and holding everything else constant.

And so it's not okay to create these different temperatures, for example, by changing the drive current, the amount of current passing through the LED. You need to accomplish that some other way so that you've isolated all other variables and all that you're looking at is this temperature effect.

You can then test additional drive currents if you want, but then for these other drive currents, you want to, again, test the same temperatures.

So, as mentioned before, we've been using the NEEP DLC QPL as kind of a point of comparison to see how this information can be used. And one thing this program does is demand that on top of the other LM-80 requirements that data is presented in a readily discernable manner.

In other words, you know, it's not okay to have everything just simply in tabular format, and it's not okay to have multiple charts. They all need to be merged into one diagram that can be understood quickly and then have a manufacturer actually identify, you know, what the lumen maintenance is, you know, at 6,000 hours based on the available information. It makes it easier for us then to check that and say yes, it looks like you interpreted that properly.

And as I mentioned yesterday, drive current appears to be something that is not fully understood by a number of people interpreting these reports. It seems like there's some confusion between drive current and the input current to the luminaires. So it's—you know, just 'cause you know how much current is going into your luminaire or your lamp, that does not mean that's the same current going through your LEDs.

And then even if you get down to the output of the driver, that's not necessarily the same current that's flowing through any given LED. So it's kind of a tricky thing. And at some point, you basically end up having to take the manufacturer at their word that yes, this is the current, which is really a rated current passing through the products once it's actually installed in a lamp or luminaire. It is measured by the LED manufacturer for the LM-80 report. The trick is then tracking it afterward.

This is an example not of an LM-80 report itself, but of literature from a luminaire manufacturer illustrating that the input current to the fixture is not necessarily the input current to the LED. So, you know, the input current to the fixture is not LED drive current. Here we've got 0.23 amps going into the fixture and 350 milliamps rated going to LEDs.

Yeah, something's going on with the formatting here. I'm not sure what's going on, but at least the bullets have it at the right look. So yesterday I referenced ENERGY STAR because that is currently referenced by the NEEP program. Here is where you find the definition of this—of, you know, TMP or ISTMT that we've been mentioning here and there.

This is basically the way that you connect LM-79 to LM-80, is through measurement of temperature on the LED after it's installed in the product, you know, the lamp or the luminaire. And so this table gives kind of an illustration of how this works. You know, you get some information on LM-79 that you don't get in LM-80 and vice versa.

And so, basically, the two are linked by ISTMT or—yeah, ISTMT. And there's also some confusion on package array module, and you can see this reflected in the revisions to the IES RP 16 document that defines these terms. So I've provided another table below to give a little bit of clarity here.

Basically, when you refer to an LED, it's really an LED die. And then a package is then a number of LED dies. It could be just one, might be several, and then a module or an array are then multiple packages assembled together.

And the reason this is important, this affects what your—like the number of tests that are required, or the number of samples required, in a given test.

So the in situ temperature measurement test is what provides the measured temperature for the LED product once it's integrated. There's a particular place where you're supposed to take this measurement on the LED product. There is some flexibility, and it's up to the manufacturer to establish exactly where it is tested.

And as you look through the literature, you'll see that manufacturers really kind of change these designations around. Beyond just picking which one they're gonna use, they'll—instead of indicating TS, they'll say TC. They'll capitalize instead of leaving lowercase, so that you see the LM-80 really doesn't get followed strictly, but the general intent is usually there.

So we've shown a couple of examples down below of some of these products. Here's a thermal image taken, I believe, by a camera to show—to give a sense of, you know, the thermal distribution on the face of the luminaire and where you're likely to find some of the hottest products.

This is not something that can be used currently for ENERGY STAR purposes. You have to stick with thermal couples, not images like this, but at some point, that may be something that's acceptable. This at least gives you a sense of what's going on.

The idea is to follow LM-80 in terms of where you measure on the LED package or module. And then going one level above this, then, you know, which LED package or module do you measure? Well, it's the hottest one in the fixture. That's the one you're most—you know, the hottest one on the luminaire, the lamp. That's what you're most interested in, the worst-case scenario.

It's worth mentioning 'cause you'll see this on these reports as well that—and, you know, as required by ENERGY STAR, the power supply should also have a measurement like this done. This is done anyway for other power supplies or ballasts used for other products like fluorescent.

And here I've shown an example. I believe—yeah, I'm not sure what kind of a ballast this is, but it's not an LED driver. Same idea. This is usually used for warranty purposes for those products, and the same pertains for LEDs. It just doesn't necessarily tie directly into LM-80. Actually, it does not tie directly to LM-80, but it can affect your liability.

So the in situ measurements basically give you—when you do LM-79 testing, your product is basically levitating in midair. It is not recessed into the ceiling if it's a downlight or installed in any other manner that's realistic.

Really, the way these things are tested is they just—they're floating in midair. So by—and one of the requirements of the in situ testing is not only that the LED is installed in situ in the luminaire or a lamp, but then in that product, the luminaire or the lamp is installed in situ itself, that it's recessed into a ceiling or something.

And so this is per the UL 1598 standard in most cases, but also for supportable products, there's UL 153. This diagram illustrates some of these different installations for luminaires. Then in addition to this, for lamps, UL 1993 is referenced. And then NEEP also references UL 1574 for track lighting fixtures specifically.

So by having these installed, say, in a ceiling, the product is gonna be running hotter. So, again, it's more conservative than just what you—what's done for LM-79. This just further worsens the picture in terms of lumen maintenance, which is a good thing when you're trying to estimate conservatively safely.

So here we've got an illustration of how you actually would go about using all of this information. And the text has moved around a little bit, so sorry again. Basically, there are several possible scenarios. Scenario 1 of ENERGY STAR compliance path 1 is that your measured TMP that you get in the luminaire ends up being greater than the maximum temperature measured during LM-80.

If this happens, then basically, you don't have useful data from the LM-80 testing. You're running your product too hot, in short. And so how far off target are you? It's kind of hard to say. It might still actually be okay for some applications, but in order to know whether this is really gonna be acceptable or not, you're gonna have to submit for ENERGY STAR Option 2 and not Option 1. And this is where you test the whole product, the whole lamp or the whole luminaire. And you don't use LM-80 at all. You test once at 0 hours and then again after 6,000 hours of operation and in the appropriate environment, as per ENERGY STAR. And then you see how well the thing has held up over time. So that's what happened when your temperature measured is higher than it should be.

So Scenario 2 is when you're actually running very cool. Your product is—your LEDs are running very cool and at a temperature lower than was actually tested during LM-80. In this case, what ends up happening is you get a specially conservative estimate. So your product is probably gonna maintain better than what ENERGY STAR estimates by its extrapolation method.

And so what you end up doing is you just simply take—you know, say you measured it—found that it was operating at 20 degrees C, you round that to the lowest temperature measured in LM-80, in this case 25 degrees. And then say that, okay, so it's 6,000 hours. That means we were at 99 percent in this case based on this chart—we were at 99 percent of initial output, which is actually doing really well.

But this is the approach you have to take. You have to be a little conservative. You might have been doing better than whatever you get by this approach in reality.

Then Scenario 3 is where most of these products are probably gonna land. And what's involved here then is a little bit of interpolation between two of these temperatures that are reported in LM-80. And so the equation is given here. And then to see how this is used, say you measure at 73 degrees and you've got data points at 55 degrees and 85 degrees, so basically curved at both of these temperatures, and these are both the same drive current.

Then you just interpolate between the two to find that you get 92 percent of initial output at 6,000 hours. From here, then, ENERGY STAR provides a means of saying—okay, so if we're going to base on a defined—curve defined by ENERGY STAR, assume the same depreciation curve, you know, the same form of a curve for all products, and then say that you have to be at a certain point at 6,000 hours. If you're below that point, it shows that your curve is gonna drop too low, too fast.

And so basically, there are a number of values given, depending on how many hours of testing you've done, it allows you to extrapolate out to a certain rated useful life. So, sorry. This is basically serving as kind of a temporary estimate while we wait for IES's TM-21 to be finalized.

So in yesterday's presentation mentioned that we really want to have all this testing done by somebody qualified and independent, if at all possible. So for LM-79, we do have that, and we provided a link yesterday to the CALiPER site where you can find a bunch of these listed.

For ISTMT, the options are NVLAP, which is also an option for LM-79, CALiPER, again, also for LM-79, and the UL Data Acceptance Program, which is not something that would work for LM-79.

LM-80, on the other hand—so those were all independent labs. LM-80, on the other hand, is currently viewed as something that is prohibitively expensive for an independent lab. There has been some interest by at least one independent lab, but for the time being, we still don't have anybody that's, say, NVLAP-qualified to do LM-80 testing.

So this is something that we're still looking into and for the time being, we're using manufactured data. Here's a recent statement. These are still using manufactured data. Here's a recent statement from ENERGY STAR now under EPA, basically acknowledging that we're still kind of in an interim mode where we will take data from manufacturers so long as they show that they're working through the NVLAP process, and likewise for other independent laboratories.

So we're still kind of just needing to work with what we've got for the time being, which is certainly better than nothing. So, as we did yesterday, and the intent of—you know, part of the intent of these three days is to leave you with a checklist. This is a lot of information, and it's nice to have something where you can kind of step through the process step by step to be sure that you're covering all the major points.

One, just to make sure that you trust the information you're looking at. And then once you've accomplished that, to then—to look at the specifics and say okay, did they actually meet our thresholds?

So first off, you want to be looking—in the case of ISTMT reports, am I looking at something that's actually from a qualified test lab? And does it look like it's—you know, does it seem possible that this information could have been manipulated by somebody, you know, whether it's a manufacturer or a salesperson. Sometimes things get changed by accident, sometimes possibly on purpose. So it's nice to, you know, see something ideally come directly from the lab, but, you know, otherwise be in a PDF format, hopefully not generated by someone other than a lab.

As with LM-79, you want to make sure that—and this is fundamental and not simple. You want to make sure that what you're looking at is actually—on the one hand, you've been submitted a product that, you know, you're deciding whether you're gonna approve and on the other hand, you're being—you've been submitted with—provided with information that might not actually apply to that product that was submitted.

That might be a report for some other fixture and does not—is not valid at all for the thing that's actually been submitted. So the first thing you want to do is compare the catalog numbers. There's no reason why the catalog number needs to be inconsistent between the two, but as it turns out, this is often the case. You know, as you look between the two, you can't reconcile the catalog numbers.

So this is something to make sure you get a comfort level with that you understand okay, this is, you know, a catalog number, format change, or something to explain why these are different or that one is more conservative than the other, part of a family, product group or family.

You also want to look at the descriptions as, again, with LM-79 to make sure that there weren't any changes or anything that's not conservative. You know, you want to make sure you're looking at the same drive current, the same thermal management, there's no fan that was attached to keep the thing cool just for the testing purposes, input power is comparable, etc., etc.

Looking at LM-80, again, for the time being, we really don't have qualified test labs. We just take the data from the manufacturers. And you—again, you want to try to make sure that you've got something that's consistent with the product that's submitted, but as we mentioned yesterday, this is a lot easier said than done.

The test labs—you know, even when you get an independent test lab reviewing these products, they're gonna have a hard time, or it may even be impossible in many cases to actually verify that the product that the luminaire manufacturer—you know, the LEDs that the luminaire manufacturer says are installed in the luminaire—are actually the LEDs that were installed in a luminaire. You know, you can't really prove that it is. You have to just take the manufacturer's word.

So—but you can still check things like color temperature, drive current and least make sure they're consistent. Also look at product photos. You probably aren't likely to see on something like this, you know, where you've got, you know, LM-80 data for one type of LED and then you see a photo of LEDs installed in the product as something else. But there's some things like this that could be a tip-off. It's worth at least looking at it.

So you want to make sure that, again, they are actually following LM-80. Again, this is not an independent test lab, so these manufacturers, it's good to actually check and make sure they're following LM-80, they're not independent.

So look at the temperatures that they actually operated these LEDs at and make sure that they're valid: 55 degrees C, 85 degrees C and something else; that their tolerances are good; that the ambient is per spec; that input current is controlled properly; the environment is controlled properly.

All this is actually detailed in the report. Voltage, all—a number of items here. Again, this is why a checklist is good 'cause otherwise, you know, reading through the LM-80 report, it's a lot of text, and it's easy to miss items. So it's nice to just work through the checklist and say yeah, I don't see where they're actually indicating that our thermal couple is ASTM E230.

So some of this, the important thing is that the manufacturer meets the intent, and some of this is—and so the previous page, some of those items may be more a matter of just, you know, as long as you meet the intent, at least you have something. But the LM-80 items are actually mandated per LM-80, not optional.

And, again, things that are simple as the minimum 6,000 hours of testing, the interval no longer than 2,000 hours between measurement points, some of these things, as simple as they seem, they gloss over. So it's good to check all these items and make sure that you really—you've got a complete report.

So getting back to actually—you know, so basically, you can then go back—once you've worked through all these items, you can go back to your set of criteria, you feel comfortable with the information you're looking at, now you can go through and start—and doing your final punch, saying that, okay, yes, you know, this threshold is met, that threshold is met, and now the product itself is okay. Not just the report is okay, but now the product is okay.

That's it for now. The purpose tomorrow, the purpose of the webinar tomorrow, is gonna be to take these checklists that we provided for LM-79, ISTMT, and LM-80 and show how you'd actually work through some of these submittals that you'd receive so that you can very quickly determine, you know, one: is the submittal trustworthy in terms of the information that's provided? And is it complete? And then two: once you've established that, then to look at okay, now, does the product actually meet our requirements?

So I think you'll find tomorrow very useful. Now we're going to get to some questions. It looks like they've been trickling in. So I think Kelly will be joining here now to help out.

Kelly Gordon: Yes, I'm on.

Jason Tuenge: Let's start up here. I have not been watching these as they come in.

Kelly Gordon: Okay. I can read the first one. The first question was: "I mostly see UL 1598 listed on in situ reports, but have seen others, as you mentioned—UL 1574, UL 1593. How do I know they're using the correct test for the particular lighting type?"

Well, the types of luminaires or products that would be addressed by each of those UL standards are listed in the standards. So for all hardwired fixtures, you know, that's UL 1598. UL 1574 is, as I understand it, for track fixtures. I'm not sure about 1593, but 153 is for portable luminaires with a plug-in cord. Do you have anything to add to that Jason?

Jason Tuenge: I think just the one thing I would say is that, you know, as we mentioned yesterday for the LM-79, at least one thing that you can trust to a certain point is you're dealing with a qualified lab, and so they should be pretty good about doing the ethical things. You know, if a fan was attached, they'll report it. They won't just gloss over that.

And so you can have a certain comfort level that they'll also test it properly, they'll choose the appropriate UL standard, as opposed to something like LM-80 where it's not necessarily something that's really qualified, right? So that's one thing that, you know, it—probably not a critical item on the checklist to make sure that the lab actually grabbed the right UL test, but it's something that, you know, is definitely worth looking at.

Kelly Gordon: Okay. The next question was: "Why not specify a third temperature for testing for consistency?"

So this is in the LM-80. And I know in earlier drafts of the LM-80 test procedure, they did have three specified case temperatures. And at some point during the process to finalize that test procedure, it was determined that they would set 55 C and 85 C as to required temperatures, but leave the third temperature to the discretion of the manufacturer.

And I think that the reason for that is that different packages and modules behave differently and, you know, there needed to be some manufacturer judgment of what is the most useful temperature to test their particular product at. So that's why that was allowed to be determined by the manufacturer.

Jason Tuenge: Yeah. And, I mean, I can think of at least one LED product that, you know—or an LED package or module that it was definitely targeted for outdoor purposes. And so they might have a sense that okay, this is most likely gonna go into a—you know, a shoebox or a cobra head, and then might also have the sense that okay, for that reason, we can test it—or we should test it at a lower temperature or a higher temperature, you know, whatever they feel is appropriate.

Kelly Gordon: Right, right, exactly. Okay. The next question was: "I missed where we can get the checklist."

So the checklist will be in these PowerPoint documents, which will be posted on the NEEP and the DOE SSL websites.

Okay. The next question: "Several slides include bullet text in quotations without a source reference. Are these quotes typically taken from the LM documents or elsewhere?"

Okay. So the lists of items to be reported, those are taken directly from the IES LM-79 and LM-80 test procedure documents.

The next question: "A certain amount of color shift is typical for SSL luminaires. Any guidance you would offer on determining how much is too much?"

Well, the guidance that is in the ENERGY STAR requirements is given in terms of the U prime V prime color space and sets a limit for the color shift over the lifetime. So it's 0.004 distance in that color space. So there is a requirement for color maintenance over time that is specific.

Jason Tuenge: Is that all we've got so far?

Kelly Gordon: I think so. Let me see. I think there was a question about the audio. I think I'm going to have Terry Shoemaker address that question.

Terry Shoemaker: Yes, the audio will be a portion of the .wmv or the video file that we plan on posting next week on the website.

Jason Tuenge: Any other questions? This is a—it's a pretty hairy topic.

Kelly Gordon: I don't see additional questions coming in. Again, we will be going through some specific examples tomorrow, looking at actual product submissions. So we will be able to get a little more into the specifics of how you interpret things that you're actually seeing coming in.

Okay. I think one more—a few more questions are coming in now. We'll wait for those to be forwarded through.

Jason Tuenge: So I think to answer this question—like, actually, Kelly's gonna be better for this 'cause you've definitely got more background on, you know, it seems like the integral lamps, you know, the integral lamp criteria is probably the most up to date and most elaborate along these lines. Is it still using the six times rule?

Kelly Gordon: Okay. So wait a minute. Here's the question that has come through. "If a manufacturer would like to claim a 50,000-hour life, how long do they need to test for?"

So in the—and this is treated in different ways in different programs. So in the ENERGY STAR for Solid-State Lighting Luminaires program, as many of you are aware, there are two life thresholds based on 70 percent lumen maintenance.

One is 25,000-hour minimum for indoor residential fixtures, and then there's 35,000-hour minimum for all commercial fixtures and outdoor fixtures. And there are specific 6,000-hour lumen maintenance thresholds associated with both of those. So if you're gonna claim a 25,000-hour life, you have to have 91.8 percent lumen maintenance at 6,000 hours. If you're going to claim a 35,000-hour life, you have to have 94.1 percent lumen maintenance at 6,000 hours.

Jason Tuenge: But they'd have to test for—

Kelly Gordon: At 50,000 hours, there is not an option in the Luminaire Program, the ENERGY STAR Luminaire Program, to claim 50,000 hours.

Jason Tuenge: So even if you had 10,000 hours of data, you still couldn't do it.

Kelly Gordon: Right. The way it's set up right now anyway is that it's—you simply have to show that you have exceeded the minimum requirement, which is 25 or 35. Now, in the LED integral lamp criteria, we did specify if you want to have a life claim that goes beyond the minimum, which is 25,000 hours, then you have to have additional testing hours.

So if you want to do 50,000 hours, you would have to have a longer test period. I believe it's about 12,000 hours.

Jason Tuenge: Okay. So they actually—they do have a structure for that. Yeah, I thought I saw something along those lines.

Kelly Gordon: Yes. There is a structure. And then in the NEEP DesignLights Consortium program, they have defined 6,000-hour thresholds associated with those longer life claims. So Sarah, I don't remember the specific number for—associated with a 50,000-hour life, but it's a specific level that you have to meet at 6,000 hours to claim that.

Sarah Eckstein: Yeah, Kelly, just to add to that, that's on the website. I don't have that information offhand.

Jason Tuenge: I'm trying to get there right now. Just a second.

Kelly Gordon: Okay. In the meantime, let me look at a couple of additional questions here. "Will LEDs be rated in color temperature as fluorescent lamps are?"

Jason Tuenge: I'm showing—

Kelly Gordon: Color temperature and the other chromaticity characteristics for LED products are defined in an ANSI standard, that's ANSI C78.377. And that lays out the nominal correlated color temperatures. There are eight of them defined there, and that adds to, in addition to the nominal color temperatures that are typically used for fluorescent. So there are eight nominal correlated color temperatures. There's a tolerance around each of those that's defined, as well as the tolerance for DUV, which is the distance from the black body curves of this, so the relative whiteness of the light, as we discussed yesterday.

Jason Tuenge: Yeah. So you can find all that in yesterday's presentation. So yeah, I'm looking at the NEEP criteria. It shows—now, I'm not sure whether this depends on product type, but for 50,000-hours rated life, you need to show 95.8 percent maintenance at 6,000 hours.

Kelly Gordon: Okay. Then it's defined.

Jason Tuenge: So—but yeah, ENERGY STAR actually pretty much holds to the six times rule, right? So if you divide the rated life by six, roughly,¬† actually look at the ENERGY STAR criteria, but if you divide it by six, then that shows you how many hours of testing you need, and that's—

Kelly Gordon: And that is based on a recommendation of the IES Procedures Committee. They recommend not estimating life out more than six times the test period.

Jason Tuenge: And that's still tentative.

Kelly Gordon: Because the data simply gets too uncertain beyond that.

Jason Tuenge: Yeah. And that's still tentative. That's not a hard science.

Kelly Gordon: Right.

Jason Tuenge: And it's their recommendation, their guidance at this point. And we don't even have TM-21 yet, so that's not even something that's written anywhere, as far as I'm aware, aside from ENERGY STAR.

Kelly Gordon: Right. And TM-21, again, is gonna be a technical memorandum that the IES is planning to publish that would provide guidance on how you use LM-80 lumen maintenance data to extrapolate and estimate useful life.

Okay. The next question: "LM-80 is for LED packages, arrays, and modules only. Can you send in the actual luminaire to a lab to get an LM-80 test report?" No.

Jason Tuenge: No. No.

Kelly Gordon: No. You can't. Because again, it doesn't apply to the luminaire. So—

Jason Tuenge: Yes. That's ENERGY STAR Option 2.

Kelly Gordon: Right. So if you wanted to test the full luminaire under the ENERGY STAR Option 2, what that entails is getting an LM-79 photometric test at initial values for the luminaire before runtime, so that's your baseline. Then you would operate the luminaire for 6,000 hours in its intended environment, so that's the UL 1598 environment for the particular luminaire type. Then at the end of that 6,000-hour period, you would do another LM-79 photometric test, and then you would compare those values.

Jason Tuenge: And that does apply to lamps as well.

Kelly Gordon: Yeah. It's the same deal for the integral lamps. Okay. Next question: "There are many lighting companies that have LED tube lights that replace T8s. Have any of these companies submitted to ENERGY STAR the LED tubes within a fixture and provided in situ temperatures and given it UL approval?"

So, at this point, the LED tube light, the T8 replacements are not eligible for ENERGY STAR, so they're not included in the integral LED Lamp Program, nor are they included in the Luminaire Program.

Jason Tuenge: Yeah. So basically, at this point, there's no way you can submit one of those products for qualification under ENERGY STAR. So yeah, basically, the categories are created basically as they make sense. So the idea being there might be some products out there that would qualify and reasonably. You know, the requirements are established based on whatever the things—whatever these products are intended to replace. And at this point, there's a reason why you don't see a category.

Kelly Gordon: Okay. Next question related to that is: "How do the best LEDs compare with high-performance T8 and T5 systems on lumen per watt efficacy?"

This is a—there's several different ways to look at this. Now, the way that we typically try to evaluate LED fixtures is on a luminaire basis. So when you measure complete light output from the luminaire over the complete—the total power into the luminaire, what is the efficacy that you get in terms of luminaire efficacy?

Now, LED products do continue to improve in efficacy, so we have seen some products, for example, two-by-two ceiling troffers that are dedicated LED fixtures that now are able to achieve very high luminaire efficacy. There was a product shown at LIGHTFAIR that was around 100 lumens per watt luminaire efficacy, and that's in a two-by-two format now. So that would compete with a two-foot or U-shaped fluorescent lamps and would exceed the efficacy that you would be able to achieve with fluorescent in that case.

For four-foot high-performance T8s and for T5s, from what we're seeing, there still is—there are not LED products that would compete at the highest efficacy levels of fluorescent. And maybe I'll let Jason elaborate on that.

Jason Tuenge: Yeah. I mean, basically, the two-by-fours, it's gonna be pretty tough for LED to compete for, you know, a while yet, and that's not even getting into cost effectiveness. That's just looking at simple, you know, application efficacies outperforming. But yeah, the two-by-twos, it seems like they are getting pretty close there.

And then there's some other applications that are not anywhere near as widespread as the two-by-four like slot lights, where the aperture is six inches or four inches by eight feet. Some of those products will have a T5 lamp, which, you know, typically those aren't efficacious as T8.

And then it's also behind a diffuser and with an optimal efficiency of 50 percent, somewhere around there. That's not very good. And so in cases like that, we'll probably see LEDs start to outcompete. And those are some pretty expensive products anyway. So I think we'll start to see LED take over, you know, application, you know, product type by product type just like we've seen it do it elsewhere, but yeah, it's not just across the board, you know, all at once, and we're not there yet.

Kelly Gordon: Okay. I think there's another question coming through. "What new changes are coming down the line for LM-80 requirements?"

I don't think we can answer that because we're not on the IES Test Procedures Committee that is considering the standard. That will have to wait until they are ready to issue something, so we cannot speak to that.

Jason Tuenge: Yeah. I think probably the only thing we can bet on is some updates in terms of like terminology, you know, just given that the RP 16 has been changing rapidly. But yeah, otherwise, I have no idea.

Kelly Gordon: And that is something to make people aware of. There is—RP 16 is a document published by the IES that provides terms and definitions for lighting products and lighting products in general, but RP 16 Addendum B was just recently published, which is specifically to do with solid-state lighting terminology.

And if—those of you who saw Addendum A, the previous addition, noted that there were some things that it was pretty hard to differentiate, for example, between an LED module and an LED array and a package. Those things have been now better defined in this new Addendum B. And that document is available, I believe for a free download—

Jason Tuenge: Yeah.

Kelly Gordon: —from the IES website.

Jason Tuenge: And actually, I can provide a link here, I think right on the screen.

Sarah Eckstein: Right. Yeah. I think they don't charge for those addendum documents, but—

Jason Tuenge: But actually, I think they're onto Addendum C. Yes. I'm looking at it right now. They're actually onto Addendum C. And yeah, I think I can show the link here. Just one second. Let's see if can pull this off. I think we can show text. Yeah, we can show text.

Kelly Gordon: If not, we can provide it in the posted version of this, of these documents.

Jason Tuenge: I'm not very good at this. Okay. Here we go. There you go. Oh, of course, it's in green font. So—

Kelly Gordon: Okay. I'm not seeing any other questions come through at this time. I think we'll give it one more minute, until about 10: 00, and then we'll wrap up until tomorrow.

Jason Tuenge: Sorry for the mess.

Kelly Gordon: Okay. This looks like a question for you, Jason. "What optical and fixture improvements will be needed before exterior LED fixtures will be able to compete with metal halide or high-pressure sodium at typical pole-mounting heights or wall applications?"

Jason Tuenge: Oh, there we go. Now I see it. All right. So I think, basically, there are two ways of looking at this. One is what kind of criteria are you looking at? What kind of application? One thing that was becoming clear early on was that in parking lots where you shoot for minimum levels, so like how dark is your darkest point, not the average level throughout your lot, just that darkest point, that's your concern with parking applications, LEDs can—already had kind of a competitive edge because they can provide better uniformity, and like much better uniformity efficiently, so that they're not wasting light, but they're making it a lot more uniform when it's delivered to the target.

By doing that, you basically cheat a little bit. You can reduce your average, and that's viewed as okay. We've demonstrated that in a number of projects where it looks great, and you're still meeting IES recommendations. So that's one case where LED has kind of had a quick advantage, so long as you weren't up against really high-wattage products.

So if you're replacing a 1,000-watt product and you're trying to do it with a single head or, you know, even with multiple heads, and if that 1,000-watt, you know, metal halide lamp was, say, vertical orientation and providing good uniformity already in the parking lot, then you're not going to be able to cheat too much in terms of improving uniformity because it was already good.

And so then what you're looking at is a lot of LEDs installed, you know, either in one head or in multiple heads on the pole, and that means a lot of cost. You know, if you really are forced to produce the same amount of output, the same lumen package as HID and truly replace it one for one, maintaining the average more or less, then you're gonna have an expensive LED product.

And so it's not so much a performance issue as it is, you know—necessarily—as it is a cost one. And HID efficacy tends to improve as the wattage goes up. The lowest wattage lamp ballast combinations for HID tend to be the lowest performance in terms of efficacy, so that's the other thing you're up against, but really cost ends up, you know, stopping a lot of these projects.

So then you look over a roadway, and you don't find a lot of 1,000-watt on roadway, right? But the difference there is that you're not looking at minimum—you know, the minimum point on a roadway. What you're concerned about with roadways is maintaining that average level. Usually, they're not over lit. And, you know, yeah, you can improve uniformity, but does IES care? No. At least as long as uniformity was already okay.

And so you really can't cheat by reducing your average in the case of roadways, but you're typically dealing with lower wattages. So LED can compete in roadways, as we've seen. You know, as far as more progress, I'm not sure how much more is gonna be accomplished by improvements in optics. I think at this point, it's mostly a matter of cost. And, of course, you know, improved efficacy also helps, and we're still seeing, you know, improvements.

Kelly Gordon: Okay. Another question has come in. It says, "ENERGY STAR will be coming out with its list of approved integral LEDs September 1. Can you speak to the number of LEDs that may be approved? I have heard no products have even been submitted."

Yeah, I think this is a little bit of a misunderstanding. The ENERGY STAR criteria for integral LED lamps goes into effect as of August 31. And what that means, those criteria were published, you know, it'll be nine months prior to that, 270 days prior to that date they're published, but everybody knows what they—what the requirements are and people can test their products and start to qualify them.

The August 31 date is the date after which products may be qualified and carry the label. Now, that doesn't mean that there will be a published list as of September 1. It just means that that's when products can start to carry the label. So from what I've heard, many products are in testing and will start to get that approval, you know, sometime in the fall once that goes into effect.

Okay. Any other questions? I'm not seeing any more. I think we're about ready to wrap up until tomorrow.

Jason Tuenge: Yeah. We're cutting it a little bit short. I think we had it allotted for a little bit longer, but yeah, if there's no more questions, we've probably gotten through the bulk of what would be out there. So yeah, tomorrow we'll see some good implementation of all that we've been discussing these last couple of days, and so I think it'll be very satisfying for everybody to see this actually put to use, to help see how you actually wade through all.