Text-Alternative Version: Walking Through Examples of Real LM-79 & LM-80 Reports

Below is the text-alternative version of Day 3 of the three-part TINSSL Webinar Series "Understanding and Evaluating LM-79 & LM-80 Reports for SSL Technology," held August 12, 2010.

Sarah Eckstein: Thank you to those that are participating in today's webcast as well as to those that have stuck around for Days 1 and 2. Lastly, I'd like to give a special thank you to NEEP sponsors and all the DLC members for providing feedback to both NEEP and PNNL on helping to frame the content for this TINSSL webinar series, so thank you.

So today's webinar is walking through examples of real LM-79 and LM-80 reports. For those of you joining this three-part TINSSL webinar series for the first time, I will give a brief overview of who NEEP is, what NEEP provides through the Technical Information Network for Solid-State Lighting, and what the DesignLightsTM Consortium for Qualified Products List is in case Kelly Gordon references it in today's presentation.

So real quick, 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. TINSSL, which stands for the Technical Information Network for Solid-State Lighting, 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 educational webinars such as this one.

So real quick, I just also wanted to mention that NEEP is actively involved in a bi-national effort, which is the DesignLights Consortium for Qualified Products List project. NEEP's role is overseeing this project and collaborates with the DLC members, who roughly cover the territory of 28 states and 2 Canadian provinces. Furthermore, NEEP coordinates with a third-party contractor for this project, and the contractor verifies and evaluates LM-79 and LM-80 test reports submitted through the DesignLights Consortium process. And one of the main goals for this project is building a prescriptive list of LED commercial products, those that essentially can judge the basis for weeding out the junk, essentially, and recognizing those that are of high quality and highly energy efficient.

So without further ado, I would like to announce our main presenter for today's webcast, which is Kelly Gordon and assisting her with today's Q & A portion of the webcast is Jason Tuenge. So I hand it off to Kelly Gordon to now present.

Kelly Gordon: Okay. Thank you very much, Sarah. I'll get started, and thanks again to those of you who are joining us for the third day or if this is your first or second day, welcome as well. The previous two days, Jason Tuenge has walked us through the LM-79 test procedure, the in situ temperature measurement test and the LM-80 test procedure, so we've seen, you know, what is required by each of those standards and test procedures, what is the information that we can expect to see, and how do you find it in the test report. And what we're going to do today is look at several real-life examples, real submissions of products that have been submitted for approval by utility programs, by the DesignLights Consortium, by the ENERGY STAR program. We've pulled examples from different places, and we're going to look fairly in depth at the information that is provided so that we can see, well, is what you get in reality what comports with the checklists that we've been talking about.

Okay, I'm going to try to advance the slide. Here we go. Okay. The first sample submittal, and what I'll show you first is kind of the overview of what was received. So there were product information sheets. This might be called a cut sheet or a specification sheet for the product. And I will note that we are not identifying the example products by manufacturer name or by model number during this presentation. In some cases you may—a product may look familiar to you, but we're not identifying them by name. These are used as examples.

So first the product information sheet. Then there were a couple of test reports included in this submission, and we're going to look at each of these more closely. There was lumen maintenance information provided. This one as well. Okay. So when you get a packet of information like this, and if you're working for a utility or you're otherwise trying to evaluate products, you know, where do you start?

So I think what we're going to do here is look at the checklist that Jason provided and presented during the past two days. We start with those checklists and work through the information that we have in front of us, so first of all, in terms of the LM-79 information, this is the photometric information. Is the report produced by a qualified test lab? That's something to check right off the bat. So we can see in this case, the test report—and I'm just going to page back up to that snapshot of it. The LTL test report, that's the one on the left there, and then we look. How do you find out if this is a qualified test lab?

Now, we've talked about how pretty soon there will be NVLAP accredited laboratories through the National Voluntary Laboratory Accreditation Program will have a list of accredited labs. We haven't had that so far. The place to look now is on the DOE SSL website under the CALiPER program. Look for laboratories that have been performing testing for CALiPER and have been qualified by DOE as being capable of providing LM-79 testing. So we look on the link that's provided there, laboratories performing integrating sphere testing for CALiPER.

Okay. We check that list. We see that Luminaire Testing Laboratory, LTL, is on that list. Laboratories performing goniophotometry testing for CALiPER, also on that list. And it's important to see that both of them are there because these are two different tests that are part of LM-79, but they do different things. And not everyone has a goniophotometer, that is a specialized piece of equipment, and specialized knowledge on how to operate that equipment, so it's important to see that they have both if they're providing you that information.

Okay. Moving on with the checklist. We want to check that data manipulation by the supplier appears unlikely. Now, how do we try to evaluate? Well, one is, you know, is it a PDF file and is it on test lab letterhead? Well, in this case, it was. Both of those things are checked off.

Next, is the information consistent with a submitted product and cut sheet? Now, as I said, we're not identifying the products during this webinar, but you would check is the catalog number complete and unique, do they match, and is there—on the test report, is there any text or any references to the product that are not clarified by the cut sheet. So something that you wouldn't be able to see consistent information on the cut sheet, that's the next thing you would check. So let's take a look at an example here.

In terms of the product description and consistency between the information you get on the manufacturer cut sheet and on the test report. Let's look at the example of input current, voltage, and wattage. So the small box on the left is the information that was on the manufacturer cut sheet, and you see they show energy data. They show the input power as 18 watts, power factor greater than .92, input current is identified as .3 amps. Okay. So if we were to multiply that out, 120 volt power times .3 amp times .92 power factor, that would give us 33 watts instead of the 18 watts, so something is going on there.

Let's look over at the test report on the right, so that's the box on the right, the zoom in. We see that they show input current as .1561 amps, and they show luminaire watts as 18.57, so there the wattage seems to comport with what the manufacturer had said, around 18 watts. It's that input current that is different. So what is going on here?

Well, did the manufacturer mean—when they said .3 amps, did they mean the rated drive current to the LEDs? Those of you who were on the line the past two days know that we discussed how the LED drive current is different than the input current to the luminaire or to the driver, so I think this is what's going on in this particular report because 300 milliamps or .3 amps is a common drive current for LEDs. So I think that's what happened with that manufacturer. Whereas on the test report, we see the input current to the luminaire is .1561, and that makes sense with the wattage of the fixture.

Okay. So the next item in terms of product description, the number of LEDs here. On the left, what the manufacturer said, equipped with seven three-watt white LEDs. Okay. That seems to match with the test report. It showed seven white LEDs.

Correlated color temperature, the manufacturer indicated 3,000K. That matches with what's on the test report with what was measured in the integrating sphere—3,051—which is within the tolerance for the 3,000K nominal.

Okay. Let's look at some of the other product description issues. "No thermal management" not indicated on the cut sheet, so we didn't see any evidence that, for example, in the example we looked at a couple of days ago where an auxiliary fan had been attached to a product during the testing. You know, that's something that wasn't included with the product and wasn't identified on the product sheet, but it showed up in the test report. We don't see anything like that on this one.

In terms of the optical distribution and the optical accessories, this is something to look at. So the information there in the top box next to the photo of the luminaire is from the manufacturer. It says "glass filters" and then off to the right it identifies various colors of filters, so you have pink, red, blue, amber, green filters that are available as options with this fixture. Now, that's important to know—which one are they testing because, as you can imagine, putting a colored filter over a luminaire is going to affect its light output and its efficacy.

So let's check the test report. What did they actually test? It indicates clear plastic optics, so this was the clear filter that they tested. And so in this case the results that you have on this test report apply only to that situation where it is a clear filter. It's not any of the colored filters, and so that's important in terms of looking at any product grouping that the manufacturer may be proposing to do. If they are going to group various products together, they need to test the worst case situation and show that it can still meet the efficacy and output requirement with that worst case optical assembly.

Okay. Continuing with the checklist, let's look at whether there are any apparent errors attributable to manual entry or mathematical errors. Something to check is power factors, so on the test report they state the power factor's .991. We can multiply that out pretty simply, so 18.57 watts divided by 120 volts divided by the amperage, the input current to the luminaire, again that .1561. That pencils out as the correct power factor. Efficacy, we look at the total lumens from the fixture. Luminous flux, 774 in this case, divided by wattage of the fixture, 18.57, 41.7 lumens per watt, so that all looks correct.

So now with the LM-79 information that was provided with this fixture entry. Now let's go onto the in situ temperature measurement test checklist and see what information was provided there. So we're looking, again, for a report that's produced by a qualified test lab, and in this case this can be a CALiPER-qualified lab that is qualified to do LM-79. It can be a laboratory that is a nationally recognized testing lab under the OSHA program or it can be in the UL data acceptance program.

Now, in this case this test report that is provided here and shown here does not indicate the name of the laboratory, so we're not able to confirm that, so that would be the first thing that would be missing, is it just says, "Thermal test." It doesn't indicate who the test laboratory is. It is in a PDF form, but again, it's not on letterhead, so that would be something to ask for.

Continuing with the checklist. Again, you would check the catalog number against what is in the product information and what is in the LM-79 report. You want to make sure that each of these reports that's being submitted is actually referencing the same product. That seems simple, but that can be the source of many problems if you don't notice that it's actually referencing a different product. Again, we're not identifying the products here, but that is definitely something to check.

Let's look at the current and input. Again, we have the issue with—evidently they are referring to the drive current for the LEDs instead of input current to the luminaire, but the electrical data on the ISTMT report does comport with what was on the LM-79 report, so the input current to the luminaire is consistent.

Okay. So we're gonna come back to some of that information in the in situ temperature measurement test in a moment, but let's take a look at some of the lumen maintenance data that was provided here. So again, you look for, you know, where did the report come from and is that identified? Is it a qualified test lab, if applicable, or is it a manufacturer test lab?

You also want to look at product photos. That helps if they're available so that you can check that this is the same—appears to be the same LED package. Product description in terms of the CCT and the drive current, and there are a couple of issues here. So recall that the product information indicated this product was a 3,000K. That would be more of a warm white type product. The lumen maintenance information that was provided, the several graphs, it was a whole packet of graphs that was provided, most of which referred to cool white results. There was one that referred to warm white, so in this case since it's a 3,000K product, we would go with the warm white results and then look at the drive current to the LEDs. Again, we're assuming that that 300 milliamps referenced in the product information applies to the LEDs, and so look at the chart on the warm white results. It does have values for a 350 milliamp current. So if the product is being driven at 300 milliamps, you could use that 350 milliamp data because the actual drive current to the LEDs in the luminaire is lower than 350.

Okay. Now, the LM-80 checklist, as you saw yesterday, is quite extensive, so this is what Jason provided you yesterday. This is based on, in the LM-80 published standard, the information that is expected to be reported. Now, you can see that with this information that was submitted in this particular case, where it's just a graph, very little of this information is actually reflected here. But just to review quickly, you're basically looking for three case temperatures, and LM-80 dictates 55 C, 85 C, and 1 other that is selected by the manufacturer. They do indicate the tolerance there to negative—it could be as much as two degrees below those target temperatures.

Ambient temperature around the LED packages during testing is within negative 5 degrees of that case temperature, and then there are restrictions for the temperatures during the photometric measurements. Then you're at 25 C and the tolerances are identified. Relative humidity needs to be less than 65 percent. I'm not gonna go through all of these parts of the checklist, but these are the other items that are identified in the LM-80 standard.

Okay. So what did we get with this information in this particular product submission? We can see that there was 6,000 hours of testing on that graph. We don't know how frequently they took that information. They show it on the graph as every 1,000 hours. They're not reporting the chromaticity shift, and they show data for 350 milliamps and for 700 milliamps. Ambient temperature is indicated to be 25 degrees C, and so it's not within 5 degrees of the case temperature.

Okay. So let's tie this information together with the Qualified Products List requirements. This is a good framework to use in looking at, well, what do we need to check off here? So some of this is LM-79 related information and some is LM-80. So let's look at minimum light output. This is the DLC requirement for track or monopoint directional lighting fixtures. Minimum light output is 250 lumens, so in this case the LM-79 information indicates they certainly meet that with 774 lumens.

The next requirement is the zonal lumen requirement, so where is the light going? The requirement is greater than or equal to 85 percent of the total light output needs to be in the 0 to 90 degree zone, so it needs to be below the horizontal. No information was submitted with this particular product that substantiated that. There was no goniophotometric part of the LM-79 test report. They only provided an integrating sphere report, so we are not able to evaluate that particular requirement.

In terms of minimum luminaire efficacy, they met that. They were above 30 lumens per watt. On the allowable CCTs, they met that requirement with about 3,000K. They were within the requirements to find in the ANSI chromaticity standard for that nominal CCT. They exceeded the minimum color rendering index requirement.

And then let's look at minimum LED lumen maintenance at 6,000 hours. The DLC requirement for this particular product category is 95.8 percent at 6,000 hours. Again, here we have an issue with the lumen maintenance data that was provided, so the drive current reported in the LM-79 report, is it less than what was indicated in the in situ temperature measurement test and less than the LM-80 measured drive current? Again, the data that was provided is not complete, but it indicates that the drive current was 350 milliamps, and as reported by the manufacturer, it appears the LEDs are being driven at 300, so that's okay.

So let's look at the temperature data—the temperature of the LEDs in the luminaire, so this was that temperature report. It didn't have a laboratory identified, but it did have some information on it. The LED board temperature evidently measured at 51 degrees C. The data that was provided in the graph for lumen maintenance for the 350 milliamp case was—had a board temperature of 42, so in this case the measured temperature exceeded what was provided in the lumen maintenance data, so we would not be able to use that information. So therefore, we do not have enough information to qualify them on the minimum lumen maintenance requirement.

Okay. So that's the first example. Let's look at another example. This is for an outdoor lighting product. Looks like a street lighting product, and this is the information that was submitted. Okay. Let's look first. Again, we check is this a qualified test laboratory? This is Intertek in New York. Checking the CALiPER list. Intertek appears on both the integrating sphere and the goniophotometry test lab lists, so that looks fine.

It's on a PDF. It's on test lab's letterhead. Again, you would check the catalog numbers, which we're not doing here, but that's something you would do right off the bat. And then let's look at this issue of the input current and voltage and wattage again. So we have an LED drive current, 315 milliamps. That is associated with 65-watt power consumption, and power factor greater than .95 is what's stated here. The test report indicates, yep, the input wattage is 65.67. Power factor was .9895. Okay. So that looks consistent.

Correlated color temperature was indicated to be 6,000K, and the test report indicates 5,855. That's within the tolerances for nominal 6,000K, so that looks fine.

Let's look at the optics. Appears to be worst case, so the optics indicated is IES type two. Type two, dark sky compliant full cutoff fixture, and that probably would be the worst case if it is at full cutoff. So that looks fine.

Checking the errors that could be attributable to math mistakes or manual entry. Again, the power factor as stated does pencil out with the—if you divide out wattage by voltage and by the input current to the luminaire. Efficacy appears to be correct, lumens divided by watts.

Okay. And then moving on to the in situ temperature measurement test. This also was provided by Intertek in this case, and in addition to being on the CALiPER list, they are a testing laboratory recognized as a—a nationally recognized testing lab by OSHA, and there is a website that you can check on the OSHA site to check the list of those laboratories.

The information on the in situ temperature measurement test does match what was provided on the LM-79 report, as you would expect since it's the same laboratory. That makes it a little bit easier. The drive current, looking at—oh, now turning to the LM-80 information, the lumen maintenance data. So again, the drive current that was listed in the product information is 315 milliamps. That is below the drive current provided in the LM-80 data that was provided with this product, so that is fine. It's less than or equal to the drive current in the lumen maintenance data.

Okay. The other thing we check with LM-80 is, you know, is the LM-80 data applicable to the correlated color temperature of the LEDs that we're looking at? In this case it's 6,000K cool white. The nominal CCTs are not explicitly identified in this LM-80 test report. It does indicate cool white, but doesn't specifically list the nominal CCTs that that applies to, so that is something to confirm.

Let's look a little bit at the temperature here. So the LM-80 data report, as shown in the text box at the bottom here, indicates in all cases that the temperature—or the case temperature and the ambient temperature meter exceed LM-80 limits, so they're saying that the case temperature is greater than or equal to 55 C and 85 C and ambient temperature is no less than negative 5 degrees—or five degrees—below the case temperature. So in that case—in the LM-80 standard they list a tolerance of negative 2 degrees C around to that 55 and 85, another target temperature. In this case it's greater than—it's on the positive side. It's greater than two degrees, but it's not more than two degrees below. So that would actually be more of a—conceivably a more conservative case to say greater than or equal to.

Okay. And relative humidity is indeed less than 65 percent. We do not have information about the voltage or the total harmonic distortion. That was not provided. And in this case there was not a diagram showing the location of the thermocouple attachment point. That would be useful to have so you could see on the luminaire where was the thermocouple attached to measure the LED in situ temperature.

The LM-80 report also indicated, as shown in the text box there, how they deal with light source failures during the test period. It says they did occur in some data sets but are not included in the attached tables. Failures may have been caused by manual handling and root cause investigations are ongoing, so that is their statement about light source failures during the testing.

Let's look at a little snapshot of these data tables. Is there an adequate number and duration of photometric measurements? It looks like the measurements have been taken. Certainly the 6,000 hours, in this case up to 9,000 hours. Has to be at least every 1,000 hours along the way. At the beginning they tested more frequently— 24 hours, 168 hours, 500, and then it went to every 1,000 hours. And the chromaticity shift was also reported, so the data shown under the red bar here is lumens, under the light green bar at the bottom that is the delta UVs, so this is the change in chromaticity over time. So for that 6,000 hours, the change in chromaticity was .0017. The ENERGY STAR threshold is .004, so this would have met that requirement.

And also what's indicated here on this data set is what test condition is this, so it says stress. They mean what is the stress test condition, 55C, 350 milliamps, and then they show the actual case and ambient temperatures and the relative humidity. So then this way you can check that you're looking at data sets for the same condition. There is also a data set dealing with the forward voltage that, again, you check at the top that it has that same condition, 55 C, 350 milliamps, so that you're looking at data for the same conditions.

Constant current and, again, you have to be comparing information for different temperatures. You need to be comparing that the current is the same, and that has to be held constant in order to be able to interpolate. Otherwise you cannot interpolate across data sets that would involve different drive currents.

Okay. So how does this information relate back to the DesignLights Consortium list as an example of how we evaluate products? The minimum light output in this case, this is for outdoor pole arm-mounted area and roadway luminaires. Minimum light output is 1,000 lumens. They seem to have met that according to their test report. The zonal lumen requirement, in this case they did provide goniophotometric information that allowed them to have a zonal lumen table that indicates that they did meet the requirement for 100 percent of the light to be in the 0 to 90 zone and that less than 10 percent of the light is in the 80 to 90 degree zone, so just below the horizontal. So we can confirm that information here.

We look under zone, 80 to 90. Over at percent luminaire, that means the percent of total lumens out of the luminaire, was only .2 percent is in that zone, so that looks fine. Luminaire efficacy had to be at least 50 lumens per watt, so they've met that, 53 lumens per watt. The allowable CCTs was less than 6,500K, so they've met that requirement.

They met the minimum CRI requirement, and if we look at the lumen maintenance, let's look again if we have everything here for lumen maintenance. Drive current, so the LM-79 reported drive current was less than or equal to that reported in the ISTMT report and less than or equal to what is in the LM-80 report. So the 315 is less than 350 milliamps. And is the temperature measured—the in situ temperature of the LED, does that show that it's less than the LM-80 data would indicate? Okay. The LM-80 data or the measured in situ temperature was 37.8 according to the test report, so that is less than the temperature that was shown in the LM-80 data, which was 55 C.

They show that they can meet the lumen maintenance requirements at 55 C, so they should be able to meet it with the lower case temperature as measured in the luminaire. So if we were to look at this lumen maintenance information on the chart where they had the tree graph, they would use the 55 C case for 350 milliamps, and it looks like that did meet the requirements for at least 95.8 percent lumen maintenance at 6,000 hours. So in that case they would have met that requirement as well. Now, some notes regarding this one, and this is some common issues with product submissions, the lumen maintenance data that was submitted did not strictly adhere to LM-80 in terms of the temperature location. Case temperature was not indicated in the long-term report. The temperature tolerances were not specifically per LM-80, but they were not less conservative but more conservative, so they were in the right direction.

The in situ LED drive current is not rated, as we discussed the last couple of days. It is rated. It's not measured, so that is something that is stated by the manufacturers. What is the drive current that is being applied to the LEDs? That is not measured, and the make and model of the LED package or module is not independently verified. Again, that is something that is stated by the manufacturer.

Just another quick example that you may see. In this case, this is another LM-80 report. Statement of the lumen maintenance at 6,000 hours. They show the ambient condition being 25 C, 30 percent relative humidity, stagnant air. So the question that would arise here is what does that mean, that ambient temperature? Is that long-term ambient temperature being 25 C or is this just during the photometric testing?

When LED packages are measured long term, they're supposed to be in a situation where the ambient temperature is held very close to the case temperature. Now, every 1,000 hours at least when they're taken out to be measured, then they can be—the photometric measurements are done at 25 C ambient, but during that aging, during that long-term run process, they need to be at the temperature that is consistent with the case temperature. So that would be an issue to clarify here with regard to this data. Okay. I believe we are at the end of our examples here, and I think we have some questions coming through. So I'm going to make sure that Jason is here on the line, and we can start answering questions. Jason?

Jason Tuenge:   Can you hear me all right?

Kelly Gordon: Yes.

Jason Tuenge: Excellent.

Kelly Gordon: Okay. Good. So I'll just read off the first question here. It's, "Could you please clarify the difference in nominal and actual input current? Is the difference a loss to the system? What is the true total wattage of the fixture in the example?" Do you want to take a crack at that, Jason?

Jason Tuenge: Sure. Yeah, and I'm not sure which example has been referenced here. I guess we can probably touch on both.

Kelly Gordon: Yeah, I think it was the first one, actually.

Jason Tuenge: Okay. So I might skip back to that slide and look at it. Oh, there's a good guess. Okay. That's probably it. So, basically—as with anything else when you're talking rated versus measured, I mean I guess the one exception in this case is that for LM-80, the manufacturer does indeed measure the drive current going through the product. But at least looking at LM-79 and cut sheets and maybe even ISTMT, when you're talking rated versus measured and just looking at the input current, it's like anything else that's rated. So if the manufacturer rates the luminous output or the CRI, you'll see that some manufacturers seem hopeful. They just put something in there and then when you actually measure it, it ends up being something else, and really this is what we reveal over and over again with the CALiPER program.

And so, you know, really in these cases it's the measured that is gonna have more meaning, you know, most likely. Now, this is one test. Just like in the case of CALiPER, often you've only got two samples being tested, and so, you know, you're not necessarily gonna capture the full range of what that product will do. You know, if you're to test 30 of their products you might see a broader range of data points.

But yeah, I mean basically, what's measured is what's measured, and in terms of reviewing submittals, you have little choice but to go with measured. Don't look at the rate. The main thing is just to, at least this stage in the game, see if they're consistent. You know, is there something wrong here? That's kind of the point of these first steps in the checklist is to, before you even decide whether you're gonna pass the product or not, decide whether you can trust the data you've got to work with. And so this is one of the things, as we're seeing here, where they can get tripped up.

Kelly Gordon: Right.

Jason Tuenge: Something's wrong.

Kelly Gordon: Right. And it's also important to distinguish between input current to the luminaire, again, and that's what's measured in LM-79, and drive current to the LED. So the luminaire has input current going to the driver. The driver is what determines the drive current to the LEDs.

Jason Tuenge: Yeah.

Kelly Gordon: Those are two different things, so what affects the wattage of the luminaire is that input current to the luminaire.

Jason Tuenge: And some cases things get further complicated by having multiple drivers in a luminaire. Some of these, you know, pole-mounted products, for example, might have a few drivers in there, and so then, yeah, you're talking about the output of each of the drivers and the input to all three combined. So yeah, it gets really confusing, and that's why you need to pay close attention to, you know, what you're dealing with and at the same time recognize that at some point you end up having to take the word of the manufacturer in terms of the drive current.

Kelly Gordon: The LED drive current, right?

Jason Tuenge: Yeah. By the time it gets installed in the luminaire or the lamp. Either that or be prepared to review a lot of documentation, if the manufacturer will share it with you, in terms of how these things are circuited. I think you can imagine it would get really difficult and confusing, so just want to at least flag that that, this is a real question mark for the whole process.

Kelly Gordon: Okay. I'm going to the next question here. "What effect does driver frequency have on efficiency and light quality?" I don't know if I can answer that. Can you?

Jason Tuenge: Yeah, that's—I actually had emailed off to one of our colleagues.

Kelly Gordon: Okay.

Jason Tuenge: And have not gotten a response yet.

Kelly Gordon: Okay. So we can come back.

Jason Tuenge: Yeah, we might loop back on this one, but in the meantime what I would say is that basically, what I would throw out there is that in the world of ballasts or other source technologies, we've seen the electronic ballasts, high-frequency ballasts have really taken over. And you know, it's largely because they're more efficient. The question though is, you know, how much of that's due to reduced losses in the circuit versus, you know, the improvements in the wave form. And that's why I was trying to get some clarification if we can get it quickly. Otherwise, you know, that's gonna have to be kind of a question mark, again, on exactly how they're achieving this.

Kelly Gordon: Okay. The next question. Is it true that LED fixtures for parking garages are not required to be ENERGY STAR per DOE? Okay. So at this point, parking garage fixtures are not included in the ENERGY STAR for solid-state lighting luminaires. If you've reviewed that spec, you'll see that there are separate requirements by product type and by lighting application, and at this point the parking garage fixtures are not included in that spec. The DesignLights Consortium does have a specification for parking garage fixtures, and they have specific requirements for that category of fixtures.

Jason Tuenge: And it's worth noting that, you know, nothing is required to be ENERGY STAR—

Kelly Gordon: Right.

Jason Tuenge: —by DOE, so that's just—I mean that could be implemented somewhere as being mandated, but ENERGY STAR is intended to be voluntary.

Kelly Gordon: I'm just reading through the next question. It says, "There seems to be too much inconsistency in testing and too much policing required by the reviewer to ensure test procedures are being followed. Are there any efforts to make adherence to proper testing procedures more rigid? What can we do to push in this direction?" 

Okay. This is a good question, and I think many of you have stuck with us for three days here and seen that it's not necessarily a simple or straightforward process, and that is absolutely true. And I think this has to do with the stage of development that we're in with the LED lighting industry. It's new test procedures, new technologies, things are changing rapidly, and that really means that things are not standard and commoditized and predictable as they are to a greater extent with the older lighting technologies. So I think everyone recognizes that this is an issue, and efforts like ENERGY STAR and the DesignLights Consortium process, I think, are—and the Lighting Facts® program as well—are all efforts to help move the industry toward greater standardization and commonality of terms in the way things are presented. So I think that's absolutely a goal. Efforts are underway, and it's going to take some time.

Jason Tuenge: Yeah, and that's really what all these standards that we're working on are for, is to help give at least some clarity to all this. And there're always gonna be loopholes, and we'll try to close 'em up as we go along and identify the loopholes. But yeah, I mean there's—everybody involved knows that, you know, we're gonna go through some growing pains.

Kelly Gordon: Right. Some of it's good too. You don't want to quash innovation.

Jason Tuenge: Yeah.

Kelly Gordon: And one of the exciting things with LEDs is that, you know, the industry keeps coming up with new and great things, different ways to use the technology. And to the extent that that improves lighting quality and improves energy efficiency and lighting delivery, we absolutely want to encourage that, but it does present some challenges for standardization and harmonization.

Jason Tuenge: Yeah, and as far as the policing goes, I mean when you look at other source types, a lot of that policing is just by recognition of a brand name and then those brand names checking each other. And right now, you know, things are just developing so rapidly. We don't have necessarily this established, you know, big brand names, and you know, things are still settling out.

Kelly Gordon: Okay. Another question is, "Could you discuss how integral lamps are tested for lumen maintenance per ENERGY STAR requirements? This is not detailed in the current version 1.1." 

Okay. So the ENERGY STAR for LED integral lamps has a requirement for at least 6,000 hours of full lamp testing. Now, in the past couple of days we've been talking about LM-80 data for LED packages, and that applies mainly to luminaires because you can use that data for the LED package. You can use the in situ temperature that's measured in a luminaire to make sense of the lumen maintenance data for the packages.

Everything gets a little bit more complicated when you start talking about integral lamps because you're putting the LEDs and all the electronics and everything into a pretty small package. You're trying to match the size of an A19 lamp or an MR16 lamp or a PAR, and it's really cramming things down. So what the ENERGY STAR program has done for those particular class of products is to say, "Look, we need to see how the whole lamp is going to do over time." There are many different components in there that could affect the long-term performance, so the proof is gonna be in the pudding. You know, does the whole lamp survive that long and does it perform well?

So the requirement is 10 samples that must be tested for at least 6,000 hours, so if you want to claim 25,000-hour L70 light, you have to have at least 6,000 hours of testing. If you want to claim a longer lifetime, there are longer test periods defined. So another thing is that lamps that are less than 10 watts are tested at 25 C ambient. Lamps that are 10 watts and above have to be tested in an elevated temperature environment, so 45 C, and that is because many light bulb replacements, as you know, will go into insulated ceiling, recessed downlights, or totally enclosed fixtures, and they will be subject to higher temperatures inside the fixture. So the long-term testing will be conducted at an elevated temperature.

Jason Tuenge: I guess one thing that was tacked onto the end of the question was it is not detailed in the current version 1.1, but it absolutely is, isn't it? I mean I think—

Kelly Gordon: Well, there's some details, but there are additional details coming out that the program will be issuing soon on greater detail about the specific test procedures.

Jason Tuenge: Okay.

Kelly Gordon: Okay. The next question: "How is the product, group, or family defined for evaluation criteria?" And they reference slide ten.

Okay. So the product grouping. And I'll just say a little bit about this, and then if Sarah wants to chime in with regard to the DLC process, that's fine too.

So the idea here is that some products, some fixtures are offered in multiple different versions. They may have different finishes and different—other components may be slightly different. You know, to avoid a case where each and every luminaire would have to be tested individually, the ENERGY STAR program allowed product families or product groups to be identified by the manufacturer.

And then again, the manufacturer, it's incumbent upon them to test the product, the version of the product that would have the most trouble meeting the requirements. So they would have to find like kind of the worst case product, test that, and show that it can meet light output requirements and efficacy requirements. And then all the other versions of that product are in the group would be assumed to meet it. Is there anything additional from the DLC on product grouping, Sarah?

Sarah Eckstein: I'd like to just point as a reference to the DesignLights.org website, 'cause our definitely is up there, and unfortunately I'm less technical. I deal with more the administration side of things. But we are—just to note, we are looking to revise our product family definition. Slightly different than the one that is up there on the website now, so we're currently undergoing some pilot studies with several manufacturers looking to submit 100-plus products, and they have different housing combinations that would accommodate this new product definition that we're looking to develop in the future. So still to come, but just taking note that we're looking on revising our current one.

Jason Tuenge: And I want to point out both ENERGY STAR and DesignLights Consortium were referenced on Day 1, if you look back at slides 18 through 20.

Kelly Gordon: Okay. Okay. The next question: "I saw that ENERGY STAR had an accelerated option of 3,000 hours of testing for integral LED lamps." 

That is true, and the way that works is you have to have it at least 3,000 hours of the long-term, full lamp testing and ultimately you're gonna have to submit the full 6,000 hours. But this is an option that allows you to apply to the program once you've completed 3,000 hours. The condition is you have to have LM-80 data for the LEDs that are used inside the lamp, and you have to provide an in situ temperature measurement test, so this just gets a little bit challenging with integral lamps because you've gotta find a way, through drilling a hole or whatever, to get a thermocouple inside the lamp and attached to the hottest LED in the array in there.

Then you gotta reseal that hole so there's not extra air getting in there. So it's just a more complicated process, but we wanted to provide that as an option for people that do have good solid LM-80 data for the LEDs they're using and they can do that in situ test. Allows them to get qualified slightly earlier, so 6,000 hours of continuous run time takes about eight and a third months, so you know, it would cut that in half potentially. Even with that early qualification, you are still expected to get the full 6,000 hour data. You know, as soon as it's available you have to submit that. Okay, are there additional questions coming in? That's the last one I saw.

Jason Tuenge: Oops, got one more.

Kelly Gordon: Okay.

Jason Tuenge: "Product for testing has a temperature requirement. Why is it important to list LED drive current, assuming it passes all temperature requirements?"

So again, as we mentioned I think yesterday, basically all the manufacturers, at least all the major ones, acknowledge that both the temperature measurement point and drive current are important for lumen maintenance and color maintenance. Not just important but critical.

These are the factors to look at and to control when you're testing. So it is not enough just to look at temperature or drive current, you want to look at both. Beyond this, although there's apparently some debate, ambient temperature is a factor as well. It's pretty clear that, you know, when your product, your luminaire is operating at different ambient temperatures, that's gonna have an effect. It's a little bit of a question apparently during LM-80 testing, and maybe Kelly can provide some verification there. But, you know, basically the drive current and the temperature both impact the—basically characterize how rough you're operating your LED and, therefore, how long it's gonna last.

Kelly Gordon: Right. And you need to know the LED drive current in order to know what LM-80 data set to look at, and whether that lumen maintenance data is relevant and usable for the product that you have.

Okay. I think there was another question. "Does the DLC have a verification process to determine whether a manufacturer is accurately defining product groups and families, or is it just ENERGY STAR?" I'll throw this to Sarah, but I assume that that is something that's checked in the DLC evaluation process.

Sarah Eckstein: Yes, this is Sarah. Essentially it's being evaluated as submittals come in, whether or not a certain product family is accepted or not depending upon how the manufacturer defines it. But the contractor verifies that by comparing what the DLC definition is with what the manufacturer states, and they have a right to turn it down if it's not matching.

Kelly Gordon: Okay. Next question is, "What determines a good product?" 

Obviously that varies by product type and by application, but I think this is where we're trying to get with the ENERGY STAR requirements for different applications and the DesignLights Consortium requirements, is looking at minimum requirements for these different types of products. So we determine that by benchmarking to good conventional technology.

So if you're looking at downlights, for example, you know, what is—you're benchmarking to existing technology. What is commonly used? What meets IES lighting requirements? What are the highest efficacy conventional technologies that are available? You're using all of that to determine, well, what is a benchmark for good performance of an LED product? We've done that with each and every of these different product categories, and then measure or evaluate measured information on LED products to see are they meeting those benchmark requirements. So a good product, that's a subjective term, but that's what we're using.

Jason Tuenge: I was just thinking one thing that might be useful to look at was a presentation we did during—it was one of the tutorial presentations on equivalency. It might be good 'cause that's kind of a qualitative overview of things. I might provide a link here in a second.

Kelly Gordon: Right. We can do the link to that. These are some tutorial presentations we provided in Philadelphia at the DOE SSL workshop.

Okay. Let's see. The next question: "Are there going to be some requirements for drivers in terms of fixture life?" 

This is a very good question and is not currently addressed in the ENERGY STAR or DLC requirements, so it'll be—the way that that life is being addressed in those programs right now is limited to lumen maintenance 'cause that's really what we have a test procedure for. It can be evaluated to some extent, but the overall luminaire life is certainly a product of many different factors, including the driver life. In the ENERGY STAR program, there is a requirement to at least verify the case temperature of the driver. That's one indication that the driver's operating within its design parameters. That can be externally confirmed. However, overall luminaire life, you know, is not something that's addressed at this point.

There are efforts within the industry with industry groups, with DOE groups to address luminaire life and reliability issues. There was recently a white paper published by DOE, and we can provide the link to that one as well, that was published by the Quality Advocates program for the DOE. So, many more issues to be addressed there.

Next question: "Does color temperature correlate to efficacy or longevity for LED lighting?" Correlated color temperature is related to efficacy of the LEDs, and this goes back to the kind of how white light is produced by LEDs today. Most white light LEDs are based on a blue LED, so it emits in a very narrow range of the visible spectrum, in the blue spectrum. And then a phosphor is applied to the LED to convert that blue light to white light, and over time—or no, on the efficacy question, so additional phosphors would have to be applied in order to make that a warm light. Cool light is apparently more blue content, so it's less phosphor that is needed to be applied to that to make it white in the cool range of the white color temperatures, so there's less for it to through and less loss. If you are trying to get a warm white product, it is less efficacious at this point.

Now, in terms of longevity, you know, over time phosphors can degrade, and so the cool white products could tend to have a longer lumen maintenance life than the warm white, but, you know, it depends on the types of phosphors that are used and how the product is used. Do you have anything to add to that, Jason?

Jason Tuenge: No. Sorry, I was just trying to figure out this font.

Kelly Gordon: Okay.

Jason Tuenge: No.

Kelly Gordon: In terms of correlated color temperature and longevity. Okay. That's it.

Jason Tuenge: Yeah, not to longevity.

Kelly Gordon: Yep. Okay. Are there additional questions? Okay. So you've provided some additional links here on equivalency and reliability. Okay. Great. Those are some of the materials from the recent workshop.

Jason Tuenge: Yeah. So I thought the reliability would cover the drivers to some extent.

Kelly Gordon: Great. Okay, I'm not seeing any additional questions at this time. We can give it a few more minutes.

Terry Shoemaker:   Kelly?

Kelly Gordon: Yes.

Terry Shoemaker: We just had one question come in.

Kelly Gordon: Okay.

Terry Shoemaker:    "Do exterior LED fixtures need convection for cooling, and how do they achieve cooling?"

Kelly Gordon: Okay. Jason, you want to take that one?

Jason Tuenge: Well, I guess I would say that you're always gonna have some sort of convective cooling going on, you know, unless the LEDs themselves are sealed in a vacuum or something. You know, there's always gonna be some kind of convective cooling, maybe only on the luminaire itself, not on the LEDs. So I think what they're getting at is active convective cooling, so you know, in other words like a fan or something. For exterior luminaires, basically you're always gonna have heat conducted away and then some kind of air movement somewhere, and it's really gonna be up to the luminaire manufacturer to decide how they want to do it and to the customer in terms of what they consider is reliable.

You know, if you've got an exposed fan somewhere, that's probably gonna be a problem, and if you've got a fan anywhere, you want to be very confident that the thing is gonna not be your first point of failure. Even if you don't have a fan and you've got a good heat sink, it's worth looking at the heat sink to see, you know, is this thing gonna become a bird's nest down the road or will it just collect a lot of dirt, you know, in a dusty environment. So it's definitely worth looking at the design, but unfortunately this is one of those areas that's gonna be pretty gray. It's gonna be very subjective. We don't have standards for, you know, quantifying the quality of a thermal management system, you know, whether it's conductive or convective.

Kelly Gordon: Okay. I don't see another question coming on at this point, so I think we can start to wrap up here. I want to thank everyone for joining in. I know many of you have come in for all three days, and I really appreciate the time, and I hope you will visit the DOE SSL website and the DesignLights Consortium website. To get additional information, please do feel free to contact us if you have follow-up questions. We'd be happy to direct you to additional information as needed, and I believe we will say everyone can disconnect at this point. Thank you very much.