November 28, 2 February 23, 2015 VIA EMAIL TO: GSL2013STD0051@ee.doe.gov Ms. Brenda Edwards US Department of Energy Buildings Technologies Program Mail Stop EE-2J 1000 Independence Ave, SW Washington, DC 20585-0121 NEMA Comments on Energy Conservation Program: Energy Conservation Standards Rulemaking Preliminary Technical Support Document for General Service Lamps Docket Number: EERE-2013-BT-STD-0051 RIN: 1904-AD09 Dear Ms. Edwards, The National Electrical Manufacturers Association (NEMA) appreciates the opportunity to provide the attached comments on the Department of Energy’s Preliminary TSD (pTSD) for Energy Conservation Standards for General Service Lamps. These comments are submitted on behalf of NEMA Light Source Section member companies. As you may know, NEMA is the association of electrical equipment and medical imaging manufacturers, founded in 1926 and headquartered in Arlington, Virginia. The National Electrical Manufacturers Association (NEMA) represents nearly 400 electrical and medical imaging manufacturers. Our combined industries account for more than 400,000 American jobs and more than 7,000 facilities across the U.S. Domestic production exceeds $117 billion per year. Please find our detailed comments below. We look forward to working with you further on this important project. If you have any questions on these comments, please contact Alex Boesenberg of NEMA at 703-841-3268 or alex.boesenberg@nema.org. Sincerely, Kyle Pitsor Vice President, Government Relations National Electrical Manufacturers Association 1300 North 17th Street, Suite 900 - Rosslyn, VA 22209 NEMA Comments on Energy Conservation Program: Energy Conservation Standards Rulemaking Preliminary Technical Support Document for General Service Lamps We are pleased to address the specific issues that the DOE posed to the public in the workshop and in the pTSD. NEMA has overarching concerns that form common themes underlying several of our comments to the specific questions. Specifically, as we discuss in response to ES.3.21 below, this is not the rulemaking that Congress contemplated in section 320 of the Energy Independence and Security Act of 2007 (EISA-2007). As we acknowledge below, NEMA appreciates the explanation provided by DOE that the Consolidated Appropriations Act of 2014 (and similar predecessor legislation) makes it difficult to consider the real baseline in this rulemaking and other issues. But we disagree fundamentally with DOE’s approach to product classes in this rulemaking and the proposal for technology neutral energy conservation standards. NEMA cannot help but conclude that the present congressional appropriations directive has influenced DOE’s selection of this approach in a manner not intended by Congress in EISA-20071. ES.3 ISSUES ON WHICH DOE SEEKS PUBLIC COMMENT DOE is interested in receiving comment on all aspects of this preliminary analysis. DOE especially invites comment or data to improve DOE’s analyses, including information that will respond to the following questions and concerns raised in the development of this preliminary TSD. ES.3.1 GSL Definitions DOE developed definitions for several terms in support of the scope of the rulemaking. DOE requests comment on the definitions under consideration including “integrated lamp,” “nonintegrated lamp,” “general service LED lamp,” “OLED lamp,” “colored lamp,” “reflector lamp,” “non-reflector lamp,” “light fixture,” “pin base lamp,” and “GU24 base.” Integrated lamp means a lamp that contains all components necessary for the starting and stable operation of the lamp, does not include any replaceable or interchangeable parts, and is connected directly to a branch circuit through an ANSI base and corresponding ANSI standard lamp-holder (socket). NEMA Comment: We disagree with the DOE’s proposed definition of Integrated Lamp. The bases on integrated lamps mentioned in the definition should be limited to those bases most commonly used with the lamps covered within the scope rulemaking. Currently these bases would be limited to medium screw bases and GU-24 bases for integrated lamps. If the scope of the regulation changes in the future, this definition can be adjusted at that time. Following is our suggested definition change: “Integrated lamp means a CFL or LED lamp that contains all components necessary for the starting and stable operation of the lamp, does not include any replaceable or interchangeable 1 For a discussion of EISA-2007 provisions pertaining to this rulemaking, see NEMA comments to DOE Framework for GSL section “General Comments” at pages 2-7, http://www.regulations.gov/contentStreamer?objectId=0900006481557fe6&disposition=attachment&contentTyp e=pdf 2 parts, and is intended to be connected directly to a branch circuit through a Medium Screw Base or a GU24 base.” Non-integrated lamp means a lamp that is not an integrated lamp. NEMA Comment: We disagree with the DOE’s proposed definition of non-integrated lamp. Many of the lamps that would be covered by this broad definition are not within the scope of the rulemaking. This definition needs to be limited to the lamps covered by the scope. We suggest the following: “Non-integrated lamp means a lamp that requires additional external components for starting and stable operation of the lamp such as a ballast or a driver and has a single-ended 2-pin or 4pin base.” General service light-emitting diode (LED) lamp means an integrated or non-integrated LED lamp designed for use in general lighting applications (as defined in 430.2). NEMA Comment: We suggest the addition of wording to clarify the use of LEDs in a GSL LED and proposed the words “that uses light emitting diodes as the primary source of light” be added to the end of DOE’s proposed definition: General service light-emitting diode (LED) lamp means an integrated or non-integrated LED lamp designed for use in general lighting applications (as defined in 430.2) that uses light emitting diodes as the primary source of light. Organic light-emitting diode or OLED lamp means an integrated or non-integrated lamp that uses OLEDs as the primary source as light. NEMA Comment: We note a typo in DOE’s proposed definition for OLED lamps and also suggest it be modified for clarity as follows: Organic light-emitting diode or OLED lamp means an integrated or non-integrated lamp designed for use in general lighting applications that uses OLEDs as the primary source of light. Colored lamp means a colored fluorescent lamp, a colored incandescent lamp, or a lamp designed and marketed as a colored lamp and not designed or marketed for general lighting applications with either of the following characteristics (if multiple modes of operation are possible [such as variable CCT], either of the below characteristics must be maintained throughout all modes of operation): (1) A CRI less than 40, as determined according to the method set forth in CIE Publication 13.3 (incorporated by reference; see §430.3); or (2) A correlated color temperature less than 2,200 K or greater than 7,000 K as determined according to the method set forth in IES LM-66 or IES LM-79 as appropriate (incorporated by reference; see §430.3). NEMA Comment: We agree with the proposed definition Reflector lamp means a lamp that has an R, PAR, BPAR, BR, ER, MR, or similar bulb shape as defined in ANSI C78.20 (incorporated by reference; see §430.3) and ANSI C79.1 (incorporated by reference; see §430.3) and is used to direct light. 3 NEMA Comment: We agree with the proposed definition Non-reflector lamp means a lamp that is not a reflector lamp. NEMA Comment: This definition can be deemed vague, and we suggest the following wording in the interest of clarity: Non-reflector lamp means an integrated or non-integrated lamp that is not a reflector lamp. Light Fixture means a complete lighting unit consisting of lamp(s) and ballast(s) (when applicable) together with the parts designed to distribute the light, to position and protect the lamps, and to connect the lamp(s) to the power supply. NEMA Comment: We agree with the proposed definition Hybrid compact fluorescent lamp means a compact fluorescent lamp that incorporates one or more supplemental light sources of different technology. NEMA Comment: DOE’s proposed definition is vague. We suggest the following changes in the interest of clarity: Hybrid compact fluorescent lamp means a compact fluorescent lamp that incorporates one or more supplemental light sources of different technology, such as halogen or LED, which are energized and operated independently and may or may not operate simultaneously. Pin base lamp means a lamp that uses a base type designated as a single pin base or multiple pin base system in Table 1 of ANSI C81.61, Specifications for Electrics Bases (incorporated by reference; see §430.3). NEMA Comment: We agree with the proposed definition GU24 base means the GU24 base standardized in ANSI C81.61 (incorporated by reference; see §430.3). NEMA Comment: We agree with the proposed definition ES.3.2 Specialty Application MBCFLs and LED Lamps DOE identified MBCFLs and LED lamps that are designed for specialty applications and are not able to provide overall illumination including: black light lamps, bug lamps, colored lamps, plant light lamps, and silver bowl lamps. DOE is considering providing exemptions for these specialty applications. DOE requests comment on the MBCFLs identified for specialty applications that cannot provide overall illumination and if there are other MBCFLs that should be considered. DOE requests comment on the LED lamps identified for specialty applications that cannot provide overall illumination and if there are other LED lamps that should be considered. NEMA Comment: We agree that black light lamps, bug lamps, colored lamps, plant light lamps, and silver bowl lamps are designed for specialty applications and do not provide overall illumination. 4 ES.3.3 GSL Exemptions DOE assessed the full list of exemptions that apply to GSILs to determine if the lamp types provide overall illumination and therefore can be used in general lighting applications. Of the exemptions provided for GSILs, DOE has preliminarily determined that appliance lamps, black lights, bug lamps, colored lamps, infrared lamps, marine signal lamps, mine service lamps, plant lights, sign service lamps, silver bowl lamps, showcase lamps, and traffic signal lamps cannot provide overall illumination and therefore cannot be used in general lighting applications. DOE is considering not establishing standards for these lamp types under the GSL rulemaking because the lamps are intended for use in non-general applications. DOE requests comment on the exemptions that DOE is considering providing for GSLs based on its assessment that the lamp types are intended for non-general applications. NEMA Comment: We agree that specialty lamp exemptions are needed for CFL and LED lamps to cover the development of future energy efficient specialty lamps as potential substitutes for incandescent technology in specialty lamps if feasible. We agree with the DOE that: Appliance, Plant, Black Light, Sign Service, Bug, Silver Bowl, Colored, Showcase, Infrared, Traffic signal, Marine Signal and Mine Service lamps should be exempted from efficiency regulations as low volume specialty lamps that are designed for specialty applications and that do not provide overall illumination. They would not provide an acceptable substitute for lamps commonly used in general service lighting applications. DOE has preliminarily determined that 3-way lamps, vibration service lamps, rough service lamps, and shatter-resistant lamps are able to provide overall illumination and therefore can be used in general lighting applications. For these reasons, DOE believes that 3-way lamps, vibration service lamps, rough service lamps, and shatter-resistant lamps are general service lamps and do not require an exemption from standards. DOE requests comment on this decision. NEMA Comment: If the CFL or LED replacement lamps for these traditional incandescent specialty product categories are regulated, special considerations may be required which will vary depending on the specialty product. If 3-way CFL or LED lamps are regulated, the efficiency requirements should be evaluated based on the highest, most energy consuming setting. As 3-way CFL lamps will operate at different efficiencies when operated at different light levels, it is important that DOE base compliance on operating only at the most efficient level, which is the highest light output level. Forcing the lower light output settings to meet high efficiency levels would be very problematic for industry and may remove this product utility from the market. This approach of testing only at the highest level is also used in other current standards (ENERGY STAR) for these products. Shatter resistant lamps normally contain a coating which absorbs a small portion of the light output. Light absorption factors would have to be considered when setting efficiency regulations covering this technology. Rough Service and Vibration Service lamp definitions are unique to incandescent technology and are not applicable to CFL or LED technology because they are more shock resistant by design. 5 During the public meeting, DOE also discussed Left-Hand Thread, Marine, Reflector and Specific lamp shapes. NEMA agrees that Left-hand Thread and Marine exemptions are no longer necessary for CFL or LED technology. NEMA believes that it would be inappropriate to include Reflector lamps as covered products in this rulemaking which is concerned with omnidirectional general service lamps. Lamps with a built-in reflector are designed for very specific applications and offer a unique product class with unique performance and efficiency features. Such lamps are not appropriate replacements for general service omnidirectional lamps and should not be analyzed in the same rulemaking. The specific lamp shapes exempted in the current incandescent rule primarily provide decorative illumination and are not purely functional as with most general service lamps. Decorative lamp shapes provide unique technical challenges for both CFL and LED technology and they cannot be assumed to be capable of reaching similar efficiency levels. The technical effort necessary to mimic the consumer-demanded performance attributes of some decorative products will come with corresponding tradeoffs in efficiency potential. Because manufacturers are only beginning to develop these types of lamps the size of this efficiency impact is not wellknown. Because of these unknowns, NEMA does not recommend regulating decorative LED or CFL lamp shapes at this time because of the lack of information and understanding of the cost and technical impacts of their performance demands. Regulating this emerging product category (primarily B, BA, CA, F, G16-1/2, G25, G30, G40, S and long T Lamps) at this time will slow product innovation and slow the development and consumer acceptance of energy efficient decorative lamps. When a regulatory specification is established – say for energy efficiency – at an early stage of a product’s life, it inhibits the manufacturer’s flexibility to experiment with other product specifications that may relate to consumer acceptance of that product or the product’s utility. As acceptable decorative energy efficient replacement lamps are developed, DOE can further assess this product category in the future. ES.3.4 OLED Lamps DOE requests comment on its consideration to continue to not establish standards for OLED lamps in this rulemaking. NEMA Comment: We agree that it is premature to establish standards for OLED products at this time. Among the reasons for our agreement is our overarching concern with regulating emerging product categories, creating a substantial risk of product slowing innovation and the development and consumer acceptance of this product category. ES.3.5 Lamps Addressed in Other Rulemakings DOE has the authority to consider additional lamp types that it determines are used to satisfy lighting applications traditionally served by GSILs. To limit the probability that one lamp type might be subject to two different standards, DOE is not considering including self-ballasted mercury vapor lamps in the scope of this rulemaking. DOE requests comment on its consideration to exclude from the scope of the GSL rulemaking lamps that are addressed in other rulemakings. NEMA Comment: We agree that self-ballasted mercury vapor lamps should not be included in this rulemaking. ES.3.6 Other Integrated Lamps 6 DOE does not believe that LED technology is currently able to provide the same utility as halogen technology in the MR16 lamp shape. Because more efficient replacements that maintain the same utility are not currently available, DOE has tentatively decided to not establish energy conservation standards for reflector pin base integrated lamps at this time. DOE requests comment on whether LED MR16 lamps are suitable replacements for incandescent/halogen reflector integrated MR16 lamps. NEMA Comment: We agree that current MR16 LED lamps cannot provide all the functionality of currently available Halogen MR16 lamps. As this is a developing product category, MR16 lamps should not be regulated during this rulemaking. DOE also requests comment on whether there are any integrated lamps with other bases than screw bases that have a significant market share. NEMA comment: We believe that the market share for integrated CFLs with GU24 bases is insignificant (less than 4%), and that GU24 based CFL products should be excluded from scope. There are no additional bases used at this time for common general service lamps that have a significant market share. ES.3.7 Other Non-Integrated Lamps Due to low market share and to avoid stifling innovation with LED non-reflector pin base nonintegrated lamps, DOE is not considering establishing standards for these products. DOE requests comment on the market share and technological feasibility of increasing the efficacy of non-reflector pin base non-integrated lamps. NEMA Comment: We do not believe that there will be any innovation in LED non-reflector pin base non-integrated lamps. These lamps are the same wattage as the CFL pin base lamps they look to replace and they are not dimmable (which represents a loss of functionality). Also the lamp-ballast pairings we have investigated do not have UL listing and we consider this to be significant. This lamp is not as simple in practice as a T8 replacement; there are multiple bases, multiple wattages, multiple sizes and multiple ballasts. We believe that if one is going to retrofit one’s CFL pin base lamps, there are more efficacious choices than the LED non-reflector pin base non-integrated lamps. Additionally, compatibility problems with reduced wattage lamps are not well understood in the DOE analysis, and could result in field issues if pursued. DOE identified screw base and pin base non-integrated lamps that meet the definition of GSL. DOE requests comment on whether there are any non-integrated lamps with other bases that meet the definition of GSL and the market share of these lamps. NEMA Comment: There are no non-integrated lamps with other bases that meet the definition of General Service lamp as defined with NEMA’s updated definitions. ES.3.8 MBCFL Metrics DOE has the authority to revise the existing metrics and consider additional metrics for MBCFLs in this rulemaking and requests comment on several topics related to this. Regarding metrics that DOE is not considering revising, DOE requests comment on maintaining the current lumen maintenance requirements. Regarding metrics that DOE is considering revising, DOE requests 7 comment on the rapid cycle stress performance of commercially available MBCFLs and the appropriateness of requiring an increased lifetime of 10,000 hours for MBCFLs. Regarding metrics that DOE is considering adding, DOE requests comment on adding a requirement for power factor and its consideration of a standard for power factor of 0.5 or greater; CRI and its consideration of a standard for CRI of 80 or greater; and the start time of MBCFLs and its consideration to require a start time of within one second of the application of electrical power. Regarding metrics that DOE is not considering adding, DOE requests comment on the consideration to not set a separate requirement for THD; the consideration to not set a CCT requirement for MBCFLs; and the consideration to not set requirements for operating frequency. NEMA Comment: We note that the ENERGY STAR requirements are voluntary standards for premium products and should not be considered as baseline requirements for medium base compact fluorescent lamps. There are energy efficient MBCFL products on the market that do not meet the ENERGY STAR specification. NEMA agrees with DOE that requirements for THD, CCT and operating frequency should not be considered. For compact fluorescent lamps, the current statutory and regulatory requirements for Lumen Maintenance, Rapid Cycle Stress testing and Life are acceptable. Besides minimum energy efficiency requirements (the primary focus of DOE regulations) these requirements were introduced by Congress to address poor quality products that damaged the reputation of CFLs and hindered consumer adoption of CFLs. This was a major issue when EPAct 2005 was enacted. These very poor quality products are now gone. These quality requirements were not introduced and are not intended as a mechanism to set ENERGY STAR-type levels for premium products through DOE energy efficiency regulations. Many of these metrics have nothing to do with energy efficiency. Increasing the Rapid Cycle Stress test to current ENERGY STAR standards is not necessary to set a minimum Energy Efficiency Standard. Increasing the minimum life standard to the ENERGY STAR level of 10,000 hours is also not necessary for energy efficiency standards. If the minimum life is increased, industry would recommend no more than 8000 hours as a national minimum to enter the market. By definition, not all products are intended to meet ENERGY STAR performance levels. Of the remaining items for consideration to add to the regulation, including Power Factor, CRI and Start Time, NEMA believes that adding power factor, CRI, or start time requirements is not necessary. NEMA does not agree with DOE’s assertion (TSD at 2-40) that ”a minimum power factor requirement could decrease energy use”. This conclusion appears to be based on a document that is not relevant to a discussion of general service lamps. In the document referenced in the TSD regarding “Reducing Power Factor Cost” the power factor scenario described is a specific case involving HID lighting systems used in industrial complexes where the lighting system represents a major portion of the power consumed. HID lamp technology is not part of the GSL rulemaking and the power factor associated with an HID lighting system is directly related to the ballast used, not the HID lamp itself. Contrary to DOE’s statement in the TSD, there are trade-offs associated with increasing the power factor in CFL and LED lamps that will reduce lamp efficiency and increase energy use. Unlike the HID system application referenced in the TSD, GSL lamps do not typically represent 8 a ‘major portion’ of the power used. More to the point, in any scenario where CFL or LED lamps are used to replace traditional incandescent lamps the substantially lower wattage of these replacement lamps will result in a reduced lighting load regardless of power factor. Based on these facts DOE should refrain from including a power factor requirement for general service lamps in the final rule. NEMA does not agree that Start Time has any effect on energy efficiency and therefore it should not be included in this standard. NEMA believes that CRI is not an appropriate characteristic for a minimum energy efficiency standard. ES.3.9 GSL Technology Options DOE requests comment on the technology options under consideration for GSLs. Specifically, DOE also requests comment on the addition of reduced current density as a technology option. NEMA Comment: CFL Highly Emissive Electrode Coatings – Has been looked at in the past but overall performance of existing emitters are already designed for energy conservation and long life. Further changes could have a dramatic effect on shortening lamp life. Any potential improvements would be minimal. Higher Efficiency Lamps Fill Gas Composition – Based on current technologies, the limits of gas fill technology has been reached by manufacturers. High Efficiency Phosphors – Current technology has reached its limit. A breakthrough is needed to improve and none is foreseen in the near future. Because of REO availability and cost issues, all coating resources are being used to reduce losses and optimize current technology performance. Glass Coatings – Improvements have been made in this area over the years. A breakthrough is needed to improve – no actions underway that would result in major improvements. Cold Spot Optimization - The effectiveness of any cold spot design is limited by the ambient temperature of a lamp in operation, for example, in fixtures. The cold spot temperature can never be lower than ambient temperature next to the cold spot. This limits the light output gains through cold spot optimization. Active cooling approaches have been studied, but are absolutely cost prohibitive, and may lower efficacy due to the power needs of the active cooling system. Higher Efficiency Diameters – Manufacturers are already producing lamps within the ideal diameters for maximum efficiency. Multi-Photon Phosphors – Has been reviewed in the past and no cost effective improved performance phosphors have been identified. NEMA agrees with DOEs decision to screen out Multi-Photon Phosphors. NEMA agrees that the use of higher grade components could slightly reduce energy loss and that the cost impact of such improvements must be evaluated in determining requirements. NEMA’s position is that for residential ballasts, increasing the power factor will raise ballast losses which will more than offset any gains in distribution efficiency and could have a negative impact on system reliability. In addition, compact fluorescent lamps used in the 9 home have a leading power factor which tends to offset the lagging power factor of motor loads and helps to balance the overall power factor of the home. NEMA is also unaware of any emerging technology that promises to lower ballast losses while maintaining the same level of performance compared to today’s state of the art ballast designs. DOE believes “[t]hese design options are being utilized in commercially available lamps and have demonstrated that they are technologically feasible, practicable to manufacture, install, and service, and do not result in adverse impacts on product utility/availability or health and safety.2” NEMA disagrees. Some of the proposals listed are not practical or are impossible in mass production, or are so cost-prohibitive that the lamps would become very unaffordable. And the technology options that truly are “...technologically feasible, practicable to manufacture...“, are currently in use. Put another way, these ideas are not new, and any that are beneficial to this mature technology are in use today. Revisiting these proposals would not only involve significant R&D, but result in few if any benefits. See our comments on Manufacturer Impact Analysis at page 29, infra. It would be incorrect for DOE to assume that the apparent differences in efficacies are the result of superior design. Those differences are more likely the consequences of differing manufacturing processes employed by different sources of CFL. LED Efficient Down Converters - These are not in use today due to the following: technical challenges surrounding narrow-band phosphors that enable high spectral efficiency include robust packaging for lumen maintenance, while achieving high quantum efficiency under high temperature and flux. Improved Package Architectures - Reliable die attachment methods are needed to enable high temperature operation and there is a need for polymer optical encapsulants to improve color stability and emitter lifetime, and high thermal conductivity to reduce downconverting layer temperatures. Another challenge is the development of high index encapsulants to increase photon extraction. Further challenges may be different depending on the architecture approach; e.g. mixed-color solutions may require additional controls increasing the cost of the package. Improved Emitter Materials - At least 4 colors, including green and yellow, are required to achieve high color quality. There is no clear development path towards efficient direct green and yellow emitters. Also, multi-color modules need more sophisticated drivers to maintain color stability (due to different degradation rates of different colors). NEMA agrees with DOEs decision to screen out this technology. Alternative Substrate Materials - The cost of GaN substrates is high. The performance of Si and GaN-on-Si-based devices is not significantly better than sapphire-based devices and does not warrant a transition to these substrates. Improved Thermal Interface Material (TIM) – Challenges: develop TIMs that enable high efficiency thermal transfer for long-term reliability and performance optimization of the LED device and overall lamp product. Optimized Heat Sink Design - The performance of the heat sink is generally compromised by material cost (e.g. metal vs. plastic) and geometrical constraints (e.g. in retrofit lamps). Active Thermal Management Systems - Reliability and cost are major concerns. 2 pTSD document item 4.4 10 Device Level Optics - The package size limits the extent of beamshaping that can be done with reasonable extraction efficiency. It may also not be desirable to integrate applicationspecific functions at a low system level for complexity management reasons. Increased Light Utilization - This is a trade-off with system level cost. Improved Driver Design – In addition to efficiency, drivers must meet many specifications (cost, power quality, flicker, dimmability, isolation, line regulation, and transient / surge protection). Optimization for specific applications often leads to a compromise in efficiency. AC LEDs - True AC-LEDs (i.e. for bidirectional bias) by definition have < 50% utilization, and often still need external components (for surge protection, flicker mitigation etc.). Challenges for HV-LEDs in general include efficiency loss due to die segmentation, increased die/package complexity to sustain the high voltage, and the wide variety of optimum forward voltages. NEMA agrees with DOEs decision to screen out this technology. Reduced Current Density - Current density is only one aspect in the design of an efficient LED die and there are many tradeoffs that take place to ensure higher efficiency. Optimizing on this one aspect will lead to lower overall efficiency. ES.3.10 GSL Product Classes DOE is considering establishing three product classes for this rulemaking. DOE welcomes comments on the product class divisions it is considering for GSLs in this preliminary analysis. Further, DOE specifically requests comments on a product class division based on lumen package for the integrated GSLs. DOE also requests comments on its preliminary determination that energy consumed in standby mode will be negligible and therefore a product class division based on standby mode operation for integrated GSLs is not warranted. NEMA Comment: NEMA disagrees with the three product classes presented for discussion by the DOE. The overriding determination that must be used to set the product classes is the technical capability and varying utility of differing technological approaches to produce the same light output. Placing all general service lamp products in common lumen ranges will result in the elimination of all technologies and all product utilities except that provided by the most efficacious technology. Therefore determining product classes based only on lumen output is not appropriate for general service lamps. For the variety of reasons discussed elsewhere in our comments, NEMA does not agree that it is good public policy to adopt a technology neutral approach to energy conservation standards for general service lamps under EPCA. This is true for both the general service CFL and LED segments presently under consideration by DOE in this rulemaking, and it is equally true for the halogen incandescent, CFL, and LED classes. In addition, we do not believe that non-integrated pin-base CFLs should be included in the scope of this rulemaking and therefore should not be given a general service lamp product class. Pin-based compact fluorescent lamps are rarely used in residential applications (the primary application for screw-based general service lamps) and cannot directly replace medium screw-based general service lamps without replacing the entire fixture. Few residential fixtures are designed to use pin-based CFLs. Fixtures using these lamp types are nearly all designed for commercial applications. Further, these mature products are already at max tech so there is very little opportunity to increase efficiency. Even if efficiency can be increased, it will generally result in lamps with slightly higher lumen output, not reduced wattage, as the dedicated ballasts will drive them at the same wattage levels. Although there are a few products designed to operate in an energy saver reduced wattage mode, they are only available for a few product types and they typically have similar efficacy; wattage and lumens are similarly reduced. Reduced lumens are not acceptable in many applications. While there are some LED lamp replacements being developed at this time, they generally have identical wattage to non11 integrated pin-base CFLs. These LEDs represent a loss of utility since they do not work with some types of controls and dimming systems. Furthermore, there is no energy savings since there is no reduction in wattage. We ask the DOE to afford the same recognition of the implications of a Lamp rule on non-integrated ballast systems as they did in the General Service Fluorescent Lamp rulemaking3 and our arguments therein regarding lamp wattage versus ballast wattage and there being no change in overall power consumption, just more lumens from a given lamp. The three product classes that are appropriate based on utility and technological capability are 1) Incandescent/Halogen, 2) CFL and 3) LED; that have a medium-screw base and designed to operate directly on 120 or 130 volts. This also represents the overwhelming majority of sockets and applications and provides the biggest opportunity for energy savings. INCANDESCENT/HALOGEN While NEMA recognizes that DOE cannot address the Incandescent/Halogen efficiency class at this time, it should be recognized as a product class with the general service lamp area which is currently regulated. The unique utility and attributes of incandescent or halogen lamp technology are: low initial cost, omnidirectional point source, point source provides very good optical control, can provide high sparkle and high brightness, operation unaffected by high or low ambient temperature, warm color appearance difficult to modify without loss of efficiency, very high CRI, relative low efficiency, relative short life, adversely affected by vibration, adversely affected by direct water contact, life not affected by frequent switching, good infrared source, immediate on to full brightness, great full range dimming in all applications. CFL The unique utility and attributes of CFL lamp technology are: relatively low initial cost, relatively long life, very diffuse omnidirectional source, provides very good general light distribution, very low pleasing surface brightness, operation is affected by high or low ambient temperature, color can be modified with some loss in efficiency at high chromaticity and high CRIs, good CRI capability, relatively high efficiency, vibration resistant, not affected by occasional direct water spray, life shortened by frequent switching, low heat source, natural slow start may be beneficial for dark area eye adaptation, can be dimmed with limitations. The CFL lamp classification should be further sub-divided by Bare-Lamp and Covered-Lamp classes because covered lamps have meaningfully lower efficiency but provide a unique utility in contrast to the Bare-Lamp. . LED LED lamps also provide unique utility vs. CFL lamps and have very different technical efficiency capability which requires them to also be placed in a separate class from the first two classes. The unique utility and attributes of LED lamp technology are: extremely long life, high initial cost, directional point source, extremely high chip surface brightness, requires special optics and diffusing materials for omnidirectional applications, operation is affected by high ambient temperature, operation is not affected by low ambient temperatures, color can be modified with some loss in efficiency at low chromaticity and high CRIs, good CRI capability, very high 3 See page 14 and related: http://www.regulations.gov/contentStreamer?objectId=0900006480f75c3e&disposition=attachment&contentTyp e=pdf 12 efficiency, vibration resistant, not affected by occasional direct water spray, low heat source, instant on, can be dimmed with limitations. Summary of Utility Differences The three technologies offer considerable differences in performance and utility and each provides the best solution for differing applications. There are many differences starting with the most basic construction aspects, filament lamps are low cost omnidirectional point sources, CFL lamps are low cost omnidirectional diffuse sources, and LED lamps are high cost directional point sources. Consumers need to be allowed to choose the best technology for their application and situation. DOE needs to consider separate efficiency levels for each technology to allow each option to remain available. LUMEN RANGE DISCUSSION Efficiency with all technologies decreases with decreasing wattage due to inescapable power losses from components. There are four natural lumen ranges where consumers select products. These are a legacy of the original 40w, 60w, 75w, and 100w traditional incandescent lamps. Lumens ranges for these products are already defined in the existing incandescent regulations: 100w=1490-2600 lumens, 75w=1050-1489 lumens, 60w=750-1049 lumens, 40 watt= 310-749 lumens. The 100 watt and 40 watt ranges are exaggerated on the high end and the low end to extend the regulated product range to just above 25 watt and just below 150w traditional incandescent lamps. As DOE proposed, the higher lumen range can be limited to 2000 lumens for current LED technology. All three product classes should be divided into these four natural lumen ranges. This will provide DOE the opportunity to set efficiency limits that are appropriate for the technologies being sold in each lumen range. This may be necessary as technical limitations and performance can vary greatly depending on the wattage and technology. This approach will allow such limits to be a single LPW or Wattage number, a linear equation, a quadratic equation or an exponential equation as necessary within the range and technology under consideration. This will allow the use of different approaches in different areas considering the widely different technologies under consideration. When complete, such as class matrix would look similar to the following: Lumens Standard Inc./Hal Mod. Spec. Inc. Hal Bare CFL Covered CFL LED 310-749 Wattage Limit 1 Wattage Limit1 .75xLumens LPW no. or equation 1 LPW no. or equation 5 LPW no. or equation 9 750-1049 Wattage Limit 2 Wattage Limit2 .75xLumens LPW no. or equation 2 LPW no. or equation 6 LPW no. or equation 10 13 1050-1489 Wattage Limit 3 Wattage Limit 3 .75 x Lumens LPW no. or equation 3 LPW no. or equation 7 LPW no. or equation 11 1490-2000 2000-2600 Wattage Limit 4 Wattage Limit 4 .75 x Lumens LPW no. or equation 4 LPW no. or equation 8 LPW no. or equation 12 None NEMA agrees that a special modified spectrum class is not needed for CFL or LED technology. NEMA also agrees that standby power for LED products will be minimal compared to impacts of the classifications shown above and would not require a separate class. The existing efficiency levels are as follows: Lumens 310-749 750-1049 1050-1489 1490-2000 2000-2600 Standard 29 w 43 w 53 w 72 w Inc./Hal Mod. Spec. Inc. 29 w 43 w 53 w 72 w Hal .75xLumens .75xLumens .75 x Lumens .75 x Lumens Bare CFL 45 LPW 45/60 LPW* 60 LPW 60 LPW Covered CFL 40 LPW 48 LPW 50 LPW 55 LPW LED N/A N/A N/A N/A N/A * This lumen range breaks across a jump in efficacy from 45 to 60LPW at a rated wattage of 15W. Proposed CSL Lumens Standard Inc./Hal Mod. Spec. Inc. Hal Bare CFL Covered CFL LED 310-749 29 w 750-1049 43 w 1050-1489 53 w 1490-2000 72 w 29 w .75xLumens 50 LPW 45 LPW 55 LPW 43 w .75xLumens 60 LPW 50 LPW 65 LPW 53 w .75 x Lumens 61 LPW 52 LPW 65 LPW 72 w .75 x Lumens 62 LPW 55 LPW 65 LPW 2000-2600 None ES.3.11 GSL Design Options DOE requests comments on the design options it is considering for GSLs. NEMA Comment: Integrated and non-integrated CFL designs are very mature. This technology was first brought to market in the 1980s. All of the practical design options to improve efficiency have been implemented over the past 20 to 30 years. There is very little room left to improve efficiency. White light LEDs are a much newer and evolving technology. Several of the technology options listed are being evaluated or optimized as a means of increasing efficiency. We note that current LED lamp designs are very efficient compared to legacy technology, and as a result of market forces manufacturers are now focused on reducing cost versus increasing efficiency even higher. The DOE should recognize this and NOT pursue extremely high efficiencies above today’s average LED product, in favor of affording industry the ability to focus on reducing cost which will directly spur increased adoption. High CSLs risk inhibiting adoption by increasing LED product costs relative to low CSL levels. This is because at the high CSL levels technology advancements will have to be used to increase efficiency rather than decrease costs. ES.3.12 GSL Data Approach DOE welcomes comment on the data approach including any additional databases that should be considered. 14 NEMA Comments: NEMA has serious reservations about the use of catalog data in assessing the performance of GSLs, especially for products without published test procedures. DOE will recall that initial or rated lumens are long term means and that rated or nominal lumens are not necessarily measured values. Furthermore, NEMA is unaware of any databases where nonintegrated CFL pin base lamp data are available. ES.3.13 GSL Baseline Selection DOE is directly analyzing all product classes and has selected one baseline lamp for each product class. DOE requests comment on its analysis of one baseline lamp for each product class. DOE also requests comment on the baseline units selected for each product class. NEMA Comment: We are concerned about the selection of baseline lamps for MBCFL and the non-integrated pin-base CFL. The MBCFL baseline lamp is practically aligned with the requirements for ENERGY STAR. CFL pin base lamps have unique base and pin configurations and if the baseline lamp is eliminated consumers will be faced with stranded assets because anyone with these fixtures will be forced to make fixture changes. We also note that the CSLs proposed for CFL are NOT for two levels of performance of the SAME product but instead for DIFFERENT products. CSL0 is clearly for a lamp with a cover, and CSL1 is for bare spiral lamps. DOE must not eliminate the consumer preference for a utility of a lamp with a cover, so these products should be divided into two classes for the purposes of standards. (See our comments above in item ES.3.10 under the heading “CFL”.) ES.3.14 More Efficacious Substitutes DOE requests comment on the criteria used in selecting more efficacious substitute lamps in the integrated product class, as well as the characteristics of the lamps selected. In particular, DOE requests comment on its assumptions that more efficacious substitutes must have lumen output within 10 percent of the baseline lamp and must be omnidirectional light sources. NEMA Comment: We agree that to satisfy consumer expectations for replacement lamps, that the substitute lamp must be within 10 percent (either higher or lower) of the lumen output from the baseline lamp. The source should also be reasonably omnidirectional if it is serving a general service lamp application. Although ENERGY STAR specifies intensity distribution requirements for omnidirectionality, CFLs are presumed to be omnidirectional and are excluded from the testing. Thus lamps investigated by the DOE as marketed as omnidirectional were most likely SSL products. Because omnidirectionality has not been a topic of standards in DOE before, requiring it in a substitute lamp may very well exclude CFLs from the discussion, which is not acceptable. We are concerned about DOEs choices of more efficacious substitutes for non-integrated pinbase CFL lamps. The baseline and more efficacious substitute lamps the DOE has chosen are 4-pin non-integrated lamps used in commercial, rather than residential applications. The two CSL1 choices are problematic. The first choice offers no energy savings - it is the same wattage as the baseline lamp. The second choice, a reduced wattage lamp, may not be supported by the existing ballast and furthermore is not within 10% of the baseline lamp lumens. ES.3.15 Non-Integrated Ballast Pairing 15 DOE pairs non-integrated GSLs with representative ballasts because the non-integrated GSLs analyzed in this preliminary analysis operate on a ballast in practice. DOE requests comment on the lamp-and-ballast systems selected for the non-integrated product class. NEMA Comment: Non-integrated lamps must be paired with a unique ballast and a very specific base to electrically and mechanically operate. DOE has selected only one system to analyze. There are dozens of pin-based systems that have not been analyzed, but are included in the scope proposal. Analyzing different lamp/ballast combinations will produce very different results. In most cases, the DOE will determine that there is no energy saving potential with most combinations. The area is very complex and the current analysis is far from adequate. Due to the complexity, the limited energy savings potential, and the maturity of this product line, NEMA strongly suggests that DOE remove this product category from the scope of this rulemaking. These commercial products are not an acceptable replacement for traditional incandescent general service lamps. We believe that if DOE performs a complete analysis on all product combinations it would support NEMA’s conclusion. ES.3.16 Candidate Standard Levels DOE requests comment on the CSLs under consideration for the integrated and non-integrated product classes, including the max tech levels. NEMA Comment: Per our comments in item ES.3.13, by placing all integrated lamps into only two categories DOE ends up with CSLs that only represent one type of technology. More concerning is that with this approach, the standards will either be set very low so all technologies, including covered-CFLs can meet all levels, or very high standards such that only the most efficiency LED lamps can meet the levels. Both results are far from ideal from either energy savings, product cost/availability or utility standpoints. To avoid either outcome, a carefully constructed class matrix must be used with separate baseline lamps for each technology class. ES.3.17 CSL Equation Methodology DOE requests comment on the methodology used to develop the CSLs equations. In particular, DOE requests comment on the use of a lumens-based equation and the equation form itself. NEMA Comment: We are very concerned how the CSL equation for non-integrated GSLs was developed. These lamps are currently unregulated and have no test procedure. We are unaware of databases for these lamps and if they exist, the veracity of the data. The DOE cannot reasonably rely upon catalog data to determine the efficacy of pin-based CFLs. Nominal and rated wattage are not measured watts and catalog initial lumens represent long term data, not individual lamp photometric performance. Furthermore, since there is no defined test procedure for non-integrated lamps, testing laboratories may not be using the same test methods and thus the information published in individual manufacturers’ catalogs may not be comparable. NEMA is also concerned about how the curves will translate across the 4 lumen ranges. There can be slight discontinuities in efficiency, up or down, depending on the technology used in the various ranges. Although a smooth curve across all 4 lumen bins may be the final outcome for some CSL levels. Each lumen bin should be evaluated separately as to the proper efficiency level for that bin, and for a specific technology. It is more likely that the curve will not connect 16 smoothly across all 4 bins at every CSL. There will be fewer CSL levels for CFL technology, whether integrated or non-integrated. ES.3.18 Non-Integrated GSL Replacement Assumptions DOE found that the fixtures frequently used with the non-integrated GSLs analyzed were available in configurations for several different lamp types, and therefore assumed that fixture compatibility would not be an issue for the vast majority of consumers. DOE requests comment on its assumption that fixture compatibility would not be a common issue for non-integrated GSL replacements. NEMA Comment: What the pTSD does not acknowledge is that, while fixtures are available in configurations for various lamps types, a particular fixture is configured for a particular base, and with the possible exception of recessed cans, a particular length and a particular shape (diameter) lamp. Just because a manufacturer makes fixtures that accommodate the available bases and sizes does not mean that the available lamps complying with the standard will have the correct base or be the correct size to fit in a customer’s existing fixture. In this situation, the fixture (and ballast) will need to be replaced and not everyone is able to afford to replace the fixture. Put another way, setting standards for non-integrated lamps significantly increases the occurrence of stranded consumer assets. DOE evaluated the impacts of CSL 1 on the individual base types in the non-integrated product class. DOE confirmed that the vast majority of base types were still available at CSL 1, and therefore consumers will not be forced to switch between lamps with differing base types. DOE also requests comment on its assumption that consumer utility will not be lost with the base types that remain at CSL 1. NEMA Comment: Pin-based compact fluorescent lamps have either 2 pins or 4 pins. Each 2-pin lamp has an internal starter and is designed for preheat, magnetic operation. The 4-pin lamps are designed for electronic ballast operation and are dimmable. These lamps have no internal starter; starting the lamps is a function of the ballast. For choke/starter operation: For electronic or dimming operation: 17 The fits are poke-yoked or keyed which means that one cannot mismatch the base and the socket unless one alters the base (which has its own safety concerns). Removing a base reduces utility. Although many lamps are still available at CSL1, these products operate at the same wattage as the base case, only with slightly higher lumen output; therefore, no energy is saved. This is the same issue as discussed in the General Service Fluorescent Lamp rulemaking4; higher efficiency at the same ballast wattage is not true energy savings, just potential over-illumination. However if the available lamps do not fit (base, size) entire fixtures will have to be replaced for most of these base types. This is not economically feasible. As the two options are no energy saved, or not economically feasible, this product class should not be regulated and should be dropped from the scope of general service lamps. ES.3.19 Product Price Determination DOE invites comment on the methodology and results for estimating end-user prices for GSLs in this preliminary analysis. DOE also requests comment on the appropriateness of the distribution channels and estimated percentage shipments through each channel used in this preliminary analysis. NEMA Comment: As stated in the January 20th public meeting, NEMA believes the DOE has erred in their assumptions about the causes of change to product prices over time. The very short amount of market exposure for LED lamps and high rate of innovation has resulted in strong price reductions with large technology improvements. DOE’s typical analysis model examined mature products well-along in their lifecycle, and assumes that improvements are more incremental. In the case of LED lamps, technology improvements have been jumping forward as new discoveries are applied, and families of products are only now evolving in a linear method like legacy technology did. Therefore, it is incorrect to compare prices for lamps for sale today with lamps for sale a few years ago. They are not the same lamps. They are new 4 See page 14 and related: http://www.regulations.gov/contentStreamer?objectId=0900006480f75c3e&disposition=attachment&contentTyp e=pdf 18 designs and they are not comparable. lamps to recognize this phenomenon. The DOE must redesign their price model for LED ES.3.20 GSL Hours of Use DOE requests comment on the data and methodology used to estimate operating hours for GSLs, particularly in the residential sector. Also, DOE seeks comment on its assumption that GSL operating hours will not vary by light source technology during the analysis period. NEMA Comment: Our comments here are related to our comments in response to the next question (ES.3.21). In chapter 2 of the pTSD, §2.9, DOE states: “DOE assumed that the higher hours of use found for CFL GSLs is based on those lamps currently disproportionately filling sockets with higher hours of use. This would not be the case during the analysis period, when CFL and LED GSLs are expected to fill all GSL sockets.” For the reasons explained in response to the next question, NEMA disagrees that “this would not be the case during the analysis period.” The sources of information cited by the DOE in chapter 2 of the pTSD, §2.9 pages 2-80 through 2-82 and the pTSD chapter 7 confirm NEMA’s own experience that CFLs and LED lamps are predominantly installed in sockets with somewhat longer operating hours than sockets where halogen incandescent lamps are installed.5 In residential applications, lamps with longer lives tend to be used in fixtures with longer operating hours, and therefore, it is likely that halogen lamps will be used in fixtures with shorter operating hours such as closets, or used on dimmers at lower wattages, which extends lamp life. CFLs are the next longest lived lamps and can be expected to be used in fixtures with somewhat longer operating hours, and LEDs are expected to be used in fixtures with longest operating hours. This natural evolution of long life lamps being matched with fixtures with long operating hours is due in no small part to consumers reducing the inconvenience and hassle of constantly replacing lamps.6 We refer to this as a “convenience” phenomenon, but we also do not discount that consumers also implicitly, if not explicitly select longer operating sockets for lower wattage, higher efficacy general service lamps due to the energy savings potential and lower total owning costs. On the other hand, a halogen incandescent lamp that is placed in a socket that is rarely used and operated can last just as long (or longer) and have a lower total owning cost for the consumer than a more efficacious lamp that has longer operating hours and is used more frequently. 5 For example, the report on Northeast Residential Lighting --- Hours of Use Study by NMR reached the following conclusion: “HOU estimates for [CFL] bulbs are significantly higher compared to those for [incandescent] bulbs. Estimates for [incandescent] and [CFL] bulbs across the five sets of estimates obtained from the hierarchical model are all statistically similar, meaning that use of [incandescent] bulbs does not vary much across the areas, and neither does use of [CFL] bulbs. While the difference in HOU between [CFL] and [incandescent] bulbs is consistent and very clear across all regions and room types, the reason for this consistency is unclear. Three competing theories could help explain the differences: differential socket selection, shifting usage, and increased usage (snapback). However, this evaluation did not collect any data to support one theory over the others. The authors of the Uniform Methods Protocol suggest that measuring snapback for residential lighting programs is not typically possible as it requires both pre- and post-metering of energy efficient lighting. The Uniform Method Protocols do not recommend adjusting for snapback effects in hours-of-use estimates. However, as differences in usage are observable in the data collected for this evaluation, the Team suggests assuming that the behavior posited by the three theories are equally responsible for the difference observed (i.e. each accounts for one-third of the difference).” 6 DOE acknowledged this behavior in chapter 2 of the pTSD, at page 2-86. 19 An energy use characterization analysis that evaluated all technologies in general service lamp sockets would recognize the dichotomy in hours of use between incandescent and CFL (and presumably LED) that the studies cited by the DOE in section 2.9 of the pTSD confirm. It is also important to make a separate point about pin base CFL. These lamps are generally used in the commercial market with common applications being hospitality, office, retail, health care, and education where the hours of operation will be longer: many of these commercial applications have 12-16 hour on cycles. These lamps are not typically used in residential applications and therefore cannot be compared with medium screw base lamps which tend to be used most in residential applications. This is important because the operating hours in the residential sector are much lower than in the commercial sector. In summary, operating hours should vary somewhat with technology and most dimming circuits should be assumed to operate halogen lamps. ES.3.21 EISA 2007 Backstop Criteria DOE requests any data suggesting that the EISA 2007 backstop criteria will not be met. NEMA Comment: This is a very difficult question to respond to given the legislative and legal posture this rulemaking finds itself in. It is not the rulemaking that Congress intended in EISA 2007, and we appreciate DOE’s explanation for this circumstance. NEMA disagrees with the premise underlying the question and the assumption that permeates many chapters of the pTSD: that the energy consumption of general service lamps in sockets after January 1, 2020 will exceed the energy consumption of such lamps if every installed socket had a general service lamp with an efficacy of 45 lumens per watt. NEMA’s present view is that by the next decade, the installed mix of halogen incandescent lamps, compact fluorescent lamps, and light emitting diode (LED) lamps in general service lamp sockets will consume less energy than the energy consumed by every general service lamp socket containing a lamp with an efficacy of 45 lumens per watt. In short, it is our expectation that the so-called “backstop” will not be triggered, but that innovation, consumer choice and the market will bring the energy consumption of a mix of lamps in the installed base of general service lamp sockets to a point where the heavy hand of what might be an unpopular regulatory intervention by the Secretary of Energy is unnecessary. The industry’s current expectation is that continuing innovation will bring about further reduction in wattage consumed by general service LED lamps (and corresponding increase in efficacy), will reduce the cost and initial purchase price of those lamps to be extremely competitive with compact fluorescent and halogen incandescent lamps, and that consumers --- both in the residential and commercial sector --- will find these lamps highly suitable for many applications, particularly where the hours of use are longer. Furthermore, the traditional incandescent lamp that no longer meets the energy conservation standards established by EISA-2007 will be nearly eliminated from general service lamp sockets. The confluence of these factors will result in the installation of general service LEDs in sockets presently occupied by a significant number of both incandescent and compact fluorescent lamps. This expected energy savings outcome does not require that the socket penetration of general service LED lamps represent more than half the installations or anywhere near that figure. NEMA’s view is not isolated. We note the report published by the DOE in August 2014 titled Energy Savings Forecast of Solid-State Lighting in General Illumination Applications. The 20 information presented in that report at pages 13-15 effectively confirms NEMA’s own expectations just expressed. Another DOE published report confirms NEMA lighting manufacturers’ actual experience: market adoption of LED lighting is occurring far more rapidly than their experience with CFLs or even they predicted just a few years ago: In NEMA’s view, the analysis in the pTSD that assumes that the “backstop” will be triggered is an unrealistic, if not artificial analysis of the general service lamp market both before and during the analysis period. Since DOE apparently cannot legally analyze a market for a mix of general service lamps that includes the halogen incandescent lamp, it becomes arbitrary and capricious to assume that the “backstop” will be triggered. ES.3.22 Fraction of Dimmable GSLs in the Residential Sector For the residential sector, DOE estimated that at the compliance year five percent of CFL GSLs will be dimmable based on manufacturer interviews, whereas no such limit was placed on LED GSLs (though they may not be installed in fixtures that employ dimmers). DOE requests comment on this approach. NEMA Comment: We agree that dimmable lamps are a SMALL portion of the CFL market. A higher percentage of LEDs are expected to be designed with dimming circuits, although it is difficult to estimate the exact number as this market is still evolving. ES.3.23 Energy Savings from Lighting Controls For this preliminary analysis, DOE estimated 30 percent energy savings for any GSL operated with lighting controls, including dimmers and controls integrated into smart LED lamps. DOE requests data and information to help compare the energy use implications of using dimmers as opposed to other lighting controls. NEMA Comment: It is important to understand that the way control systems work with ballasted lamps and luminaires (individual control) is not the way (phase cut) circuit control systems work with non-ballasted lamps and luminaires (simultaneous control of groups of lamps). Dimming integrated CFLs on phase cut dimmers does not save as much energy as dimming incandescent or halogen. NEMA is concerned that arguments claiming energy savings on very low wattage lamps are suspect. With respect to LED lamps, there are little energy savings to be gained from dimming these already low-power sources. Savings from operation on residential dimmers will vary widely depending on room use. If used for deep dimming such as for TV viewing savings can be significant at those times. Other times dimming will be based on mood lighting. Dimming for dining can vary significantly depending on the occupant. While individual dimming experiences will differ significantly, a 30% overall energy reduction for lighting systems that are dimmed is a reasonable estimate. 21 There are fewer opportunities for pin-base CFLs to operate on control systems or dimmers; there is no dimming in retail applications and dimming is limited in hospitality applications to meeting and banquet rooms (not hallways or lobbies). We would estimate 5% overall energy savings for pin-based systems used in commercial applications. Additionally, control systems associated with smart LED lamps are used for architectural and entertainment purposes, not energy savings. For residential users, it is all about the experience and task tuning. Currently, smart LED lamps are more common in residential use than in commercial use. It is our understanding that utilities are working on models to determine the potential energy savings of controls for commercial LED users. ES.3.24 Integrated Low Lumen GSL Market Distribution Estimates DOE seeks comment on the market distribution estimates for the four lumen ranges analyzed as part of the integrated low lumen product class. NEMA Comment: The market distribution analysis further illustrates the problems with putting all technologies together in the same product classes. CSL 1 is essentially all compact fluorescent lamps; CSL 2 and CSL3 represent older LED technology that is still on the market because LED product evolution is occurring so rapidly. CSL 4 and 5 represent differing types of LED technology at levels that could never be met by CFL lamps. The very strange distribution shown here is further evidence that these products should not be analyzed using this classification method and why NEMA has proposed an alternative classification method. ES.3.25 Consumer Purchases in Standards Case Analyses In each of the 10,000 sampled purchases used to determine the average LCC savings, DOE assumes that in the standards case consumers purchase lamps that are at least as efficient as the ones they would purchase in the absence of standards. DOE seeks comment on this assumption. NEMA comment: In the absence of standards, products may develop on other attributes such as color temperature, color rendering (or whichever new metric is developed) or long life that are more important to consumers but have lower efficiencies. These products are not developed in the standards case. However, we recognize it is difficult to predict the development of products not currently on the market and therefore this may be the only practical assumption possible. ES.3.26 Commercial Hours of Use Variability DOE invites comments and data on its approach to account for variability in hours of use in the commercial sector. NEMA Comment: We believe that 10-12 hours is appropriate for non-integrated pin-base CFLs. ES.3.27 GSL Service Life Scenarios DOE is analyzing three GSL service life scenarios in its analyses. DOE invites comment on the lifetime scenarios considered in the LCC, PBP, and subsequent analyses. 22 NEMA Comment: We agree that Scenario 17 is the most likely scenario and that the effects of frequent starting on CFL life should be taken into account in residential applications. In Scenario 2, again we agree that few lighting systems will last longer than 20 years and have a high likelihood of being replaced after two decades. Scenario 3 contains an assumption that LED lamps may last a median of 5 years, similar to consumer electronics. At this time, there is very little evidence to either support or contradict this assumption. ES.3.28 Lifetime Distribution Modeling DOE invites comments and data on the assumptions and methodology used to calculate GSL survival probabilities as a function of GSL age. NEMA Comment: Non-integrated pin-base CFL life is dependent of number of starts per day. If there is frequent switching then the life can be significantly reduced. Lamps are usually rated at 3 and 12 hour starts just like GSFL. Weibull analysis, while mathematically correct, often does not reflect the real life results for MBCFL well – due to application conditions and cycle time variations. We note that the DOE based the Weibull analysis on a graph of fluorescent lamp mortality from the IES Handbook (note “e” on page 8E-8). The mortality graph in the IES Handbook represents linear fluorescent lamps. The Weibull shape factor will be different for integrated lamps and may affect the DOE’s conclusions. ES.3.29 Installation Costs For this preliminary analysis, DOE assumed that the installation costs for GSLs are identical for all CSLs and product classes in the residential sector as well as the commercial sector (though they may differ by sector). Therefore, DOE did not include installation costs. DOE welcomes comment on this approach. NEMA Comment: Many consumers will require an electrician if the non-integrated pin-base lamp ballast fails and they need to replace the ballast or the fixture. ES.3.30 LCC and Consumer Impacts from Potential Standards For a potential standard at each CSL, DOE presents the resulting average LCC savings and the percent of consumers affected by the standard using the base-case and standards-case efficiency distributions calculated in the shipments analysis. DOE seeks comment on this approach. NEMA has no comment. ES.3.31 GSL Disposal Costs DOE requests comment and relevant data on the disposal cost assumptions used in its analyses. NEMA Comment: The Universal Waste Rule designates the generator of the waste (e.g., user of the lamp) as the responsible party for its disposal. In Maine and Vermont, Extended Producer Responsibility (EPR) laws have been passed, and manufacturers operate and fund recycling programs for residential lamps through NEMA. In Minnesota, some utilities add a small charge to utility bills to fund recycling programs. Starting in January 2015, $0.25 per lamp 7 pTSD page 2-86 23 will be added to the purchase price of mercury containing lamps sold or sold into Washington. Some retailers (Lowes, Home Depot, IKEA) as well as drug stores, hardware stores and grocery stores, collect and pay for recycling lamps. ES.3.32 LCC and PBP Methodologies DOE requests comment on the overall methodology and results of the LCC and PBP analyses. NEMA Comment: We suggest the DOE re-run their analyses once issues regarding price methodology are resolved, per our response to item ES.3.19 ES.3.33 GSL Shipments Data DOE requests any representative data on GSL shipments as they become available in order to improve the accuracy of the shipments analysis. NEMA Comment: we are still evaluating available shipment data for use by the DOE in this analysis. ES.3.34 Residential Integrated High Lumen GSLs DOE requests comment on its assumption that approximately 3 percent of all residential-sector GSLs with integrated ballasts or drivers are brighter than 2,000 lumens. NEMA Comment: we agree that these products are approximately 3% or less. ES.3.35 Integrated LED Fixture Penetration DOE assumed that integrated LED fixtures will capture 0 percent, 15 percent, and 50 percent of the fixture market by 2049 in the low-penetration, reference, and high-penetration scenarios, respectively. DOE invites comment and data on these scenarios. NEMA has no comment at this time. ES.3.36 Non-Integrated CFL GSLs DOE assumed that non-integrated CFL GSLs will remain a constant fraction of the installed GSL stock in the commercial sector, after accounting for the incursion from integrated LED fixtures into the commercial building stock. DOE seeks comment and data on this assumption. NEMA Comment: We do not agree with the assumption that CFL pin in commercial will remain constant. We anticipate a significant decline over the analysis period. The CFL pin recessed cans have lower efficiencies relative to LED. The trend has already started to move to LED retrofit kits (NOT non-integrated pin-base LED) in existing commercial applications and new LED fixtures in new construction. We anticipate that this product area will be significantly replaced by new LED fixtures over the next 30 years. ES.3.37 Rare Earth Price Scenarios DOE invites comment on the two rare earth materials price scenarios considered in its analyses. 24 NEMA Comment: A rare earth price spike hurts everyone not just industry but consumers as well. We refer the DOE to our comments on the GSFL rulemaking8. ES.3.38 Electricity Price Scenarios DOE invites comment on the electricity price projection scenarios considered in its analyses. NEMA has no comment at this time. ES.3.39 LED Learning Rate Scenarios In its reference scenario for this preliminary analysis, DOE assumed that the LED GSL learning rate will slow in the near future to equal the historical learning rate for CFL GSLs. In an alternative scenario, DOE assumed that the LED GSL learning rate will remain at the faster value observed in recent years. DOE invites comment on these scenarios. NEMA Comment: We again recommend the DOE use multiple product classes as suggested in item ES.3.13 so as to separate the two technologies and allow different learning rates. This is because we believe that high efficiency requirements from DOE and other entities such as the California Energy Commission will cause prices to remain at current levels as manufacturers are forced to pursue efficiency over other tradeoffs. Alternatively, lower CSLs will allow manufacturers to continue to pursue cost-savings and result in lower prices and increased adoption. We suggest the DOE explore both scenarios, and we believe that the lower CSLs with lower cost are capable of achieving a higher national energy savings. ES.3.40 Incremental Price of Brighter LED GSLs DOE is assuming that both the price of LED GSLs and the incremental price of brighter LED GSLs are falling. DOE requests comment and data on this assumption. NEMA Comment: While both prices may be falling, a price differential between high and low brightness LED products will still remain because more lumens come at a cost of more materials; more LED dies or larger drivers, or both. We recommend the DOE investigate material and other cost contributors to pricing in a family of products. As we recommended at the DOE Public Meeting of January 20th, the DOE should avoid comparing separate generations of lamp designs, which has contributed to the mistaken conclusion that higher lumen lamps are cheaper or of equal price. The misunderstanding results from first generation LED lamps of low lumen being the first to market at high prices. Innovation has allowed brighter, cheaper LED lamps in relatively short time. These lamps are different generations and cannot be compared. ES.3.41 Pre-Compliance Year Shift Away from Incandescent GSLs DOE assumed in its shipments projections that some fraction of the market for GSILs shifts to CFL or LED GSLs in each year prior to 2020, with the remainder shifting to CFL or LED GSLs in 2020. DOE assumed that the remaining market for GSILs in the commercial sector is already negligible, so this shift was assumed to occur entirely within the residential market. DOE requests comment and data on this assumption. 8 See page 34: http://www.regulations.gov/contentStreamer?objectId=0900006480f75c3e&disposition=attachment&contentTyp e=pdf 25 NEMA Comment: We agree that there will be some shift from GSILs to CFL or LED GSLs as well as some shift from CFL to LED each year up to 2020. We expect this shift to be significant, and increasing over the time period. LED products purchased between now and 2020 will still be expected to be installed in 2020. For the reasons explained in our response to Question ES.3.21, NEMA does not expect the entire market or substantially the entire market to shift suddenly to CFL and/or LED in 2020, and we submit this is an erroneous assumption.9 The DOE cannot assume the shift in shipments as it has without consideration given to the probability that the current enforcement prohibition may continue into 2020 and beyond. Because the DOE lacks analysis on this issue, this is still an open question. Table 9.5.3 of the pTSD shows decreasing sales of the combination of CFL sales plus LED sales from 2015 to 2020. With projected declining LED lamp prices, industry believes that the number of LED sales will significantly increase, more than offsetting the reduction in CFL sales such that the combined total of both technologies will continue to increase, not decrease, from 2015 to 2020. Table 9.5.3 also shows a sales spike that is both unrealistic and impossible to achieve. Production capability is scaled to market demand. Manufacturers, individually and collectively, could not economically justify an enormous increase in production capability for one year, especially if it was known, as represented in the Table, that the entire market would return to “normal” the following year. If this were to occur, there would be severe product shortages and disruption in the market. This is illustrative of the artificiality of the analysis that we mentioned in our response to ES.3.21. ES.3.42 GSL Market Data DOE requests comment and data on current market shares and market trends for GSLs in the commercial sector. NEMA Comment: Most of the products covered under this rulemaking are used in the residential sector. The exception is pin-based lamps which are used almost entirely in the commercial sector. However, pin-based CFL lamps represent a very small percentage of the commercial market. Their use has been flat to declining over the past 5 years and a continued decline in their use is expected. As this is a declining market area with very little potential for energy savings, NEMA believes these products should be removed from this rulemaking. (See our chart in item ES.3.7) ES.3.43 GSL Rebound Effect Scenarios DOE analyzed three rebound effect scenarios in the NIA: 1) 0 percent rebound for both the residential and commercial sectors (the reference scenario), 2) 8.5 percent rebound and 1 percent rebound for the residential and commercial sectors, respectively, and 3) 15 percent rebound for both the residential and commercial sectors. DOE requests data that can be used to further refine the rebound effect assumptions used in the NIA. 9 It is anybody’s guess, but the statutory basis for the DOE’s inability to conduct a full analysis of the general service lamp market brought about by the Consolidated Appropriations Act of 2014 can equally be assumed to remain in place in 2020 and beyond and that fact alone would render the DOE’s premise and assumptions unrealistic. 26 NEMA Comment: While we believe there will be little to no rebound effect in the commercial sector, we do not rule out some rebound effect with the growing presence of LED products in the residential sector as their extremely long life and low energy use could lead some consumers to operate their lamps for longer periods of time without producing nuisance replacements or high energy bills. An 8.5% to 15% rebound effect for LED lamps used in the residential sector is foreseeable. ES.3.44 Penetration of LED GSLs with Standby Functionality DOE assumed that residential LED GSLs with standby functionality will represent 0 percent, 50 percent, and 100 percent of residential LED GSLs in the market by 2049 in the low-penetration, reference, and high-penetration scenarios, respectively. DOE invites comment and data on these assumptions. NEMA Comment: While use of standby functionality will increase, it is very unlikely that it will ever reach 100% because adding this functionality will add cost, and there will always be a market for the lowest cost, no-features product. ES.3.45 Penetration of Commercial GSLs with Controls DOE analyzed two scenarios to account for the fraction of GSL shipments with controls in the commercial sector: 1) The current fraction of GSLs with controls remains constant over the analysis period, and 2) The fraction of commercial floor space utilizing various types of controls grows from 30 percent today to a projected value of 80 percent by the end of the analysis period (the reference scenario). DOE invites comment and data on these scenarios. NEMA Comment: Assuming that ASHRAE 90.1 continues to be the national energy code throughout the analysis period (or IECC equivalent), it would be reasonable to assume that most commercial floor space as well as other qualifying spaces will have controls. Almost any renovation will necessitate the use of a control system and these controls would include manual on or partial automatic on, bi-level dimming, automatic off or scheduled shutoff, and automatic daylight responsive controls where appropriate. ES.3.46 Shipments Analysis and NIA Scenarios DOE is considering a number of scenarios in its shipments and NIA analyses. DOE asks for comment on whether there are other scenarios which should be considered. NEMA Comment: DOE should consider a parallel scenario where halogen lamps remain on the market. ES.3.47 Consumer Subgroup Analysis DOE welcomes input regarding which, if any, consumer subgroups should be considered when developing potential energy conservation standards for GSLs. NEMA Comment: Low income consumer groups will be most affected if low cost halogen or low cost CFL lamps are no longer available in 2020. We refer to our comments in item ES.3.40 above about the tradeoff of efficiency versus cost and adoption rates. ES.3.48 Emissions Analysis 27 DOE requests comment on its approach to conducting the emissions analysis for GSLs. NEMA has no comment at this time. ES.3.49 Monetization of Emissions Reductions Benefits DOE invites input on the proposed approach for estimating monetary benefits associated with emissions reductions. NEMA Comment: As we have previously noted in appliance efficiency proceedings under the Energy Policy and Conservation Act: Given the enormous uncertainty in the IAMs models, these models --- even “averaged” as the Interagency Working Group has done --- are poor tools for agency decision-making, particularly with respect to products regulated by EPCA that are not themselves a source of emissions. Reliance on the SCC to justify a standard could be socially counterproductive: An agency could very well end up justifying the imposition of enormous costs upon a non-polluting, energy-saving product that it wants to see penetrate the market in greater numbers that could not otherwise be justified without the uncertain benefits attributed to the SCC. NEMA believes that DOE should base its net benefit determination for justifying a particular energy conservation on the traditional criteria relied upon by DOE --- impacts on manufacturers, consumers, employment, energy savings, and competition. If there are estimated benefits from reduced carbon emissions, they should be noted --subject to the inherent uncertainty and difficulty in monetizing those benefits --- as benefits over and above the net benefits from DOE’s traditional analysis under EPCA. We incorporate by reference our entire comment on this subject in the recently completed incandescent reflector lamp and general service fluorescent lamp rulemaking,10 which the DOE acknowledged in the Final Rule published in that rulemaking. 80 Fed.Reg. 4042, 4099 (January 26, 2015). ES.3.50 Utility Impact Analysis DOE seeks comment on the planned approach to conduct the utility impact analysis. NEMA has no comment at this time. ES.3.51 Employment Impact Analysis DOE welcomes input on its proposed approach for assessing national employment impacts. NEMA has no comment at this time. ES.3.52 Regulatory Impact Analysis 10 Energy Conservation Program: Energy Conservation Standards for General Service Fluorescent Lamps and Incandescent Reflector Lamps, Docket No. EERE-2011-BT-STD-0006. NEMA Comments dated June 30, 2014, pages 38-44. 28 DOE requests any available data or reports that would contribute to the analysis of alternatives to standards for GSLs. In particular, DOE seeks information on the effectiveness of existing or past efficiency improvement programs for these products. NEMA Comment: Millions of dollars have been spent by Utilities through their rebate programs to reduce the price of energy efficient products within this product scope. Depending on the standard level selected the Utilities may, or may not, continue these programs to reduce product price, thus adoption could actually slow under a given pricing scenario. We again refer to our comments about efficiency, price and adoption in item ES.3.40. Manufacturer Impact Analysis NEMA Comment: The DOE should appraise the fact that the lighting manufacturing industry has been and is undertaking a significant amount of investment that is integral to a market transformation, developing a wide-range of new and innovative solid-state lighting products. This innovation and investment is expected to continue for several years to come. Innovation has been and continues to occur rapidly, making products previously introduced to the market quickly obsolete. Investment will continue to be directed in the future at making these products even more affordable to consumers and possibly delivering energy savings beyond what energy conservation standards under consideration will deliver. The amount of this investment directed at one segment of general service lamps, LEDs, makes it very difficult to justify any significant new investment in other segments of the general service lamp definition that represent the more mature product lines. Stated another way, regulatory action that forces manufacturers to make incremental investments in mature products that generate only modest benefits can make it more difficult to invest in the future. Given the intense competition that is occurring in developing new and better solid-state lighting products, it is unlikely that manufacturers will invest in old technology at this point of the transformation. Therefore DOE should only be evaluating the efficiencies of the products that are on the market today and not assume manufacturers will invest to make these even incrementally more efficient. Mandatory investment in the mature energy-efficient products can hinder competition and American competitiveness. DOE should not lose sight of the market transformation and trend that has already occurred and will continue to occur and avoid unintended consequences that could impact competitiveness, American workers and consumers in this time of change. 29