[Federal Register Volume 61, Number 40 (Wednesday, February 28, 1996)]
[Proposed Rules]
[Pages 7431-7436]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 96-4512]



=======================================================================
-----------------------------------------------------------------------

DEPARTMENT OF ENERGY

Office of Energy Efficiency and Renewable Energy

10 CFR Part 430

[Docket No. EE-RM-94-220-IF]
RIN 1904-AA61; RIN 1904-AA70


Energy Conservation Standards Program for Consumer Products: Test 
Procedures for Fluorescent and Incandescent Lamps

AGENCY: Office of Energy Efficiency and Renewable Energy, Department of 
Energy.

ACTION: Notice reopening comment period.

-----------------------------------------------------------------------

SUMMARY: On September 28, 1994, the Department of Energy (Department or 
DOE) published an interim final rule and a proposed rule regarding 
energy conservation test procedures for fluorescent and incandescent 
lamps. Based on the public responses, the Department is considering 
certain revisions of the interim final rule and proposed rule and seeks 
public comment on options it is considering. The options involve the 
following topics: determining the wattage of a fluorescent lamp for 
purposes deciding whether the energy conservation standards and test 
procedures apply to it; the confidence limit, ``derating factor'' and 
statistical test used in the test procedure sampling plan; definition 
of colored lamps; determining the rated voltage or rated voltage range 
of an incandescent lamp for purposes of deciding whether the energy 
conservation standards and test procedures apply to it; defining rated 
voltage for testing incandescent lamps; and defining the bulb shapes 
for elliptical reflector (ER) and bulged reflector (BR) incandescent 
lamps.

DATES: Written comments in response to this notice must be received by 
the Department by April 15, 1996. The Department requests 10 copies of 
the written comments and, if possible, a computer disk. (The Department 
uses WordPerfect.)
    There will be a public meeting to gather input on these issues in 
Washington, D.C., on March 5, 1996. The meeting will begin at 9:30 a.m. 
and will be held at the U.S. Department of Energy, Forrestal Building, 
Room 2E-069, 1000 Independence Avenue, S.W., Washington, D.C.

ADDRESSES: Written comments are to be submitted to: U.S. Department of 
Energy, Office of Energy Efficiency and Renewable Energy, Ms. Sandy 
Beall, ``Energy Conservation Standards Program for Fluorescent and 
Incandescent Lamps, Docket No. EE-RM-94-220-IF,'' EE-431, Forrestal 
Building, 1000 Independence Avenue, SW, Washington, DC 20585-0121. 
Telephone: (202) 586-7574; Telefax: (202) 586-4617.
    Copies of the transcript of the July 19, 1995 lamp workshop and of 
the public comments on the interim final rule may be read at the 
Department of Energy Freedom of Information Reading Room, U.S. 
Department of Energy, Forrestal Building, Room 1E-190, 1000 
Independence Avenue, S.W., Washington, D.C. 20585, (202) 586-6020, 
between the hours of 9:00 a.m. and 4:00 p.m., Monday through Friday, 
except Federal holidays.

 
[[Page 7432]]

FOR FURTHER INFORMATION CONTACT:
Terrence L. Logee, U.S. Department of Energy, Office of Energy 
Efficiency and Renewable Energy, Mail Station EE-431, Forrestal 
Building, 1000 Independence Avenue, S.W., Washington, DC 20585-0121, 
(202) 586-1689
Edward Levy, Esq., U.S. Department of Energy, Office of General 
Counsel, Mail Station GC-72, Forrestal Building, 1000 Independence 
Avenue, S.W., Washington, DC 20585-0103, (202) 586-2928

SUPPLEMENTARY INFORMATION:

1. Authority

    Part B of Title III of the Energy Policy and Conservation Act, Pub. 
L. 94-163, as amended (EPCA) or the Act, created the Energy 
Conservation Program for Consumer Products other than Automobiles 
(Program). The products currently subject to this Program include 
certain fluorescent and incandescent lamps and medium based compact 
fluorescent lamps. EPCA sets minimum energy conservation standards for 
general service fluorescent and incandescent reflector lamps and 
requires the Department to develop test procedures.

2. Background

    On September 28, 1994, the Department published an interim final 
rule establishing test procedures for general service fluorescent and 
incandescent lamps and for medium based compact fluorescent lamps, 59 
FR 49468, and a Notice of Proposed Rule for definitions of rough and 
vibration service incandescent reflector lamps and colored fluorescent 
and incandescent lamps, 59 FR 49478. In addition DOE held a hearing on 
the proposed rule on November 15, 1994 and a workshop on these issues 
on July 19, 1995. The Department received many comments on the interim 
final rule and on the proposed rule including comments from 
manufacturers, a national trade association, a professional society, a 
utility, and a Federal agency. The comments included requests that the 
Department: (1) modify its test procedure sampling plan to change the 
confidence limit, ``derating factor,'' and statistical test used to 
determine compliance of certain lamps with the energy conservation 
standards; (2) permit testing and compliance for incandescent lamps at 
a lamp's design voltage, and expand the voltage range from the 
statutory requirement of 115 through 130 volts to 100 through 150 
volts; (3) define the exemption for the bulged reflector (BR) and 
elliptical reflector (ER) incandescent reflector lamp by reference to 
the American National Standards Institute (ANSI) C79.1-1994; (4) 
determine that a new product coming on the market, a fluorescent lamp 
rated at 25 watts, which is below the 28 watt threshold for coverage 
under EPCA, is actually a 40 watt fluorescent that is covered by the 
statutory standards and test procedures; and (5) revise its proposed 
definition of colored fluorescent and incandescent lamps.
    In response to the foregoing suggestions, the Department is 
considering various options to alter the Interim Final and Proposed 
Rules. Because the issues raised by these options were not expressly 
considered in either the preamble to the Interim Final or Proposed 
Rules, the Department is now seeking comment from interested parties on 
these options. In particular, the Department seeks any new factual 
information and data that will assist it in addressing these issues.

3. Discussion

    a. Revision of the Sampling Plan. DOE's Energy Conservation Program 
for Consumer Products has been developed and refined since its 
inception in 1978. Compliance with energy efficiency standards has been 
assured in part by having each manufacturer certify that its covered 
products comply with the applicable energy efficiency standard. The 
certification must be based on tests of the product in accordance with 
test procedures prescribed by DOE.
    In promulgating test procedures applicable to certification, one of 
the major goals has been to provide a statistically valid approach so 
that there is a high probability that products which have been tested 
and certified as being in compliance with the applicable efficiency 
standards actually comply with those standards. Each DOE test procedure 
incorporates a sampling plan, and that sampling plan is designed to 
give reasonable assurance that the true mean performance of the product 
being manufactured and sold meets or exceeds the DOE energy efficiency 
standard.
    DOE recognizes that units of a product may vary in energy 
efficiency for a number of valid reasons, including differences in 
component parts, production and testing. The risk to the public of 
purchasing a non-complying product, the risk to manufacturers of 
selling such a product, and the burdens of performing representative 
testing, are reduced through the application of a statistically 
meaningful sampling plan and basing the certification decision on the 
mean energy performance of the sample units.
    There are several critical elements of a sampling plan. One is the 
selection of units for testing. Units must be representative of the 
product, and be selected randomly from a batch. Sample size is also a 
critical element of a sampling plan. The results yielded by energy 
efficiency testing of a product, consisting of tests conducted on a 
sample of units, will be increasingly more reliable as the size of the 
test sample increases. This, however, increases the testing burden on 
the manufacturers. Also, as the variability in performance increases 
among individual tested units of a product, the reliability of the test 
results decreases. As a result, DOE's test procedures require sampling 
plans based on a confidence limit approach. This approach is designed 
to minimize the manufacturers' testing burden while ensuring accurate 
determination of compliance within a specified level of confidence.
    The interim final rule prescribing test procedures for lamps 
requires a minimum sample size of 20 units for each model, which must 
be randomly selected during seven out of 12 months of production. The 
rule further provides in essence that the lamp efficacy for a given 
model of lamp shall be the average efficacy for the tested lamps of 
that model, and ``shall be no greater than the lower of (i) the mean of 
the sample or (ii) the lower 99 percent confidence limit of the true 
mean divided by 0.99.'' DOE views the latter calculation as being a 
one-sided confidence interval using the t-statistic, with the 0.99 
divisor constituting a ``derating'' factor. The confidence limit would 
be calculated using generally accepted methods found in statistics 
textbooks, based on the sample mean and sample standard deviation.
    DOE included the derating factor to take into account variability 
in the efficiency of products due to many factors, including 
manufacturing variability, variations in the material (e.g., 
phosphors), and testing errors, including reference lamp calibration 
errors. Furthermore, this format (confidence limit divided by a 
derating factor) is similar to the format required for other appliance 
products for which DOE has authority to require testing.
    The National Electrical Manufacturers Association (NEMA) has 
proposed loosening the confidence interval to 95 percent, and changing 
the derating factor to 0.97, which increases its derating effect. It 
justifies this proposal on the basis of typical production variations 
and measurement uncertainties, including calibration issues. NEMA 
submitted estimates of 

[[Page 7433]]
the magnitude of these variations and their effect on compliance 
determinations. They estimated an uncertainty of 2 percent due to the 
reference lamps used in the measurement process, with additional 
variability among different laboratories.
    NEMA has also proposed substituting the z-statistic procedure for 
the t-statistic procedure. The z-statistic procedure is similar to the 
t-statistic procedure, except that, for each model of a product, it 
uses the standard deviation, , that applies to the entire 
population of manufactured units for that model. That standard 
deviation is assumed to be known from previous measurements. The t-
statistic procedure, by contrast, uses the standard deviation, s, of 
the sample units tested. The z-test also replaces the factor t with 
another factor z, both of which are found in standard tables.
    The effect of going to a 95 percent confidence limit will be to 
make it slightly easier to demonstrate compliance, while also slightly 
increasing the chance that a noncompliant product will be judged to be 
in compliance. In other words, when testing demonstrates compliance at 
the 95 percent confidence level, there would be a one in twenty chance 
that a non-tested unit of the product may not meet the standards 
instead of a one in one hundred chance under the procedure promulgated 
by the interim final rule.
    The effect of using the z procedure instead of the t procedure will 
be to produce lower confidence limit values which are more favorable to 
the manufacturers, because the value of the z factor from the tables is 
less than the value of the t factor, unless the number of sample units, 
n, is very large. However, the z procedure is more representative than 
the t procedure because the standard deviation in the z method is 
determined from a larger population than the standard deviation in the 
t method. Use of the z procedure requires an accurate measurement of 
the population standard deviation for each model. Accurate measurement 
would appear to require, for example, prior tests of a large number of 
units of that model selected at random, conduct of the prior testing in 
accredited laboratories, and prior testing conducted under conditions 
and using test procedures that are comparable to current conditions and 
procedures.
    The Department is considering the option of permitting a 
manufacturer to use the ``z'' statistic as an alternative to the ``t'' 
statistic, for tests of any product for which the following criteria 
are met: (1) the standard deviation used in the test procedure was 
derived from a minimum sample of 60 or more randomly selected lamps of 
the same basic model; (2) the statistical data was measured by 
accredited laboratories; (3) the prior testing was conducted under 
conditions and using test procedures comparable to current conditions 
and procedures. When these criteria are not met, a manufacturer would 
be required to use the ``t-statistic.'' The Department specifically 
seeks input on whether lamp manufacturers can derive standard 
deviations for their products from historic test experience. The 
Department is seeking comment on this approach or other possible uses 
of the ``z'' statistic. The Department is also considering, and seeks 
comments on, modification of the derating factor and confidence 
interval, as suggested by NEMA.
    b. Definition of Rated Voltage, Determination of Test Voltage and 
Determination of Voltage Range. When the Department considered test 
procedures for incandescent lamps in the interim final rule, it noted 
that neither the definition of incandescent lamp in Section 321(30)(C) 
of the Act, 42 U.S.C. 6291(30)(C), nor Illumination Engineering Society 
LM-20, ``Approved Method for Photometric Testing of Reflector-Type 
Lamps'' defined the test voltage. Therefore, in the interim final rule, 
the Department requires testing of all incandescent lamps at 120 volts 
to be consistent with the statutory requirements for labeling. 10 CFR 
Part 430, Subpart B, Appendix R, Section 4.2.1.
    In its comments, NEMA requested that the Department allow testing 
of incandescent lamps at their design voltage. Otherwise, NEMA claimed 
that certain 125 and 130 volt lamps would be banned from the market by 
failing to meet the standards if tested at 120 volts. The industry and 
NEMA also claim that 125 and 130 volt lamps serve two market niches: 
regions in the country where power line voltage is greater than the 
nominal 120 volts and applications requiring long life lamps. 
Manufacturers claim that they would be forced to sell lamps with 
decidedly shorter lives than the 125 and 130 volt lamps currently in 
the marketplace if DOE requires compliance with the standards at 120 
volts.
    In response to queries by NIST, Philips proposed that the 
Department consider requiring testing of incandescent lamps at the 
rated voltage marked on the lamp. Furthermore, when a lamp is marked 
with a voltage range, Philips proposed that the rated voltage should be 
taken as the mean of the voltage range. This wording is based on text 
taken from the International Electrochemical Commission Standard 432-1.
    The Department believes that requiring compliance for incandescent 
lamps at 120 volts will reduce lamp life for some consumers and may 
also remove most 125 and 130 volt lamps from the marketplace. However, 
none of the manufacturers define what is meant by design voltage. 
Therefore, since the statute uses rated voltage, the Department is 
considering adopting the definition of rated voltage from the Institute 
of Electrical and Electronics Engineers Standard Dictionary of Terms 
which defines rated voltage as ``the voltage to which operating and 
performance characteristics are referred.'' Furthermore, the Department 
is considering a requirement to test incandescent lamps at the rated 
voltage, as marked on the lamp, or at the mean of rated voltage range, 
as marked on the lamp. This approach would provide for testing 
incandescent lamps at a known reference voltage for certification to 
the energy efficiency standards while agreeing with the Federal Trade 
Commission (FTC) requirements for labeling. The Department is also 
considering the option of requiring that lamps not marked with a 
voltage will be tested at 120 volts.
    With respect to the issue of ``rated voltage range'' the definition 
of ``incandescent reflector lamp'' in the Act, refers to a ``rated 
voltage or rated voltage range at least partially within 115 to 130 
volts.'' Section 321(30)(C)(ii), 42 U.S.C. 6291(30)(C)(ii). NEMA 
recommended expansion of the voltage range in the statute to 100 to 150 
volts, asserting that the statutory limit could unintentionally allow 
evasion of the standards requirements for certain products. Under the 
language in the statute, for example, a product could be rated at 131 
volts, thereby removing it from the standard. Yet this product would 
perform acceptably in a 130 volt environment and could be sold for such 
applications.
    The interim final rule incorporates the statutory definition of 
incandescent lamp including the voltage range. The Department will 
continue to use this definition. The Department notes that only one 
manufacturer currently markets lamps with design voltages greater than 
130 volts. However, in response to queries by NIST, several 
manufacturers agreed that the nominal tolerance for incandescent lamp 
voltage is 10 percent. The Department believes that the 
statutory range of 115 to 130 volts may also be subject to this 
tolerance. Therefore, the Department is considering the option of 
treating lamps with voltages greater than 103.5 volts 

[[Page 7434]]
and less than 143.0 volts as being ``at least partially within a rated 
voltage range of 115 to 130 volts,'' and subject to the energy 
efficiency standards.
    The Department is seeking comments on the acceptability and 
workability of these options for rated voltage, test voltage and rated 
voltage range. Alternative proposals are welcome but the Department 
requests that these proposals be supported by references to existing or 
draft industry standards or that the proposals be supported by data.
    c. ER and BR Reflector Lamp Definitions. The Act contains 
exemptions for several types of incandescent reflector lamps including 
those for ER (elliptical reflector) and BR (bulged reflector) bulb 
shapes. Section 321(30)(C)(ii), 42 U.S.C. 6291(30)(C)(ii). However, 
these lamps are not defined in the statute or the interim final rule 
and DOE is concerned that the exemption may be abused without a clear 
definition of what constitutes an ER or BR bulb.
    One commenter provided copies of ANSI Standard C79.1-1994 which 
contain descriptions of the ER and BR bulb shapes. Another commented 
that if the ANSI definition was different than what some manufacturers 
have been using, there would be tooling costs to conform the lamp 
envelope to the new shape definition and DOE should provide time for 
manufacturers to implement the new ANSI requirements. In its comments 
to the workshop, NEMA claimed that there was a consensus to define ER 
and BR lamps by reference to ANSI Standard C79.1-1994.
    An Osram-Sylvania Inc. (OSI) comment claims that: (1) the BR lamp 
is not marketed for recessed applications; (2) BR lamps are more 
efficient than rough/vibration service R lamps; (3) the BR lamp is less 
costly for the residential market than the halogen PAR lamp; (4) OSI 
has introduced a 65 watt BR lamp which meets the efficiency standards; 
and (5) the ANSI C79.1-1994 bulb shape standard is a result of the 
mandatory ANSI 5-year revision cycle and it is fundamental to all lamp/
fixture interchangeability. The Department notes, however, that the 
previous ANSI revision to the bulb shape standard was published in 
1984.
    During the workshop, the American Council for an Energy Efficient 
Economy (ACEEE) commented that this exemption was placed in the statute 
to protect one small manufacturer and that the drafters of the Energy 
Policy Act of 1992 (EPACT) believed that these products were not sold 
in large quantities and were expected to disappear from the market. 
Furthermore, ACEEE comments suggested that the exemption was meant to 
apply to lamps that are rated with lower wattage than their reflector 
(R) or parabolic aluminized reflector (PAR) counterparts. In its 
written comments, ACEEE requested that DOE define ER and BR lamps in a 
way that would limit exemptions for these lamps as originally intended 
in EPACT.
    The Department believes the definitions of ER and BR bulb shapes in 
ANSI Standard C79.1-1994 (Figure 1 on page 7) are new definitions of 
the ER and BR bulb shapes because earlier versions of ANSI Standard 
C79.1 did not include definitions for either ER or BR bulb shapes in 
their current form. ANSI C79.1-1984 discusses the bulged (B) and 
elliptical (E) shape designations as basic bulb shapes of general 
service incandescent lamps. The ``RE'' elliptical reflector shape in 
the 1994 ANSI C79.1 standard could be described as cutting off the top 
half of the basic ``E'' bulb shape in the 1984 document since the 
elliptical portion of the ``E'' bulb forms that part of the ``RE'' bulb 
below the major axis or lens of the reflector bulb. However, the bulged 
reflector bulb would represent a greatly diminished ``B'' shaped bulb 
with a reflector bulb connected to the top of this small ``B'' shaped 
bulb. For these reasons, the Department believes the 1994 ANSI document 
represents a major modification of elliptical and bulged bulbs from the 
1984 document.
    ER and BR reflector bulb shapes typically have a long neck, a 
characteristic which is not addressed in ANSI C79.1. This is presumably 
to extend the lens closer to the end of recessed ceiling fixtures in 
the ER bulb. Therefore, the Department believes the ANSI C79.1-1994 
definitions of the ER and BR bulb shapes are subject to interpretation, 
and questions whether these definitions agree with the commonly 
understood bulb shapes being manufactured and which were contemplated 
by exclusion of ER and BR bulbs from EPCA coverage.
    Although the Department believes the ANSI Standard C79.1-1994 does 
not fully prescribe the ER and BR bulb shapes, the Department is 
considering adopting ANSI Standard C79.1-1994 as part of the definition 
of an ER or BR bulb shape, subject to additional criteria, to capture 
the characteristics of ER and BR bulbs in the marketplace at the time 
the exemptions were established. One criterion being considered is a 
longer neck than an R or PAR lamp with either a specified dimension or 
a dimension stated as a comparison, such as 25 percent longer than 
similar wattage R or PAR lamps. An additional criterion under 
consideration for the BR lamp is to require that the bulged shape must 
be reflectively coated and large enough to redirect light emitted by 
the filament to the side and rear of the lamp toward the lens. The 
Department is also considering a requirement for a reduced wattage 
filament for both ER and BR lamps. The Department is seeking comment on 
whether to specify a certain wattage reduction or to state this 
reduction as a percentage comparison to standard R or PAR lamps.
    The Department invites comments on the definitions for ER and BR 
lamps it is considering. The Department also requests copies of catalog 
listings and other data to help it determine the extent of reduced 
wattage ER and BR lamps offered in the market.
    d. Determination of Rated Wattage for a Fluorescent Lamp. EPCA sets 
standards for fluorescent lamps 48 inches long with rated wattages of 
28 watts or more, 96 inches long with rated wattages of 52 watts or 
more, and 2 foot U-tube lamps with rated wattages of 28 watts or more. 
Sections 321(30)(A) and 325(i)(1), 42 U.S.C. 6291(30)(A) and 
6295(i)(1). The standard levels have the effect of prohibiting the 
sale, after October 31, 1995, of certain lamps previously on the 
market, including 4-foot, 40 watt cool white fluorescent lamps.
    The 4-foot, 40 watt cool white fluorescent lamp consumes 40 watts 
of power when used with a conventional high power factor ballast. High 
power factor ballasts are used in over 85 percent of the fluorescent 
fixtures using four foot lamps. Such high power factor ballasts are 
typically used in commercial applications. If a 40-watt cool white lamp 
is used with the type of low power factor ballast generally used in 
residential applications, the lamp will consume about 25 watts, which 
is below the 28 watt threshold that defines the lower limit of coverage 
in the standards.
    Neither the statute nor DOE's existing regulations specify the type 
of ballast to be used in determining the rated wattage of lamps. In the 
absence of a specification, some have argued that 4-foot lamps could 
have their rated wattage determined using a low power factor ballast 
and if, using this testing method, the rated wattage was less than 28 
watts, the lamp would be exempt from the standard.
    DOE believes that it is unreasonable to apply this statute so as to 
permit the continued manufacture and sale of lamps that when used with 
the most common types of ballasts (i.e., high power factor) would 
consume 28 or more watts, but fail to meet the 

[[Page 7435]]
standards prescribed by the statute. In an attempt to address this 
concern, DOE sent a letter on August 30, 1995, to lamp manufacturers 
indicating that it would consider any lamp that was electrically the 
same as the 40-watt cool white lamp to be subject to the same statutory 
standards. However, manufacturers have since begun to introduce, or 
indicated that they plan to introduce, slight variations on the 40-watt 
cool white lamp that would be rated at 25 watts based on use of low 
power factor ballasts. Despite these modifications, the lamps being 
marketed or developed would still perform like 40-watt cool white lamps 
when used in high power factor ballasts.
    The Department believes that Congress intended the rated wattage of 
fluorescent lamps, for purposes of defining the universe of lamps 
covered by the standards, to be determined by using a high power factor 
ballast. The wattages included in the table that now appears in section 
325(i) of the Energy Policy and Conservation Act appear to assume the 
use of high power factor ballasts. 42 U.S.C. 6295(i). In addition, when 
Congress had previously set efficiency standards for ballasts, those 
standards were only applied to high power factor ballasts.
    The Department is now considering a requirement that the rated 
wattage of a fluorescent lamp, for purposes of determining coverage by 
the standards, is the measured wattage when the lamp is used with a 
high power factor ballast. The Department is soliciting public comment 
on the possibility of requiring the use of high power factor ballasts 
in determining the rated wattage of fluorescent lamps. Before making a 
final determination on this matter, the Department also intends to 
consider other possible means to achieve comparable objectives.
    For example, the Department is considering the approach used in the 
Canadian lamp regulations issued in the November 29, 1995 Canada 
Gazette, Part II, Volume 129, No. 24, pg 3073. Under this possible 
approach, the Department would add an additional phrase to the 
definition for general service fluorescent lamp specifying that, 
``General service fluorescent lamp means any fluorescent lamp that is a 
physical and electrical equivalent of a lamp described in paragraph 
(a), (b), (c), or (d).'' However, the Department believes that this 
approach may suffer the same weakness as DOE's attempt to elaborate on 
the definition of basic model discussed in the DOE letter of August 30, 
1995.
    The Department also will consider determining whether a particular 
lamp is covered by the standards by requiring that its measured wattage 
be compared to the measured wattage of a similar covered lamp using the 
same ballast. The wattage of the covered lamp divided by the wattage of 
the lamp in question would be multiplied by the wattage marked on the 
covered lamp to determine the rated wattage of the lamp in question. 
However, this approach may not work for new products.
    The Department is concerned, however, that if it requires rated 
wattage to be determined using a high power factor ballast, 
manufacturers might be inhibited from producing certain products 
designed and marketed for use exclusively with low power factor 
ballasts. Even though there are now available a number of lamps that 
can be safely used in low power factor ballasts, and which would be 
unaffected by this proposal, the Department does not want to restrict 
unnecessarily the choices that might be available to users of low power 
factor ballasts in the future. For this reason, the Department is 
soliciting public comment and proposals on how it might use its 
discretionary regulatory authority or its authority to grant certain 
waivers or exemptions to address this possible problem. Specifically, 
DOE is interested in identifying specific technical features or 
performance or other characteristics of lamps that would provide 
reasonable assurance that such lamps would be used exclusively in low 
power factor ballasts.
    At least one manufacturer has indicated that it believes that a 
substantially reduced lamp life (e.g., 6,000 hours compared to the 
industry norm of 20,000 hours) should restrict the usage of such lamps 
to low power factor ballasts in the residential sector. But DOE is 
concerned that lamps with useful lives of 6,000 hours may still be 
widely used with high power factor ballasts. DOE is also concerned that 
accurately determining average lamp life can be difficult and time 
consuming and questions the utility to consumers of a requirement that 
may discourage manufacturers from increasing product life.
    The Department recognizes that one of the motivations for 
introducing modified 40 watt lamps is industry concern that residential 
and other users of low power factor ballasts might use 34 watt lamps in 
their fixtures, which would increase the risk of overheating and fires. 
While consumers have a range of safe alternatives to the 34 watt lamp, 
and 34 watt lamps are being labeled to warn consumers against their use 
with low power factor ballasts, DOE believes that these industry 
concerns may be valid. DOE solicits public comment on these concerns 
and how DOE might best use its regulatory authorities to ensure 
consumers are adequately protected.
    Finally, in order to better assess these issues, the Department is 
seeking more information on the size and characteristics of the market 
for lamps used in low power factor ballasts.
    e. Definition of Colored Fluorescent and Incandescent Lamp. In the 
Notice of Proposed Rulemaking, the Department defined colored 
fluorescent and colored incandescent lamps because Sections 
321(30)(B)(iii), 321(30)(C)(ii), 42 U.S.C. 6291(30)(B)(iii) and 42 
U.S.C. 6291(30)(C)(ii) of the Act contain exemptions for these lamps 
without defining them. The Department is seeking definitions of colored 
lamps which can be determined by measurement of certain 
characteristics. Therefore, the Department proposed to define colored 
incandescent and fluorescent lamps by using suitable minimum values of 
the Color Rendering Index (CRI) or correlated color temperatures (CCT). 
(59 FR 49478).
    Several manufacturers suggested that the upper limit for CRI for 
colored fluorescent lamps be increased to 40. Phillips Lighting states 
that a CRI of 40 will prevent the exclusion of gold fluorescent lamps 
which are used in printing applications. OSI also recommends that the 
acceptable CRI for amber and red incandescent lamps be raised but DOE 
believes that this is not necessary with the proposed revisions to the 
colored incandescent lamp definition because these lamps have a low 
CCT.
    In its comments to the July 19, 1995 lamp workshop, Durotest 
suggests that CCT limits for colored fluorescent lamps be less than 
2,500 deg. K or greater than 6,600 deg. K or with a CRI less than 40. 
For incandescent lamps, Durotest suggests that the CCT parameters 
should be less than 2,500 deg. K or greater than 4,600 deg. K or CRI 
less than 50. NEMA also suggests using the same CCT and CRI parameters 
as Durotest. It asks DOE to clarify in the preamble that a lamp is 
considered colored if its CCT falls outside the range above or if its 
CRI falls below the values above.
    The Department appreciates the industry suggestions for revised 
limits on CCT and CRI. DOE's original proposal would have defined 
certain green lamps as white lamps based on their CRI. This problem is 
caused by the difficulty of choosing a reference lamp of equal CCT to 
the lamp in question and because CRI was originally intended to 
characterize non-colored lamps.
    As a result of industry suggestions and comments, one option the 
Department is considering is to revise 

[[Page 7436]]
its proposed definition of a colored lamp by using a maximum value of 
CRI or a suitable band of CCT. Therefore, the Department is considering 
a definition of colored fluorescent lamp as a lamp with a CRI value 
less than 40 or a color correlated temperature not above 2,500 deg. K 
for red and yellow colors or not below 6,600 deg. K for blue and green 
colors. The Department is also considering a definition of colored 
incandescent lamp as a lamp with CRI values below 50 or a lamp color 
correlated temperature either not above 2,500 deg. K for red and yellow 
colors or not below 4,600 deg. K for blue and green colors. The 
Department believes that the measurements required to determine if a 
lamp is colored by the above definitions are minimal. The CRI is a 
required measurement for fluorescent lamps and manufacturers would only 
have to make a CRI measurement for lightly tinted incandescent lamps. 
The color temperature is derived from spectroradiometric measurements 
and this data already exists for most lamps.
    However, at the July 19, 1995 lamp workshop, NEMA proposed an 
alternative definition of colored lamps which depends on the excitation 
purity of a colored source. Excitation purity is defined as the ratio 
of two collinear distances (NC/ND) on the Commission Internationale de 
L'eclairage (CIE) chromaticity diagram. NC is the distance between the 
point representing the sample lamp and a specified reference point. ND 
is the distance between the point locating the dominant wavelength of 
the sample lamp and the specified reference point. NEMA suggests that a 
value of excitation purity greater than 50 percent would be a 
reasonable lower limiting value defining a colored lamp. NEMA claims 
that a single definition will suffice for all colors. Plotting one 
number on the x,y chromaticity diagram which shows the 50 percent 
excitation purity area marked on it will quickly determine whether a 
lamp is colored. Furthermore, NEMA requested that the Department not 
finalize the colored lamp definition until they complete their 
specification of chromaticity coordinate boundaries.
    NEMA notes that the excitation purity method proposed will not 
discriminate between clear and colored lamps with CCT's from slightly 
above 2,856 deg.K and lower. This is an inherent drawback of the 
chromaticity diagram and redefining the excitation purity limit will 
not correct it. NEMA suggests that the Department define a colored 
region around the black body locus on the chromaticity diagram as 
white. The area within the 50 percent excitation purity area is called 
pastel and lamps in this area must be marked for a specific application 
to be called colored. Although the excitation purity method fits DOE's 
criteria for a measurable colored lamp definition, the Department is 
not inclined to adopt this method because it is complicated to describe 
due to the use of three zones on the chromaticity diagram.
    As a second option, the Department is considering a colored lamp 
definition using x, y chromaticity coordinates which lie outside of the 
area bounded by the following points: (0.285,0.332); (0.453,0.440); 
(0.500,0.440); (0.500,0.382); (0.440,0.382); (0.285,0.264). These 
boundaries are taken from CIE Publication No. 2.2, Colors of Light 
Signals.
    The Department believes that defining a colored lamp by using the 
chromaticity coordinates above will satisfy manufacturers' concerns 
that lamps of low color temperature but near the black body locus 
should be considered white. Likewise, this method satisfies a DOE 
concern that valid orange and red colored lamps on or near the black 
body locus would not be considered colored.
    Since an incandescent lamp creates light by heating a filament 
``white hot,'' some lightly tinted incandescent lamps lie very near the 
black body curve on the x-y chromaticity diagram. The Department 
believes that the x-y chromaticity definition of colored lamps will 
apply to nearly all colored lamps with a few significant exceptions. 
Very lightly tinted incandescent lamps, such as jeweler's blue and 
plant grow lamps, may not meet the colored lamp definitions as they are 
currently proposed. NEMA recommends an exemption for colored 
incandescent plant lamps because there is a filter in these lamps which 
affects the yellow and green parts of the spectrum. NEMA also suggests 
that DOE require manufacturers provide a generic description of a plant 
lamp's features and require that these lamps be marketed and designated 
for plant lighting applications. In addition to the above, GE Lighting 
proposes to add that colored lamps are not suitable for general 
lighting applications. Therefore, the Department is considering an 
additional criteria in the definition of colored incandescent lamps 
that would require application specific incandescent colored lamps to 
be designated as such on the lamp and in marketing materials.
    Additionally, Durotest has urged the Department to provide an 
explicit exemption for neodymium lamps because they claim that the 
color is doped directly into the glass bulb. Therefore, the Department 
is considering specifying that incandescent lamps with lens filters 
containing 5 percent or more neodymium are colored lamps. The neodymium 
filter adjusts the light spectrum for reptile lighting applications.

4. Public Meeting Procedure

    At the public meeting, DOE will seek discussion of the points 
discussed in this notice. Should any party wish to raise any other 
matter addressed in the Interim Final or Proposed Rules, they should so 
notify DOE by February 29, 1996.
    The meeting will be conducted in an informal, conference style. A 
court reporter will be present to record the minutes of the meeting. 
There shall be no discussion of proprietary information, costs or 
prices, market shares, or other commercial matters regulated by 
antitrust law. After the meeting and period for written statements, the 
Department will consider the views presented in formulating a Final 
Rule regarding fluorescent and incandescent lamp test procedures.

    Issued in Washington, DC, February 22, 1996.
Brian T. Castelli,
Chief of Staff, Energy Efficiency and Renewable Energy.
[FR Doc. 96-4512 Filed 2-27-96; 8:45 am]
BILLING CODE 6450-01-P