[Federal Register Volume 76, Number 164 (Wednesday, August 24, 2011)]
[Proposed Rules]
[Pages 52892-52899]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2011-21636]


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

DEPARTMENT OF ENERGY

10 CFR Part 430

[Docket No. EERE-2007-BT-STD-0016]
RIN 1904-AB50


Energy Conservation Program: Energy Conservation Standards for 
Fluorescent Lamp Ballasts

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

ACTION: Notice of data availability and request for public comment.

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

SUMMARY: On April 11, 2011, the U.S. Department of Energy (DOE) 
published a notice of proposed rulemaking (NOPR) proposing new and 
amended standards for fluorescent lamp ballasts (ballasts) pursuant to 
the Energy Policy and Conservation Act of 1975 (EPCA). During the 
subsequent public meeting and in written comments, stakeholders 
provided additional data and raised concerns regarding the test data 
DOE used in support of the NOPR and DOE's approach to accounting for 
measurement variation and compliance certification requirements. In 
response to several of those comments, DOE conducted additional testing 
and is publishing this notice to: announce the availability of 
additional data provided by the National Electrical Manufacturers 
Association (NEMA) and additional DOE test data; address the 
differences between the DOE test data and the data submitted by NEMA; 
describe the methodological changes DOE is considering based on the 
additional data and present efficiency levels developed using the 
revised methodology and all available test data; and request public 
comment on the updated analyses, as

[[Page 52893]]

well as the submission of data and other relevant information.

DATES: DOE will accept comments, data, and information regarding this 
notice of data availability submitted no later than September 14, 2011. 
See section VI, ``Public Participation,'' of this notice for details.

ADDRESSES: Any comments submitted must identify the notice of data 
availability (NODA) for fluorescent lamp ballasts and provide the 
docket number EERE-2007-BT-STD-0016 and/or Regulatory Information 
Number (RIN) 1904-AB50. Comments may be submitted using any of the 
following methods:
    1. Federal eRulemaking Portal: http://www.regulations.gov. Follow 
the instructions for submitting comments.
    2. E-mail: [email protected]. Include the Docket 
Number EERE-2007-BT-STD-0016 and/or RIN number 1904-AB50 in the subject 
line of the message.
    3. Postal Mail: Ms. Brenda Edwards, U.S. Department of Energy, 
Building Technologies Program, Mailstop EE-2J, 1000 Independence 
Avenue, SW., Washington, DC 20585-0121. If possible, please submit all 
items on a compact disc (CD), in which case it is not necessary to 
include printed copies.
    4. Hand Delivery/Courier: Ms. Brenda Edwards, U.S. Department of 
Energy, Building Technologies Program, 950 L'Enfant Plaza, SW., Suite 
600, Washington, DC 20024. Telephone: (202) 586-2945. If possible, 
please submit all items on a CD, in which case it is not necessary to 
include printed copies.
    No telefacsimilies (faxes) will be accepted. For detailed 
instructions on submitting comments and additional information on the 
rulemaking process, see section VI of this document (Public 
Participation).
    Docket: The docket is available for review at http://www.regulations.gov, including Federal Register notices, comments, and 
other supporting documents/materials. All documents in the docket are 
listed in the http://www.regulations.gov index. However, not all 
documents listed in the index may be publicly available, such as 
information that is exempt from public disclosure.
    A link to the docket Web page can be found at: http://www.regulations.gov. The http://www.regulations.gov Web page contains a 
link to the docket for this notice, along with simple instructions on 
how to access all documents, including public comments, in the docket. 
See section VI.A for further information on how to submit comments 
through http://www.regulations.gov.
    For further information on how to submit a comment or review other 
public comments and the docket, contact Ms. Brenda Edwards at (202) 
586-2945 or by e-mail: [email protected].

FOR FURTHER INFORMATION CONTACT: Dr. Tina Kaarsberg, U.S. Department of 
Energy, Office of Energy Efficiency and Renewable Energy, Building 
Technologies Program, EE-2J, 1000 Independence Avenue, SW., Washington, 
DC 20585-0121. Telephone: (202) 287-1393. E-mail: 
[email protected].
    Ms. Elizabeth Kohl, U.S. Department of Energy, Office of the 
General Counsel, GC-71, 1000 Independence Avenue, SW., Washington, DC 
20585-0121. Telephone: (202) 586-7796. E-mail: 
[email protected].
    For information on how to submit or review public comments, contact 
Ms. Brenda Edwards, U.S. Department of Energy, Office of Energy 
Efficiency and Renewable Energy, Building Technologies Program, 
Mailstop EE-2J, 1000 Independence Avenue, SW., Washington, DC 20585-
0121. Telephone: (202) 586-2945. E-mail: [email protected].

SUPPLEMENTARY INFORMATION: 

Table of Contents

I. Introduction
II. Additional Data
III. Comparison of NEMA-Provided Data and DOE Data
    A. NEMA Reduction Factor
    B. Sample Size
    C. Lab Accreditation
    D. Measured Versus Calculated BLE
    E. Total Lamp Arc Power Approximations
IV. Accounting for Variation and Compliance Certification Procedures
    A. Compliance Certification Requirements and Design Variation
    B. Measurement Variation
V. Efficiency Levels
    A. Functional Form
    B. Preliminary Efficiency Levels
    1. IS and RS Ballasts
    2. PS Ballasts
    3. Eight-Foot HO Ballasts
    4. Sign Ballasts
    5. Residential Ballasts
VI. Public Participation
    A. Submission of Comments
    B. Issues on Which DOE Seeks Comment
VII. Approval of the Office of the Secretary

I. Introduction

    The EPCA establishes energy conservation standards for certain 
ballasts and requires that DOE conduct two cycles of rulemaking to 
determine whether to amend the standards for ballasts, including 
whether to adopt standards for additional ballasts. (42 U.S.C. 
6295(g)(5)-(8)) To complete the first of these rulemakings, DOE 
published the 2000 Ballast Rule. 65 FR 56740 (Sept. 19, 2000). To 
complete the second rulemaking, DOE is considering amendments to the 
existing standards for ballasts and evaluating standards for additional 
ballasts.
    In April 2011, DOE published a notice of proposed rulemaking (NOPR) 
that proposed new and amended energy conservation standards for 
fluorescent lamp ballasts (hereafter the April 2011 NOPR). 76 FR 20090. 
In conjunction with the NOPR, DOE also published on its Web site the 
complete technical support document (TSD) for the proposed rule, which 
described the analyses DOE conducted and included technical 
documentation for each analysis. The TSD also included the engineering 
analysis spreadsheets, the life cycle cost (LCC) spreadsheet, the 
national impact analysis spreadsheet, and the manufacturer impact 
analysis (MIA) spreadsheet.\1\
---------------------------------------------------------------------------

    \1\ The spreadsheets developed for this rulemaking proceeding 
are available at: http://www1.eere.energy.gov/buildings/appliance_standards/residential/fluorescent_lamp_ballasts.html.
---------------------------------------------------------------------------

    DOE held a public meeting on May 10, 2011, to hear oral comments on 
and solicit information relevant to the proposed rule (hereafter the 
May 2011 public meeting). At this meeting, NEMA presented test data 
that they found inconsistent with the data collected by DOE and that 
could affect the standards established in the final rule. In general, 
NEMA's ballast luminous efficiency (BLE) values appeared to be lower 
than those obtained by DOE. These observations caused NEMA to question 
the validity of the data collected by DOE for the April 2011 NOPR. NEMA 
specifically cited lab accreditation, sample size, and calculations of 
BLE as potential sources of the discrepancies they observed. Other 
stakeholders agreed that there were discrepancies between the two data 
sets and emphasized the importance of identifying the source of the 
differences. In addition, DOE received comments on the methodology used 
to account for compliance certification requirements, design variation, 
and measurement variation. DOE also received comments on the 
appropriate shape of DOE's proposed efficiency level curves.\2\
---------------------------------------------------------------------------

    \2\ Comments referenced here are available in the docket for 
this rulemaking, which can be found at regulations.gov under docket 
number EERE-2007-BT-STD-0016.
---------------------------------------------------------------------------

    Since the publication of the NOPR, DOE has analyzed NEMA's data and 
conducted additional testing to enhance

[[Page 52894]]

its analysis. In order to incorporate these additional results, DOE has 
modified slightly its approach to the engineering analysis and thus is 
considering efficiency levels that differ from those presented in the 
April 2011 NOPR.
    DOE is publishing today's NODA to: (1) Announce the availability of 
the additional NEMA test data and the additional test data developed by 
DOE; (2) address the differences between test data obtained by DOE and 
test data submitted by NEMA; (3) describe the methodological changes 
DOE is considering based on the additional data and present efficiency 
levels developed using the revised methodology and all available test 
data; and (4) request public comment on these analyses, as well as the 
submission of other relevant information. The following sections 
describe the additional data and revised methodology in more detail. 
After considering the comments received, DOE will publish a final rule 
by October 28, 2011.\3\
---------------------------------------------------------------------------

    \3\ Under the consolidated Consent Decree in New York v. Bodman, 
No. 05 Civ. 7807 (S.D.N.Y. filed Sept. 7, 2005) and Natural 
Resources Defense Council v. Bodman, No. 05 Civ. 7808 (S.D.N.Y. 
filed Sept. 7, 2005), the U.S. Department of Energy was required to 
publish a final rule amending energy conservation standards for 
fluorescent lamp ballasts no later than June 30, 2011. The consent 
decree was later modified, requiring DOE to publish a final rule no 
later than October 28, 2011.
---------------------------------------------------------------------------

II. Additional Data

    For the April 2011 NOPR, DOE tested more than 450 ballasts to 
develop proposed energy conservation standards. At the time the NOPR 
was published, DOE posted test data to its public Web site in Appendix 
5C of the TSD. Appendix 5C contained a listing of all ballast models 
tested at DOE's primary lab for the April 2011 NOPR, including 
identifying characteristics such as lamp type operated, number of lamps 
operated, starting method, ballast factor, input voltage, and catalog 
performance value. For each ballast model, DOE also reported average 
\4\ tested values for input power, total lamp arc power, and BLE.\5\
---------------------------------------------------------------------------

    \4\ The average across several samples for each model number.
    \5\ DOE obtained these values in accordance with the active mode 
test procedure in Appendix Q1 of 10 CFR part 430.
---------------------------------------------------------------------------

    At the May 2011 public meeting, NEMA presented data collected from 
several manufacturers. These test results were contained in a power 
point presentation that was subsequently posted to the public meeting 
Web site (http://www1.eere.energy.gov/buildings/appliance_standards/residential/fluorescent_ballasts_nopr_public_meeting.html). NEMA's 
data included average BLE values from three manufacturers that were 
reduced by 0.8 percent to account for compliance certification 
requirements. Attendees of the public meeting noted that the BLE values 
of the most efficient ballast models tested by NEMA appeared to be less 
than the most efficient ballast models tested by DOE. These 
stakeholders emphasized the importance of identifying the reasons for 
the differences between the two data sets. In addition, several 
stakeholders requested that DOE provide more information, including 
data for individual ballast samples and test results from other labs at 
which testing was conducted. NEMA also noted that about 60 percent of 
DOE's test data represented ballast models with less than four tested 
samples, which is not consistent with the minimum number of samples 
required to demonstrate compliance with DOE's standards. The California 
Utilities (CA Utilities) stated that if possible, DOE should conduct 
testing of four or more samples to more accurately reflect the testing 
process that must be completed by manufacturers for certification 
purposes.
    Following the May 2011 public meeting, DOE posted to the public 
meeting Web site a more comprehensive set of test data used to develop 
the April 2011 NOPR, which specified ballasts by serial numbers, added 
round robin test results, and included results for each sample tested, 
rather than the average across several samples for each model number. 
DOE also purchased and tested additional ballasts to increase tested 
models' sample size to a minimum of four samples consistent with 
compliance certification requirements in 10 CFR 429.26. DOE also tested 
additional ballast models, particularly for sign ballasts and 
residential ballasts, to gain more market information about these 
ballasts. This NODA announces the availability of all available test 
data--the NEMA-provided data, the data utilized for the April 2011 
NOPR, and the results of additional testing conducted after publication 
of the April 2011 NOPR--on DOE's Web site: http://www1.eere.energy.gov/buildings/appliance_standards/residential/fluorescent_lamp_ballasts.html.

III. Comparison of NEMA-Provided Data and DOE Data

    At the May 2011 public meeting, NEMA presented test results for its 
highest efficiency NEMA Premium products. NEMA explained that the data 
contained in the presentation represented the mean of four or five 
samples that was then decreased by 0.8 percent to account for 
compliance requirements. NEMA stated that this reduction, consistent 
with DOE's proposed reduction to efficiency levels in the April 2011 
NOPR, was calculated using the same methods that are required to 
certify with new standards.
    In addition to their observation that the manufacturer-provided 
data was lower in efficiency than DOE's data, NEMA expressed concern 
regarding DOE's data collection methods. NEMA commented that the number 
of samples DOE tested for several ballast models was too small, 
potentially resulting in test data not representative of the mean 
efficiencies of the ballast model's population. They pointed out that 
for the majority of ballast models included in the analysis, DOE tested 
fewer than four samples, which is not consistent with the minimum 
number of samples required to demonstrate compliance with DOE's 
standards. NEMA also commented that the difference between the data it 
collected and DOE's results may be due to DOE's labs not having proper 
accreditation. Furthermore, NEMA stated that the measured BLEs reported 
in appendix 5C of the NOPR TSD were not consistent with the BLEs 
calculated by NEMA (using data from the same appendix).
    Following the May 2011 public meeting, several manufacturers 
provided the model numbers and corresponding efficiencies for the 
ballasts included in NEMA's data set. Upon receiving this information, 
DOE conducted a comparative analysis and evaluated potential sources 
for the apparent discrepancies between the DOE and NEMA data sets: The 
reduction factor NEMA applied to its average BLE values, sample size, 
lab accreditation, the calculation of BLE, and the arc powers reported 
for NEMA's results.
    After considering all of the potential sources, discussed in the 
following sections, DOE preliminarily concludes that, after removing 
NEMA's reduction factor as discussed in section III.A., the remaining 
differences between the two data sets arise primarily from normal 
measurement variation. This remaining variation generally falls within 
the expected measurement variation of  2.5 percent of the 
mean efficiency, suggested by NEMA. Additional testing has increased 
sample size such that it is consistent with compliance certification 
requirements. DOE has also confirmed that its testing was conducted in 
accordance with the active mode test procedure and that its 
calculations of BLE are accurate.

[[Page 52895]]

A. NEMA Reduction Factor

    As stated earlier, the ballast efficiencies presented by NEMA at 
the May 2011 public meeting represent the mean of four or five samples 
decreased by 0.8 percent. To calculate this 0.8 reduction factor, NEMA 
referred DOE to an analysis NEMA conducted and submitted as a comment. 
In that analysis, NEMA calculated the 0.8 percent reduction factor 
based on an application of the certification equation described in 10 
CFR 429.26. NEMA assumed that each sample set's three standard 
deviation spread was equal to five percent of the mean efficiency (2.5 
percent for design variation and 2.5 percent for measurement 
variation). NEMA then calculated a mean efficiency adjustment factor 
(for sample sizes of four and five) by inserting this standard 
deviation into the certification equation. This adjustment factor 
represented an estimate of the percent difference between the sample 
mean and the value NEMA anticipated reporting to DOE for certification.
    To understand potential discrepancies between NEMA and DOE's test 
data, it is necessary to ensure that similar calculation methodologies 
have been undertaken for the two data sets. Therefore, for the purpose 
of comparing the efficiency data, DOE removes the 0.8 percent reduction 
from NEMA's presented ballast efficiencies, resulting in values that 
represent mean tested efficiencies. These efficiency values are 
analogous to DOE's mean tested efficiencies presented in the NOPR. 
However, DOE recognizes the importance of accounting for measurement 
variation and certification requirements in establishing efficiency 
levels. Additional discussion of these issues and how DOE is 
considering addressing them is provided in section IV.

B. Sample Size

    NEMA noted that less than 40 percent of DOE's test data for the 
April 2011 NOPR represented ballast models with four or more tested 
samples. They stated that the large standard deviation in efficiency 
among DOE's samples, as well as the discrepancy in tested values versus 
catalog reported values, indicates that DOE potentially did not use a 
sufficient number of samples to calculate the mean efficiencies of the 
ballast models analyzed. The California Utilities (CA Utilities) stated 
that if possible, DOE should conduct testing of four or more samples 
per ballast model to more accurately reflect the testing process that 
must be completed by manufacturers for certification purposes.
    Since the publication of the April 2011 NOPR, DOE has conducted 
additional testing to increase the sample size of selected ballast 
models. Over 90 percent of tested ballast models now have a minimum of 
four samples. Only in those cases where models have been discontinued 
or were unavailable for purchase was DOE unable to test a minimum of 
four samples.

C. Lab Accreditation

    NEMA also commented that the difference between the data it 
collected and DOE's results may be due to DOE's labs not having proper 
accreditation. DOE notes that 10 CFR 430.25 requires testing of 
fluorescent lamp ballasts to be performed in accordance with Appendix 
Q1 of 10 CFR part 430 subpart B by test laboratories accredited by 
National Volunteer Laboratory Accreditation Program (NVLAP) or a NVLAP-
recognized organization, Underwriter Laboratories, or Council of Canada 
in accordance with ISO 17025. 76 FR 25211, 25219 (May 4, 2011). ISO 
17025 is an international standard that outlines general requirements 
for the competence of testing and calibration laboratories. NVLAP 
operates an accreditation system that requires applicant laboratories 
to be assessed against all ISO 17025 requirements.
    DOE has contacted both test laboratories utilized for DOE testing 
and verified each is properly accredited and that all testing was 
conducted in accordance with the active mode test procedure in Appendix 
Q1. However, DOE recognizes that lab-to-lab variation can still be 
present among NVLAP-accredited test labs following the prescribed test 
procedure. DOE accounts for lab-to-lab variation in the establishment 
of efficiency levels as described in section IV.B.

D. Measured Versus Calculated BLE

    NEMA identified several samples in DOE's test data for which the 
measured BLE reported in appendix 5C of the NOPR TSD was not consistent 
with the BLE calculated by NEMA. Though some of the differences were 
small, NEMA provided examples of four ballast models with differences 
up to 8 percent.
    To address the small discrepancies, DOE notes that the information 
provided by NEMA is consistent with calculating the BLE values by 
dividing the average arc power of all samples by the average input 
power of all samples. NEMA's method is not consistent with the active 
mode test procedure. In contrast, DOE's measured BLE reported in 
appendix 5C of the TSD was determined, as required in the test 
procedure, by averaging the BLE of each individual sample. Based on 
DOE's analysis, this difference in methodology accounts for the small 
discrepancies observed between the values reported in appendix 5C and 
those calculated by NEMA.
    DOE also worked to resolve the larger differences cited by NEMA in 
their presentation at the May 2011 meeting. DOE identified six samples 
with measured-versus-calculated BLE differences ranging from 7.8 to 8.0 
percentage points, which included the specific examples cited by NEMA. 
These six samples were all magnetic ballasts; in accordance with active 
mode test procedure (see Table A, Appendix Q1 of 10 CFR part 430 
subpart B), DOE calculated BLE by reducing the measured ballast 
efficiency (lamp arc power divided by ballast input power) by a 
frequency adjustment factor (1.00 for high-frequency ballasts and 
values ranging from 0.93 to 0.95 for low-frequency ballasts). These 
larger discrepancies are consistent with NEMA not including this 
adjustment factor in its calculation of BLE. Thus, DOE believes its 
measured BLE values are correctly calculated and consistent with the 
active mode test procedure.

E. Total Lamp Arc Power Approximations

    Due to the relationship between total lamp arc power and ballast 
efficiency, in the NOPR, DOE proposed establishing efficiency levels as 
logarithmic equations dependent on measured total lamp arc power. When 
NEMA plotted their test data against the DOE proposed efficiency 
levels, however, NEMA paired their ballast efficiency test data with 
approximated total lamp arc powers rather than measured arc powers. DOE 
found these approximations to be higher than typical test results for 
similar ballast types in DOE's data set, with differences as high as 
27.6 percent overall. As this discrepancy could potentially cause 
NEMA's test data to appear to have artificially lower efficiencies 
relative to DOE's efficiency levels, DOE has revised NEMA's approximate 
lamp arc powers using American National Standards Institute (ANSI) 
rated high frequency lamp arc powers to calculate total expected lamp 
arc power. These lamp arc powers better align with expected total lamp 
arc powers for similar ballast types.
    For example, NEMA associated the efficiency of a ballast with a 
normal ballast factor that operates two 4-foot medium bipin (MBP) T8 
lamps with an arc power of 55 W. To correct the

[[Page 52896]]

approximated arc power, DOE calculated the typical arc power (51 W) by 
multiplying the ANSI-specified high frequency arc wattage for an F32T8 
lamp (29 W) by the number of lamps operated (2) and the most common 
normal ballast factor (0.88). DOE used this calculated arc power when 
comparing its efficiency levels to the manufacturer-provided data as 
discussed in section V.

IV. Accounting for Variation and Compliance Certification Procedures

    In the April 2011 NOPR, DOE accounted for measurement variation and 
certification requirements by calculating reduction factors for each 
and adjusting the efficiency levels accordingly. DOE calculated a 0.6 
percent reduction factor for measurement variation by comparing the 
data from the primary laboratory, which conducted the majority of DOE's 
testing, with data from its secondary laboratory, which tested a 
limited number of identical samples. DOE applied the 0.6 percent 
measurement variation reduction to the efficiency curves so that the 
standard level could, on average, be met by ballasts tested at the less 
efficient lab. To account for certification requirements, DOE 
calculated the difference between the output of the compliance 
certification equation in 10 CFR 429.26 and the sample mean of DOE's 
test data to be 0.2 percent. As DOE's certification requirements at 10 
CFR 429.26 require manufacturers to report the lower of these two 
values, DOE reduced the efficiency levels, based on average BLEs, by 
this value. Using the data that DOE made available immediately 
following the May 2011 public meeting, both NEMA and the CA Utilities 
submitted analyses to determine how DOE's data should be adjusted to 
account for certification requirements and measurement variation.
    NEMA's analysis used an assumed design variation and a calculated 
measurement variation in the compliance certification equation to 
adjust each ballast efficiency data point. NEMA then suggested that DOE 
base its efficiency levels on these adjusted data points rather than 
mean efficiency values. Specifically, NEMA determined the mean BLE for 
each ballast model by averaging all tested values of that particular 
model. NEMA then calculated the maximum measurement variation across 
labs for each category of fluorescent lamp ballast (e.g., 4-foot MBP, 
4-foot miniature bipin (MiniBP), or 8-foot recessed double contact 
(RDC) high output (HO)). NEMA added this highest calculated measurement 
variation for each ballast type to a 2.5 percent assumed design 
tolerance to characterize the total variation. NEMA then entered these 
variations into the compliance equation to calculate a reduction factor 
based on sample size of each tested model.
    The CA Utilities also conducted an analysis on the data DOE 
provided following the May 2011 public meeting. They agreed with NEMA 
that compliance certification requirements should be considered when 
assessing whether products will meet each standard level. However, they 
pointed out that NEMA had employed methods to characterize the reported 
value that were not consistent with the requirements specified in 10 
CFR 429.26. Instead, the CA Utilities used individual samples of DOE's 
efficiency data to calculate both the sample mean and the value 
determined by the compliance certification equation in 10 CFR 429.26. 
Then, as directed by the compliance certification regulations, they 
represented reported efficiency as the lower of the two values. They 
suggested that DOE base its efficiency levels on these reported values.
    Consistent with the April 2011 NOPR, DOE recognizes the importance 
of considering the variation present in the test data when developing 
efficiency levels. DOE acknowledges that due to design variation, the 
reported value for compliance certification may deviate from the sample 
mean and must be accounted for. As described in the following sections, 
DOE is considering modifying its approach to account for variation and 
compliance certification procedures based on the comments provided.

A. Compliance Certification Requirements and Design Variation

    DOE agrees with both NEMA and the CA Utilities that standard levels 
should account for the procedures manufacturers must follow to certify 
compliance with standards. As stated earlier, 10 CFR 429.26 requires 
manufacturers to test a minimum of four fluorescent lamp ballasts and 
report the minimum of either the mean efficiency of the samples or the 
output of a compliance certification equation based on the lower 99 
percent confidence limit of the sample. The lower 99 percent confidence 
limit equation requires a calculation of the standard deviation of the 
sample set to account for design variation.
    Both the NEMA and CA Utilities approaches recommend that, in order 
to develop efficiency levels, DOE should adjust its mean efficiency 
data points to represent values similar to those manufacturers would 
report to DOE for compliance certification. However, their approaches 
differ in how they computed the standard deviation to input into the 
compliance certification equation. The CA Utilities calculated the 
standard deviation among all samples of a particular ballast model 
tested at a single lab. NEMA, however, calculated the standard 
deviation by assuming a 2.5 percent design variation and then adding an 
additional measurement variation based on DOE's lab-to-lab test data 
for each ballast category.
    DOE disagrees with NEMA's method of applying the compliance 
certification requirements. Firstly, the test procedure's compliance 
requirements direct manufacturers to calculate the standard deviation 
of the tested sample, rather than an assumed population standard 
deviation. Secondly, this calculation would likely not include data 
from more than one lab unless manufacturers chose to test their samples 
of a single ballast model at more than one location. DOE is considering 
accounting for measurement (specifically lab-to-lab) variation as a 
separate adjustment to efficiency levels as discussed below in section 
IV.B.
    The CA Utilities evaluated both the sample mean and compliance 
equation for each ballast model and compared the lower of the two, the 
reported value, to the standard level. DOE believes the CA Utilities 
approach for accounting for compliance certification requirements is 
more consistent with the procedures laid out in 10 CFR 429.26 and is 
therefore considering using this methodology in the final rule. To 
facilitate this approach, as discussed earlier, DOE conducted 
additional testing since publication of the NOPR to increase the sample 
size of several ballast models in accordance with compliance 
certification requirements. To account for both certification 
requirements, DOE has calculated a new data set which represents the 
reported value for all ballast models. DOE used these reported values 
to develop the efficiency levels described in section V of today's 
NODA.

B. Measurement Variation

    DOE is also considering revising its methodology to account for 
measurement variation, specifically lab-to-lab variation. DOE received 
test data from NEMA following the May 2011 public meeting and also 
received test data from NEMA-member manufacturers. The data from 
manufacturers allowed DOE to match NEMA test data with the same ballast 
models tested at DOE's primary and

[[Page 52897]]

secondary labs. Using the model-specific test data supplied by several 
manufacturers (representative of three different manufacturer labs) and 
DOE's BLE data (representative of the two labs used by DOE), DOE 
determined that on average, the BLE test data from DOE's primary lab 
was 0.7 percent more efficient than the average test lab. DOE 
attributes this offset to systematic lab-to-lab variation and therefore 
is considering reducing the efficiency levels by 0.7 percent so that 
they are representative of ballasts tested at the average test lab. 
This approach is slightly different than that taken in the April 2011 
NOPR, which applied a 0.6 percent reduction to efficiency levels, 
representing the average offset between DOE's primary lab and the least 
efficient lab (in that case, DOE's secondary lab). DOE believes that 
adjusting efficiency levels so that they represent the average test lab 
better characterizes the mean performance of products currently being 
sold.

V. Efficiency Levels

A. Equation

    In the NOPR, DOE proposed establishing efficiency levels as 
logarithmic equations dependent on total lamp arc power. DOE developed 
this logarithmic relationship by empirically fitting curves to 
manufacturer product lines present in DOE's test data. DOE is 
considering changing the contour of the efficiency levels for the final 
rule to better fit all of the available data. Upon analysis, NEMA's 
test data show a larger efficiency decrease at lower powers than DOE's 
data indicate. Although DOE and NEMA generally tested the same types of 
ballasts, NEMA tested more permutations of ballast factor and number of 
lamps for each product line, particularly at lower wattages. For 
example, NEMA's data contained BLE values for 1-lamp 4-foot MBP 
ballasts with both low and high ballast factors, whereas DOE's data 
included 1-lamp 4-foot MBP ballasts with only normal ballast factors. 
Therefore, based on an application of several equation forms of 
efficiency levels, DOE concluded that a power law equation fits both 
the NEMA data and DOE's data better than the logarithmic relationship 
proposed in the April 2011 NOPR. A power law equation takes the form:
[GRAPHIC] [TIFF OMITTED] TP24AU11.001

Where: Power = total measured lamp arc power

    Because the NEMA data represents the most complete product lines 
and thus may represent a more accurate depiction of a BLE-lamp arc 
power relationship than DOE's initial test data, DOE fit power law 
regressions to the NEMA test data to calculate the exponent ``C.'' For 
the instant start and rapid start (IS/RS) ballasts, DOE found the 
exponent ``C'' to be 0.25. The exponent 0.25 is also a quantity used in 
relating power to relative losses (analog of efficiency) for 
distribution transformers, and fluorescent lamp ballasts similarly 
employ transformers and inductors. The programmed start (PS) NEMA data, 
however, suggested a different exponent for ballasts that use the PS 
starting method. DOE believes that this alternate shape is attributable 
to the PS ballasts' higher fixed losses due to internal control 
circuitry and heating of lamp electrodes (cathode heating). As these 
losses are a larger proportion of total losses at lower powers, the PS 
product classes have a steeper slope across the range of wattages. 
Using NEMA's data for PS ballasts, DOE found the exponent ``C'' to be 
0.37.
    With exponents set for the two starting method categories, DOE fit 
the power law equation to the reported value data (calculated in 
accordance with 10 CFR 429.26 as discussed in section IV.A) by 
adjusting the coefficient ``B'' to delineate among criteria such as 
different product lines, ballasts that operate different lamp types, 
and other clusters in efficiency data. The most efficient (maximum 
technologically feasible) efficiency levels closely approximate the 
NOPR proposals for the highest wattages, but better follow product line 
efficiency trends at lower wattages.

B. Preliminary Efficiency Levels

    Using the methodology described in the previous section, DOE 
developed a complete set of efficiency levels for this NODA, which are 
being considered for the final rule. DOE developed power law curve-fits 
based on the DOE test data. Then to develop efficiency levels, DOE 
applied a lab-to-lab adjustment factor (derived from all available test 
data) to these curve-fits (as discussed in section IV.B). In addition, 
DOE compared the resulting efficiency levels against the NEMA data to 
confirm the impacts of the efficiency levels on product availability 
indicated by the analysis of the DOE data. The following sections 
describe the efficiency levels considered for each representative 
product class. An Excel spreadsheet summarizing these levels is 
available on DOE's Web site: http://www1.eere.energy.gov/buildings/appliance_standards/residential/fluorescent_lamp_ballasts.html. The 
final rule and accompanying TSD will include the complete downstream 
analyses on these levels and results.
1. IS and RS Ballasts
    DOE developed three efficiency levels for the IS/RS product class. 
EL1 was designed to eliminate 4-foot MBP T12 ballasts while allowing 4-
foot MBP T8 ballast and 8-foot slimline ballasts to comply with energy 
conservation standards. EL2 corresponds to a level which allows the 
highest-efficiency product lines from each of the four major ballast 
manufacturers to comply. DOE defines a full product line as spanning a 
sufficient diversity of products (spanning several ballast factors, 
numbers of lamps per ballast, and types of lamps operated). EL3 is the 
maximum technologically feasible (max tech) level which DOE defines for 
fluorescent lamp ballasts as the highest level, regardless of 
manufacturer, that is technologically feasible for a sufficient 
diversity of commercially available products. Use of those criteria 
results in an EL3 with which nearly two manufacturer product lines 
comply.
2. PS Ballasts
    DOE developed three efficiency levels for the PS product class. The 
least efficient level (EL1) was designed to eliminate the lowest 
efficiency 4-foot MBP, 4-foot T5 high output, and 4-foot T5 standard 
output PS ballasts. This also corresponds to a level at which each of 
the four major fluorescent lamp ballast manufacturers maintain a 
diversity of products. EL2 allows full product lines from two major 
manufacturers. Finally, EL3, the maximum technologically feasible 
level, was designed to represent the most efficient PS ballasts tested 
by DOE. EL3 is the highest level that allows one full line of products, 
regardless of manufacturer.
3. Eight-Foot HO Ballasts
    For the 8-foot HO IS/RS product class, DOE developed three 
efficiency levels. For this product class, DOE tested ballasts that 
operate two lamps, the most common lamp-and-ballast combination. EL1 
was designed to just allow the least efficient T12 electronic ballasts, 
eliminating magnetic ballasts. EL2 allows the least efficient T8 
ballast tested and eliminates the vast majority of T12 electronic 
ballasts. Finally, EL3 was designed to just allow the most efficient T8 
ballast tested by DOE.

[[Page 52898]]

4. Sign Ballasts
    The sign ballast market comprises primarily magnetic and electronic 
ballasts that operate T12 HO lamps. DOE tested sign ballasts that 
operate up to one, two, three, four, or six 8-foot T12 HO lamps. The 
test data showed that sign ballasts exist at two levels of efficiency. 
Therefore, DOE analyzed a baseline and one efficiency level above that 
baseline. EL1 was designed to allow a full line of electronic sign 
ballasts, including ballasts that operate one through six lamps.
5. Residential Ballasts
    In the April 2011 NOPR, DOE had proposed that both residential and 
commercial ballasts could achieve similar levels of efficiency at the 
highest levels analyzed. Based on the similarity in efficiency, DOE 
included both ballast types in the same product class. However, for the 
final rule, after conducting additional testing which indicate that 4-
lamp residential ballasts may not be able to achieve the same levels as 
commercial ballasts, DOE is considering a separate product class for 
residential ballasts. The additional data for residential ballasts is 
also available at http://www1.eere.energy.gov/buildings/appliance_standards/residential/fluorescent_ballasts_nopr_public_meeting.html. Consequently, DOE has derived and is considering two 
separate efficiency levels for residential ballasts to incorporate the 
new data. EL1 was designed to just allow the least efficient T8 
ballasts, eliminating T12 residential ballasts. EL2, the maximum 
technology feasible level, is the highest level that allows a full 
range of T8 products (including both two- and four-lamp ballasts) to 
comply.

VI. Public Participation

A. Submission of Comments

    DOE will accept comments, data, and information regarding this NODA 
no later than the date provided in the DATES section at the beginning 
of this notice. Interested parties may submit comments, data, and other 
information using any of the methods described in the ADDRESSES section 
at the beginning of this notice.
    Submitting comments via http://www.regulations.gov. The http://www.regulations.gov Web page will require you to provide your name and 
contact information. Your contact information will be viewable to DOE 
Building Technologies staff only. Your contact information will not be 
publicly viewable except for your first and last names, organization 
name (if any), and submitter representative name (if any). If your 
comment is not processed properly because of technical difficulties, 
DOE will use this information to contact you. If DOE cannot read your 
comment due to technical difficulties and cannot contact you for 
clarification, DOE may not be able to consider your comment.
    However, your contact information will be publicly viewable if you 
include it in the comment itself or in any documents attached to your 
comment. Any information that you do not want to be publicly viewable 
should not be included in your comment, nor in any document attached to 
your comment. Otherwise, persons viewing comments will see only first 
and last names, organization names, correspondence containing comments, 
and any documents submitted with the comments.
    Do not submit to http://www.regulations.gov information for which 
disclosure is restricted by statute, such as trade secrets and 
commercial or financial information (hereinafter referred to as 
Confidential Business Information (CBI)). Comments submitted through 
http://www.regulations.gov cannot be claimed as CBI. Comments received 
through the Web site will waive any CBI claims for the information 
submitted. For information on submitting CBI, see the Confidential 
Business Information section below.
    DOE processes submissions made through http://www.regulations.gov 
before posting. Normally, comments will be posted within a few days of 
being submitted. However, if large volumes of comments are being 
processed simultaneously, your comment may not be viewable for up to 
several weeks. Please keep the comment tracking number that http://www.regulations.gov provides after you have successfully uploaded your 
comment.
    Submitting comments via e-mail, hand delivery/courier, or mail. 
Comments and documents submitted via e-mail, hand delivery, or mail 
also will be posted to http://www.regulations.gov. If you do not want 
your personal contact information to be publicly viewable, do not 
include it in your comment or any accompanying documents. Instead, 
provide your contact information in a cover letter. Include your first 
and last names, e-mail address, telephone number, and optional mailing 
address. The cover letter will not be publicly viewable as long as it 
does not include any comments.
    Include contact information each time you submit comments, data, 
documents, and other information to DOE. E-mail submissions are 
preferred. If you submit via mail or hand delivery/courier, please 
provide all items on a CD, if feasible, in which case, it is not 
necessary to submit printed copies. No facsimiles (faxes) will be 
accepted.
    Comments, data, and other information submitted to DOE 
electronically should be provided in PDF (preferred), Microsoft Word or 
Excel, WordPerfect, or text (ASCII) file format. Provide documents that 
are not secured, that are written in English, and that are free of any 
defects or viruses. Documents should not contain special characters or 
any form of encryption and, if possible, they should carry the 
electronic signature of the author.
    Campaign form letters. Please submit campaign form letters by the 
originating organization in batches of between 50 to 500 form letters 
per PDF or as one form letter with a list of supporters' names compiled 
into one or more PDFs. This reduces comment processing and posting 
time.
    Confidential business information. Pursuant to 10 CFR 1004.11, any 
person submitting information that he or she believes to be 
confidential and exempt by law from public disclosure should submit via 
e-mail, postal mail, or hand delivery/courier two well-marked copies: 
one copy of the document marked ``confidential'' that includes all the 
information believed to be confidential, and one copy of the document 
marked ``non-confidential'' with the information believed to be 
confidential deleted. Submit these documents via e-mail or on a CD, if 
feasible. DOE will make its own determination about the confidential 
status of the information and treat it according to its determination.
    Factors of interest to DOE when evaluating requests to treat 
submitted information as confidential include: (1) A description of the 
items; (2) whether and why such items are customarily treated as 
confidential within the industry; (3) whether the information is 
generally known by or available from other sources; (4) whether the 
information has previously been made available to others without 
obligation concerning its confidentiality; (5) an explanation of the 
competitive injury to the submitting person which would result from 
public disclosure; (6) when such information might lose its 
confidential character due to the passage of time; and (7) why 
disclosure of the information would be contrary to the public interest.

[[Page 52899]]

    It is DOE's policy that all comments may be included in the public 
docket, without change and as received, including any personal 
information provided in the comments (except information deemed to be 
exempt from public disclosure).

B. Issues on Which DOE Seeks Comment

    Although DOE welcomes comments on any aspect of this notice, DOE is 
particularly interested in receiving comments and views of interested 
parties concerning the following issues:
    (1) The conclusion that after removing 0.8 percent NEMA's reduction 
factor and recalculating lamp arc powers, the remaining differences 
between DOE and NEMA-provided data are likely due to normal measurement 
variation;
    (2) The methodology used to account for compliance certification 
requirements and measurement variation in developing efficiency levels;
    (3) The appropriateness of using a power law equation to develop 
efficiency levels and the chosen values for the exponent ``C''; and
    (4) The efficiency levels considered.

VII. Approval of the Office of the Secretary

    The Secretary of Energy has approved publication of this notice of 
data availability.

    Issued in Washington, DC, on August 18, 2011.
Timothy Unruh,
Program Manager, Federal Energy Management Program, Energy Efficiency 
and Renewable Energy.
[FR Doc. 2011-21636 Filed 8-23-11; 8:45 am]
BILLING CODE 6450-01-P