International Harmonisation of CFL Specifications
First Draft Proposals for Performance Specification Tiers
The following table shows the draft specification for bare bulb CFLs with three sets of performance
requirements for each of the ‘core’ criteria relating to CFLs. These have been based on a range of
existing specification currently used by energy efficiency programs, details of which are provided
below the proposed tiers. The discussion paper which provides the rationale and background for
the development of this proposals is attached at the end of this document.
The sets of requirements reflect different levels of stringency, although for some of the criteria the
specifications may be the same or similar. The stringency of each set have been set to be
appropriate for the following applications:
Low: A mandatory minimum performance level
Medium: An endorsement label or procurement program in a developing country
High: An endorsement label for the best products available currently
These draft specifications are provided with the express intent to gaining comments from all
stakeholders. These comments should be posted either directly on the website (www.apec-
esis.org/cfl) or may be submitted by email to the Performance Standard Working Group Facilitator
Steve Beletich (email@example.com) or the Overall CFL Initiative Coordinator Stuart Jeffcott (firstname.lastname@example.org). Further, stakeholders are invited to submit other existing, or alternative proposed, performance specifications which can be used as information to raise
knowledge in the ongoing debate. Unless specifically requested otherwise, it is intended that all
draft specifications and comments received from stakeholders will be posted on the public website.
Coordinator for the Performance Specifications Working Group
Draft specification for bare bulb CFLs:
Low Medium High Starting Time 2.5 seconds 1.5 seconds 0.5 seconds Run-up Time 180 seconds 120 seconds 60 seconds
Input power(W) Initial Luminous Efficacy Input power(W) Initial Luminous Efficacy Initial Luminous Efficacy
<15 W: ? 45 <15 W: ? 45 Correlated Color Temperature (CCT) Input power(W)
?15 W ? 55 ?15 W ? 55 6500K,5000K 4000K,3500K,3000K,2700K Initial Efficacy(without cover) ? 5 to <10 46 50
? 10 to <15 52 55 ? 15 to <25 57 60
? 25 to < 60 62 65 Lumen Maintenance (2000 80% 80% 88% hours)
Premature Lamp Failure Rate ? 15% ? 10% ? 5% (1000 hours)
Lifetime Min 4000 Min 6000 Min 8000 Power Factor 0.5 0.5 0.9
Lamps must be Lamps must be The color The color marketed as one of marketed as one of Lamps must be marketed as tolerance shall be tolerance shall be ? 90% of sample must fall the following: 2700K, the following: 2700K, one of the following: 2700K, Colour Appearance within 5SDCM within 5SDCM withing 7-step ANSI MacAdam 3000K, 3500K, 3000K, 3500K, 3000K, 3500K, 4100K, 5000K, from the target from the target ellipse for that color temperature. 4100K, 5000K, or 4100K, 5000K, or or 6500K. values values 6500K. 6500K.
Colour Rendering Index CRI > 75 CRI > 80 CRI > 80 Maximum Mercury Content 10mg/lamp 5mg/lamp 5mg/lamp EMC As per local requirements As per local requirements As per local requirements Safety As per local requirements As per local requirements As per local requirements
Existing Performance Specifications
CECPELIENERGY STAR (proposed)Core Criteria
For electronic ballast:?4s; For magnetic ballast:?10sCFL must continuously illuminate within 1.5 second of being switched Time after switching on until full start (and remain lighted), average of ten samples shall be less than 1.00 second. Starting Timeon. NANABare (non-amalgam)Average of 10 samples tested Bare (amalgam), Average of 10 samples tested Run-up Timeshall be less than 1.0 minutecovered and outdoor shall be less than 3.0 minutereflectorsInput power(W)Initial Luminous EfficacyInput power(W)Initial Luminous EfficacyInput PowerInitial Luminous EfficacyInitial Efficacy(without Correlated Color Temperature (CCT)Correlated Color Temperature (CCT)Lamp power < 1050cover)6500K,5000K4000K,3500K,3000K,2700K6500K,5000K4000K,3500K,3000K,2700K10 < Lamp power < 1555
? 5 to <94650? 5 to <9465015 < Lamp power < 2560
? 9 to <155458? 9 to <155255Lamp power > 2565
? 15 to <256165? 15 to <255760
? 25 to < 606770? 25 to < 606265
The requirements of minimum initial luminous efficacy of a lamp The mininum initial luminous efficacy of a lamp model with a cover(no Input PowerInitial Luminous EfficacyInitial Efficacy(with a model with a cover are same as the above table.reflect)shall be no less than 85% of the requirements indicated in the cover)Lamp power < 1045above table.10 < Lamp power < 1550
15 < Lamp power < 2555
Lamp power > 2560N/aN/aInput PowerInitial Luminous EfficacyInitial Efficacy (Outdoor Reflectors)Lamp power < 2033
Lamp power > 2040The lumen maintenance shall be no less than 80% when burned for The luminous flux of the lamp must be ? 80% of initial levels at 40% of Average lumen output > 90.0% of initial lumen output @ Average lumen output > 80.0% of initial lumen output @ Lumen Maintenance2000hrs.model’s rated lifetime.1,000 hours of rated life, with 2 samples having lumen 40% of rated life, with 3 samples having lumen output < ??output < 85.0%75.0%
NANANAPremature Lamp Failure RateMust have a minimum rated lifetime of 6,000 hours.Must have a minimum rated lifetime of 6,000 hours.Must have a minimum rated lifetime of 6,000 hours.Lifetime
The actually power facter shall be ?0.05 of rated value.Power factor shall be ? 0.5 at maximum power.Average of sample must be ? 0.5.Power Factor
The color tolerance shall be within 5SDCM from the target valuesMust comply with IEC 60969 and the color tolerance shall be within Lamps must be marketed as one of the following: 2700K, 3000K, 3500K, 4100K, 5000K, or 6500K. ? 90% of sample Colour Appearance5SDCM from the target values.7-step ANSI MacAdam ellipse for that color temperature. must fall withing Correlated Color TemperatureColor Rendering IndexColor Rendering Index (CRI) should be at least 80.Average of the 10 samples CRI > 80.0, and no more than 2 samples can have a CRI less than 77.0Colour Rendering Index6500K,5000K?77
NALamp and lamp package must comply with any local regulations N/aMercury Contentregarding disclosure and disposal, including regulations regarding toxic
Vietnamese Procurement Specification (2005)EU Quality CharterEST SpecificationCore Criteria
CFL must continuously illuminate within 1.5 seconds of being Measured after 2 secondsStarting Timeswitched on The time to 75% of stabilized light output 100 seconds, or, the The time to 60% of stabilized light output ? 60 secondsMeasured after 60 seconds?Run-up Timetime to 80% of stabilized light output 120 seconds ?Input PowerInitial Luminous EfficacyClass A of the EU energy labelClass S - which meet the energy labelling requirements for Class A of the Initial European Directive 98/11/EC. Efficacy(without <15 W: ? 45cover)?15 W ? 55
Class L – which meet the energy labelling requirements for Class B of the European Directive 98/11/EC. Additionally they shall also meet the energy labelling requirements of at least 85% of the minimum Class A requirements of Directive 98/11/EC.N/aClass P – which are used for decorative effect lighting and which meet Initial Efficacy(with a the energy labelling requirements for Class B of the European Directive cover)98/11/EC. Additionally they shall also meet the energy labelling requirements of at least 45% of the minimum Class A requirements of Directive 98/11/ECN/aClass R Initial Efficacy (Outdoor Reflectors)The efficacy of the centre beam intensity in candelas per watt of rated power shall not be less than five times that of the incandescent filament reflector lamp it is claimed to replace.After 2000 hours of operation the luminous flux of 80% sample Bare LampsThe lumen maintenance shall Covered LampsThe lumen maintenance Lumen maintenance measured at 2000 hours life, 5000 hours life, and at Lumen MaintenanceCFLs must be ? 80% of initial levels be no less than 88% when shall be no less than 83% declared median lifeburned for 2000hrs.when burned for 2000hrs.For both types of The lumen maintenance shall be no less than 75% at 100% of rated life.lampsNAPremature Lamp Failure RateMust have a minimum rated lifetime of 6,000 hours.Must have a minimum rated lifetime of 6,000 hours. If defined as 'Long Life' must have a rated Lifetime of Type 'S' lamps ? 12,000hrs. Lifetime of Type 'L' & 'P' ? Lifetimelife 12000 hrs6,000hrs.?Power factor shall be ? 0.5 Power factor shall be ? 0.55 or ? 0.9 for lamps claiming to be high power factor Power Factor
The CFLs supplied shall be “daylight” color temperature (5500 The rated Correlated Colour Temperature (CCT) shall lie between 2650K Colour AppearanceK to 6500 K). and 2800K. Color Rendering Index (CRI) should be at least 80.CRI > 80CRI > 80Colour Rendering Index
International Harmonization of CFL Specifications
Discussion Paper Vs2 (Altered to reflect comments from workshop in Korea)
CFL Performance Specifications
(Working Group 2)
Mark Ellis, Mark Ellis & Associates, Australia
David Fridley, Lawrence Berkeley National Laboratory, United States
22 November 2005
The performance characteristics for compact fluorescent lamps (CFLs) which are important to
consumers include more than just energy efficiency, and for this reason most programs (MEPS,
Labeling and Certification, Bulk Procurement, etc) include a set of criteria covering a range of
performance characteristics (a performance specification) which must be met by a lamp to
comply/qualify. Lamp life, lumen maintenance and start-up time are examples of these „other‟
performance characteristics which are typically covered by specifications.
As documented elsewhere (eg. Conference proceedings of Right Light 6), there are now over 20 energy
efficiency programs for CFLs around the world that have been developed over recent years. Some of
these programs cover different performance characteristics, and the thresholds for each characteristic
varies, so that there exist many performance standards, many of which are similar but not exactly the
same. As the market for CFLs has expanded to become truly global, the presence of such a
proliferation of specifications complicates the compliance tasks for CFL manufacturers and probably
increases the final costs to consumers. A separate project is attempting to quantify the impact on CFL
prices as a result of these compliance issues.
While it is recognised that any jurisdiction has the right to set individual performance specifications,
the benefits of reducing the costs of CFLs to consumers without sacrificing quality (and thereby
increased penetration rates) is a compelling reason for rationalizing the number of different
specifications. Further, it is likely that enforcement costs for authorities could also be reduced, or at
least targeted more effectively.
Having a small number of clearly identified sets of performance specifications available to the
international community may also appeal to jurisdictions which do not current run programs to promote
CFLs, and encourage them to launch new initiatives based on the limited number of established
performance specifications due to the associated reduced costs/skills required to establish a program.
At the Shanghai launch of the International CFL Initiative, it was agreed that this working group should
examine performance specifications for CFLs:
“Through 2005 and part of 2006, develop a number of performance specifications for
self-ballasted CFLs of increasing stringency. This framework or scale facilitates any
interested party (government, industry-operated voluntary scheme or other proponent)
to select the performance requirements that best fit their particular needs (potentially as
minimum performance requirements or to indicate better or even best-available
technology).Any CFL scheme would still determine if and when any of these
performance specifications should operate.”
This paper discusses what is involved in setting performance specifications, highlights issues which
require resolution and presents draft specifications for consideration.
We welcome the views of manufacturers and suppliers on the current cost implications of meeting a
range of performance specifications in their markets.
OUTLINE OF PROPOSAL
A performance specification comprises a set of thresholds for identified criteria. In view of the range
of potential uses for performance specifications, (including minimum energy performance regulations,
procurement policies, endorsement labeling) it is considered that more than one set of performance
specifications will be required.
Such a tiered performance system also enhances the capability of nations wanting to use a mix of both
minimum performance standards and higher efficiency requirements for product endorsement.
The task of this working group is therefore to propose several sets of specifications, which represent
increasing levels of stringency. These could be conceived as follows:
? The lowest specification may represent a level suitable for adoption as a mandatory minimum
performance threshold in the current market.
? The next specification might be suitable for a future minimum performance level, perhaps in three
to four years time.
? A third level may represent the requirements of an endorsement programme which aims to have a
high number of compliant products.
? The most stringent level might be set to represent current best practice, met by only a small
proportion of current models.
Further improvements in technology over time may warrant a higher stringency level than currently envisaged, and the system should incorporate the capacity to improve and display compliance with that improvement over time.
It is intended that once these sets of performance specifications are agreed upon, they would be available for adoption by organisations responsible for deciding the specifications for individual energy efficiency programs. The actual adoption of these performance specifications by an individual program would be voluntary as outlined in the section below.
HOW WILL THE RATIONALIZATION OF SPECIFICATIONS OCCUR?
It should be stressed that there is no intention to impose new specifications on jurisdictions who currently run energy efficiency program for CFLs, or on those that wish to do so in the future. The selection of specifications is entirely the responsibility of each individual jurisdiction, and the international harmonization project will not alter that in any way.
However, most energy efficiency programs for CFLs periodically revise their specifications every few years, and at this point there is the opportunity for each jurisdiction to consider whether to adopt one of the sets of internationally agreed performance specifications proposed here.
Similarly, when a new program is being established, the proponents can choose the set of internationally agreed specifications that best meets its requirements.
The range of criteria included in performance specifications are generally designed to address the concerns of consumers, who typically compare the performance of CFLs with that of incandescent lamps. For example, issues such as lamp lifetime are important to a consumers‟ assessment of value, since the higher capital cost of a CFL is only repaid by energy savings if longer lifetime claims are met.
Most existing programs include the following criteria in their performance specifications:
? Efficiency level
? Lumen Maintenance
? Lifetime (both traditional lifetime and switch withstand)
? Start-up and Run-up time
? Colour rendering
? Colour temperature
Some programs also specify:
? Power Factor
? Starting temperature
? Cold start ability
? Lifetime under rapid cycle stress
? Variation of luminous flux with lamp orientation
? Marking requirements re: GLS equivalence, Correlated colour temperature and Lifetime (hours
? Mercury content of lamps
? Light distribution
A new criteria, “early failure rate”, indicates the number of lamps from a sample which fail within the first 1,000 hours. This is not generally used but would provide a useful means of differentiating products.
National requirements with regard to electromagnetic disturbance (EMC), total harmonic disturbance and safety requirements are usually noted in program specifications, either through an external reference or repeated in full.
In order to meet the objective of rationalizing the number of different performance specifications, the principle adopted is that there should be a set of „core‟ criteria for which performance specifications are
identified. This would enable any jurisdiction to add further criteria as required, on top of this list of core criteria.
The recommended approach for this project is that the following criteria should be included in harmonized performance specifications.
? Efficiency level
? Lumen maintenance
? Lifetime under rapid cycle stress
? Early failure rate
? Start-up and Run-up time
? Colour rendering
? Colour temperature
? Power factor
? Mercury content of lamps
? Marking requirements re: GLS equivalence, Colour temperature and Lifetime (hours vs years)
These core criteria reflect comments at the workshop in Korea. Further consideration will be given to the inclusion of a minimum level of EMC compliance in high
ELECTRICITY SUPPLY CONDITIONS
Currently CFLs are manufactured to operate at the voltage/frequency combinations in the target market,
eg 240v/50Hz in EU, 110v/60Hz in North America, etc. While it is technically feasible to produce
lamps which operate at all common voltage/frequency combination, this is not current practice.
It is likely that some of the performance characteristics of identical lamps would vary slightly under
different electricity supply conditions (efficacy, for example). In other words, it may be technically
more difficult to achieve a higher lamp specification for one market than another.
This raises the issue of whether the performance specifications need to be different for each
However, since the extent to which efficacy (and other performance aspects such as lifetime) vary with
voltage/frequency for CFLs is not well known, it would be difficult to quantify equivalent performance
requirements for each market. Further, considering that lamps currently meet high specifications in
both Europe and the US, it is likely that the variations are in fact quite small.
In any case, since the intention is to have several sets of performance specifications, it is still possible
for a 240/50Hz country to use a different set of specifications to those adopted in the 115/60Hz country,
this largely negates the need for equivalent specifications.
At the workshop in Korea it as agreed that there is no need for different specifications for different
voltage/frequency supply conditions.
SETTING PERFORMANCE REQUIREMENTS
At this stage we do not propose to identify the specific performance specifications for each of the levels.
Detailed specifications will be proposed following discussion of the methodology in Korea.
Table 1 below illustrates how the specifications will be set: for each „Level‟ there will be a defined minimum performance requirement for each of the „core‟ criteria.
Overall, the requirements for Level 2 will be more stringent than for Level 1, however for some
individual criteria the specifications may be the same. For example, A and E may be more stringent 22
than A and E (efficacy and mercury content), although D and D (colour rendering) may have the 1121same specification.
Table 1: Illustration of Performance Specifications
Criteria Minimum Performance Requirements
Efficacy Lifetime Run-up time Colour Mercury Other criteria
rendering content Level 1 A B C D E F 111111Level 2 A B C D E F 222222
Considerable investment has been made by program designers and manufacturers in the specifications
used by existing programs, including those currently under revision. It is therefore proposed that,
where feasible, the most widely implemented specifications should be used as the basis for the
As shown in Table 2, the establishment of five levels should provide sufficient delineation in the
marketplace, while satisfying the needs of most program designers. Further levels could be added, but
a proliferation of levels will begin to erode the benefits of rationalizing the number of specifications.
Table 2 also locates the relative position of each level by identifying an equivalent program for each
level (where available). These programs could be used to set the performance specifications for each
Table 2: Levels of Performance Specifications
Tier Stringency Description Example
Lowest performance specifications: suitable for MEPS Eg. Current China MEPS level Level 1
Possible MEPS level in the future Level 2
Inclusive endorsement label/procurement Eg. Current ELI specifications Level 3
Current best practice level endorsement Eg. EU Quality Charter, China Level 4 label/procurement certification, new Energy Star
Most stringent level, future endorsement Reserved for future use Level 5 label/procurement
At the workshop in Korea it was agreed that initially three levels representing low, medium and high
performance should be sufficient initially.