Technical specifications for CFLs
Technical specifications for compact fluorescent lamps with integral ballast
Eur Ing Ir David Cogan BSc, CEng, MIEE, MIPENZ, MIESANZ APEC Energy Efficiency Standards Coordinator 1. Introduction This report provides an outline and commentary on international specifications1 for compact fluorescent lamps with integral ballasts (CFLs) and the moves taking place to improve alignment between them. Some of the reasons for there being different requirements between regions are discussed. The report indicates the factors to consider when developing a CFL programme specification so that it is appropriate for local conditions but also aligned with international practice, and thus able to use CFLs available on the market. 2. Background Compact fluorescent lamps with integral ballasts intended as replacements for incandescent GLS lamps have been available for nearly twenty-five years. Initially they were large, expensive, and heavy due to their electromagnetic ballasts. They also had other drawbacks, including poor starting performance. However, they were still appropriate and economic for certain applications, in places where starting was infrequent, hours of use per year high and where replacement was difficult and costly. Over the years, their design has improved, electronic ballasts have been introduced, and with an increase in production volumes their price has reduced. It is now economic for households to use CFLs in place of most GLS lamps that are used in areas that are occupied for significant periods. The economy arises from the CFLs using only 20% to 25% of the electricity consumed by a GLS incandescent lamp that provides a comparable light output. In New Zealand, for example, BRANZ’s household energy end-use project has shown that lighting accounts for over 10% of domestic sector electricity use. According to the EECA Energy Database, the domestic sector electricity consumption is just over 35% of the total. CFLs use only 20 to 25% as much electricity as the lamps they replace, so if the majority of the lighting use changes from GLS incandescent to CFL, the energy saving potential is of the order of 6% of domestic sector electricity use, or 2% of total electricity use. Despite its relative inefficiency and short life, the GLS incandescent lamp produces good quality light with no unexpected problems or side-effects. Some CFLs, on the other hand, have exhibited unwanted properties. Safety aspects are covered by the established Standard IEC 60968 and its clones, while electromagnetic compatibility aspects are dealt with by CISPR 15. However, the international standard on the performance of CFLs, namely IEC 60969, specifies only the methods of test to be used when verifying some of the manufacturer’s claims.
In this report, the term “international specifications” includes not only the Standards produced by the international standards bodies such as ISO and the IEC, but includes other specifications that are recognised and used by a number of countries.
1
450019-1
Technical specifications for CFLs
As may be expected, there are a number of technical specifications in existence, and this can represent a nuisance to the manufacturers. Therefore there is a move towards aligning the specifications where reasonable and practical. The International CFL Harmonization Initiative is a global activity, with a strong lead from APEC member economies and support from industry and expert organisations, such as the Australian Lighting Council. The initiative covers test protocol, performance requirements, and compliance issues. The main source specifications are those produced by the Efficient Lighting Initiative2, the European Union, the Energy Star specification from the United States Environment Protection Agency (EPA), and the China Centre for Certification of Energy Consuming Products (CECP). (China is the major producer of CFLs.) Unlike some energy using products, it is not reasonable to expect that a single specification will cater for all needs; local conditions make certain performance parameters more or less important in different regions. Some examples: — In the Philippines, the voltage is nominally 230 V, but due to high demand and low supply capacity the actual voltage is often much less. Therefore CFLs need to be able to start under low-voltage conditions; a minimum operating voltage of 170 V is normally specified. A disadvantage of these lamps when sold elsewhere is that they tend to take a long time to reach full output, especially in low ambient temperatures. — In a few regions, including the colder regions of New Zealand in winter, CFLs used in outdoor fittings, and even some of those indoors, may be required to start in low ambient temperatures of a few degrees Celsius. The CFL design may need to be adapted to suit, but perhaps compromising other performance aspects, such as the ability to start and run at low voltage. — In some rural areas of Vietnam and Ethiopia, lighting represents over 70% of the electrical load. A change to CFLs in place of GLS incandescent lamps would enable more households to afford and install lighting, but the power quality and the capacity of the distribution system could suffer if the CFLs have poor power factor and a high harmonic content. Therefore, in the International CFL Harmonization Initiative, the initial outputs will be a single, comprehensive test protocol that specifies how performance parameters are measured, plus a model specification that will include a restricted number of limiting values for each parameter that may be selected using a “mix and match” approach to create a specification appropriate to an individual energy efficiency campaign. Not all performance parameters will be applicable to all applications. For some of the performance parameters one of the limiting values may be “no requirement”. This model specification is still being developed, so not all preferred limiting values for the performance parameters have been settled. It is likely these will be a mixture of values established by existing international specifications and values that reflect lamp performance that is now available due to recent technological development. Information about the International CFL Harmonization Initiative may be found at the website http://www.apec-esis.org/cfl/www/.
2
The Efficient Lighting Initiative (ELI) was a UN sponsored programme that took place in Argentina, the Czech Republic, Hungary, Latvia, Peru, the Philippines, and South Africa. It is now being extended and is based in China. For details see http://www.efficientlighting.net/.
450019-1
Technical specifications for CFLs
3.
Performance Parameters The present draft testing protocol being prepared for the International CFL Harmonization Initiative includes measurement procedures for the following performance parameters: (a) Starting time; (b) Run-up time; (c) Low temperature starting; (d) Efficacy; (e) Luminous flux distribution; (f) Lamp life; (g) Lamp life (accelerated lamp life test) (h) Lumen maintenance; (i) Switching withstand; (j) Colour appearance and colour rendering index; (k) Power factor (real and apparent); (l) Harmonics; (m) Interference with infra-red devices; (n) Mercury content. The initiative specification will also specify how certain comparative claims may be made, including: (o) Lamp life (p) The equivalent GLS incandescent lamp The importance of these and the limiting values that are or may be associated with them are covered individually below.
3.2 Starting Time This is the time taken for the lamp to strike and produce a continuous light output, but possibly at reduced light output. With older CFLs, there was often a pause before the lamp attempted to strike, and there may have been several attempts to strike — a series of flickers — before it finally lit. Present day lamps are less prone to this problem, but some lamps may exhibit a short delay before striking. Where the light is turned on once per day this may not pose a problem. However, in a country such as New Zealand, occupants often move from room to room in a house, requiring a fast start to provide immediate light in the room they are entering. Specified starting times are currently 4 seconds (CECP), 1.5 seconds (ELI) and 1 second (Energy Star). The unified requirements are under consideration, but a starting time of 1 second is likely to be one suggested value in the harmonized specification. Many lamps now available start virtually instantaneously, so an alternative starting time could be 0.3 second, which equates roughly to one pace into a room. However, a longer starting time of 4 seconds may need to be allowed where the lamp has to meet other special performance requirements, such as the ability to start on low voltage. 3.3 Run-up time Run-up time is the time taken for the lamp to reach full brightness, or near full brightness. The human eye operates over an extremely wide range of illumination, and does not distinguish small differences in illumination. Once a lamp produces 75% of full output, users would be unlikely to distinguish between that and full output. However, some lamps do stay “dark” for an appreciable time, perhaps taking several
450019-1
Technical specifications for CFLs
minutes before providing enough light for a user to be able to read comfortably by it. Therefore, run-up time is one of the performance parameters taken into consideration by specifications. Run-up times currently specified include the ELI requirement for the lamp to produce 75% of full output within 100 seconds; the CECP requirements for 80% of full output within 3 minutes and fully stabilized within 40 minutes; and the Energy Star requirement of being stabilized within 3 minutes. The definition of “stabilized” may vary somewhat! For testing purposes, stablization will probably be determined by taking a series of readings of lamp output at one-minute intervals until the maximum difference between any two of the most recent eight readings is less than 1% of the average of those reading. This definition obviously cannot be used to determine whether a lamp is fully stabilized at a time of three minutes! However, long run-up times are often associated with low ambient temperatures and with lamps that operate over a wide voltage range, while the testing is generally carried out at “normal” ambient temperatures and at full voltage. Under these conditions, it is probably reasonable to expect a CFL to produce a “stabilized” output — say, has a light output that does not thereafter vary by more than 5% from the subsequent average output — within 3 minutes. 3.4 Low temperature starting Normal starting tests are carried out at “normal” ambient conditions of 25±1°C. However, some lamps may be required to start in low ambient temperatures. Accordingly, a test is being introduced in which the ability of a CFL to start in a low ambient temperature of 4°C to 6°C is checked. (This value of low ambient temperature was selected for pragmatic reasons, as it means that a normal refrigerator may be used as the test chamber.) Currently the only requirement is in the Energy Star specification, which requires the supplier to state the minimum temperature at which the lamp will start. 3.5 Efficacy Efficacy — the ratio of light output to electricity consumption3 — is the performance parameter where the CFL is greatly superior to the GLS incandescent lamp. Over the range 15W to 100W, the efficacy of a GLS incandescent varies typically between 7 and 14 lumens per watt. For the same output, the initial efficacy of a CFL will be at least 40 lumens per watt and generally better than 55 lumens per watt. This means that it is possible to realise significant savings by substituting CFLs for GLS incandescent lamps. It also means that, paradoxically, efficacy is one of the less important performance parameters; savings are going to be very worthwhile in any event. It is not worth compromising other performance parameters at the expense of consumer satisfaction for small gains in efficacy. However, CFLs with good efficacy are available and established, and as this is the performance parameter of greatest interest to promoters of energy efficiency, it receives close attention. Specifications generally have a range of efficacy requirements, depending on lamp output, lamp type and lamp power. For CFLs with a lamp colour of 2700 K and output
“Efficacy” is used in place of “efficiency” as the latter is essentially a dimensionless quantity expressed in percentage or per unit terms; so for dimensioned quantities, such as lumens per watt, the correct term is “efficacy”.
3
450019-1
Technical specifications for CFLs
roughly equivalent to 75 W to 100 W GLS lamps, specified efficacies are 60 lumens per watt in the ELI and present Energy Star specifications, and 65 lumens per watt in the CECP and the proposed Energy Star specifications. Australia (and New Zealand) regulators are considering a mandatory minimum energy performance requirement of 55 lumens per watt for this range. Lamp efficacy is based on the initial lamp output, which is the output of the lamp after it has been “run-in” for 100 hours; in the early hours of operation the lamp output may vary up and down for a while before it stabilises. 3.6 Luminous flux distribution CFLs come in a variety of configurations and sizes. Some configurations are such that the peak light output will be perpendicular to the lamp axis (i.e. delivered sideways) whereas the light output from a GLS incandescent lamp is relatively even in all directions except, of course, in the direction of the base. While a different light output characteristic is not inherently a problem, most light fittings in use are designed to be suitable for the light output characteristics of a GLS lamp. A replacement lamp with markedly different light output characteristics may prove to be unsatisfactory. Therefore, it is likely to become a requirement that the CFL manufacturer provides data or other indication of the light output characteristic so that the CFL may be matched with an appropriate light fitting. 3.7 Lamp life Lamp life is defined as the time at which the portion of lamps in a sample that have failed is 50%. Some programme requirements place a limit on early failures. The minimum lamp lifetimes required by specifications at present are 5,000 operating hours in the case of CECP and 6,000 hours for ELI and Energy Star. The manufacturer may claim a longer life, in which case, of course, the claimed performance must be met. An optional standard value of 10,000 hours is being considered for the international harmonized specification. 3.8 Accelerated lamp life Some programmes allow an initial assessment of lamp life to be made using a fast switching regime. This is somewhat similar to a switching withstand test (see 3.10 below) but the two tests ought not to be confused. The full lamp life test should always be used as the definitive one. 3.9 Lumen maintenance The light output of any fluorescent lamp decreases over its operational life. This is due to a combination of causes, including absorption of arc plasma material into the phosphor layer, deposition of cathode material on the inside of the tube, and the very gradual increase of the pressure inside the arc tube due to migration of gases through the lamp material. Modern lamps exhibit better lumen maintenance characteristics than older lamps, due to the modern practice of sealing the phosphor layer to avoid the absorption of mercury within the phosphors. (This also allows the quantity of mercury introduced into the lamp to be greatly reduced.) The CECP and ELI specifications both require lumen maintenance to be better than 80% of the initial (100 hour) value after 2000 hours of operation, while the Energy Star specification requires a minimum of 80% of the initial (100 hour) output at 40% of rated lamp life. This requirement may affect the rated life claimed by the manufacturer.
450019-1
Technical specifications for CFLs
3.10
Switching withstand Historically, the applications for which a CFL was appropriate were those where the lamp was on for long periods at a time. Therefore the life test features one switching cycle during every three hours of operation. This is still appropriate for regions where a dwelling consists essentially of a single room where the light is on from dusk to bedtime. However, in developed countries there is a far greater incidence of movement between rooms and consequently of switching — in some rooms there may be several switches per hour of lamp operation. Therefore, switching withstand tests are being introduced. The Energy Star requirement is that a CFL must withstand a rapid cycle test (five minutes on, five minutes off) of one switching per two hours of rated life. Europe is arguing for the requirement to be one switching for each hour of rated life.
3.11
Colour appearance and colour rendering index Incandescent lamps produce a “natural” light that is on the locus of black body radiation. CFLs produce their light by the radiation of phosphors; depending on the mix of phosphors used, the light produced by CFLs can differ markedly, both in terms of appearance, and in how it makes the colours of objects it shines on appear to beholders (colour rendering). Appearance is to some extent a matter of user preference. Where CFLs are used to replace incandescent lamps, it is normal to specify a colour temperature of 2700 K. This is a “warm” colour and is compatible with incandescent lamps. Higher temperature colour lamps appear bluer, and are more compatible with daylight. International specifications used to specify a colour temperature range, e.g. “between 2700 K and 3000 K”. But this actually permits lamps to the same specification to have noticeably different colour appearance. Specifications now tend to require a very specific colour with a tolerance. The colour rendering of CFLs, and other fluorescent lamps, has improved markedly with the development of new phosphors. A colour rendering index (CRI) of 80 or over can be considered good quality, and is generally what is called up by specifications. A CRI of 90 or over is excellent quality, and just about acceptable for colour matching and for doing (artistic) painting by.
3.12
Power factor (actual and displacement) CFLs tend to have a low power factor unless steps are taken in the design to improve it. A low power factor means the current taken by the lamp from the supply is unnecessarily high. As an incandescent lamp is almost a pure resistance with unity power factor, replacement by CFLs will introduce a lagging power factor and out-ofphase current. Unless compensated for, the effects of this are additional load on the electricity supply system and additional system losses. Large consumers of electricity, including large hotels, may have demand charges as part of their electricity tariff; a high power factor reduces the cost of that tariff component. Generally, however, CFLs are not concentrated in a few premises. The least complicated way of addressing this problem is to specify a high power factor type of CFLs rather than installing some form of power factor control to the electricity distribution system, which would also be more expensive overall. The power factor of CFLs is an important consideration where lighting constitutes a high proportion of the connected load, and also where there are system constraints (or potential constraints) to electricity transmission. There are a few places in large metropolitan areas overseas where the distribution system itself introduces a leading
450019-1
Technical specifications for CFLs
power factor component. In such a case some lagging load helps to provide a balance. Because of this, the international harmonized specification will feature alternative power factors of 0.5 and 0.9. But in the New Zealand context, a high power factor for CFLs is indicated not only to obtain the greatest system savings but also to help with providing a satisfactory harmonic characteristic. There is a relationship between the harmonic content and the upper bound of the associated power factor. Displacement power factor is derived from the phase difference between the voltage and the fundamental (the 50Hz component) of the current. It may be thought of as what the real power factor would be if harmonics were ignored. Determination can be relatively straightforward by observation of an oscilloscope trace, although in some instances the shape of the current trace can obscure the phase angle difference between voltage and current. 3.13 Harmonics CFLs with electronic ballasts typically draw irregular current from the supply, resulting in a complex current waveform and the production of harmonics — higher frequency currents that cause additional heating in the supply and also interfere with wireless devices and, in New Zealand, with ripple control systems. Harmonics may also interfere with or even damage other equipment, especially electronic appliances. The generally accepted standard for limitation of harmonics from CFLs is the international Standard IEC 61000-3-2. The CECP specification details the limit for each individual harmonic, but upon examination it is found that this specification is actually aligned with the IEC requirements. At present the Energy Star specification is slightly different, being merely a recommendation of 32% total harmonic distortion, but again this is fairly compatible with the requirements of IEC 61000-3-2. Harmonics can contribute to a degradation of power factor, with the relationship being set by the equation: 1 pftrue ≤ pfdist = 1 + (THDI / 100) 2 Thus the true power factor of a lamp with total harmonic distortion of 35% cannot be better than 0.95. 3.14 Interference with infra-red devices There have been cases of a CFL mimicking an infra-red remote control and causing unwanted changes to other appliances, such as changing channels on a television. This is probably related to poor harmonic performance, but is still under investigation. A test to detect any tendency for a CFL to act in this way is under development. 3.15 Mercury content CFLs use a small amount of mercury in gaseous form within the tube where it conducts the electric arc that excites the phosphors. The amount is now much smaller than for earlier CFLs due to the technology of providing a sealing layer to the phosphors to prevent them absorbing the mercury. This feature also assists in providing a good lumen maintenance characteristic. In addition, the amount of mercury in the lamps is small enough not to represent an environmental hazard if the CFL is disposed on in a landfill. Environmental concerns have been instrumental in having limits placed on mercury content of all fluorescent lamps, and in devising a procedure to check on the amount of mercury used in each lamp.
450019-1
Technical specifications for CFLs
3.16
Stated lamp life If claiming a lamp life in years, the practice used by Energy Star is for these to be based on lamp usage of 3 hours per day, i.e. 1095 hours per year. This may explain the Energy Star choice of a minimum operating time of 6000 hours (or nominally 5 years) instead of the previously common value of 5,000 hours, which would qualify for a claim of only four years.
3.17
GLS Equivalence One area where there is a noticeable misalignment between specifications concerns the claims for GLS equivalence that may be made. As may be seen from Figure 1, the ELI and European specifications permit comparison to be made on the basis of initial (100 hour) figures, with the ELI values being a bit more lenient in favour of CFLs. However, the output from CFLs does diminish with use, and the average output throughout life of a CFL is going to be less — see 3.9 above. The Energy Star requirement appears to allow for an average lumen maintenance of 0.85, and initially a CFL that complies with the Energy Star requirements will be brighter than the equivalent GLS lamp. This practice avoids user dissatisfaction in the later period of the CFL life, when its output could be around two thirds of its initial output. (GLS lamps also suffer some degradation of light output, due partly to deposits evaporated from the filament settling on the inside of the glass envelope, and partly to external dust setting on the lamp. But for a GLS lamp used base-up the former effect is not so severe as for a CFL, and the latter effect may be countered by cleaning or, more usually, by replacing the lamp when it fails.) There is a powerful school of thought in New Zealand that considers claims of equivalence ought to be made on the basis of average light output, and not on initial light output. In this case, claims of equivalence ought to follow the Energy Star requirement, or else be based on the light output at 40% life (which approximates to the
2500
2000 Australia Output lumens Energy Star 1500 ELI ECL/SEAL 1000 ECNZ Incandescent 500
0 0 20 40 60 80 100 120 140 160 Incandescent equivalent
average light output). Figure 1: Permissible claims for equivalence of CFLs with GLS incandescent lamps
450019-1
Technical specifications for CFLs
In the future, once CFLs have largely replaced incandescent lamps, the concept of equivalence will become meaningless, and lamp output will have to be indicated by another method. The most obvious one will be the output in lumens. It is notable that even now the CECP specification does not include GLS equivalence as a way of denoting lamp output. 4. Summary of technical requirements Table 1 shows the salient requirements of the international specifications for CFLs in summary form. Table 1: Summary of technical requirements of international CFL specifications
APEC (proposed basic) Starting time Under consideration 75% in 100 secs No requirement 55 lm/W Under consideration >6000 h ≥ 80% L100 After 2000 hours Under consideration ≥ 10,000 h ≥ 80% L100 at 40% life APEC (proposed enhanced) 1 sec 75% in 100 secs and stabilized within 3 mins Starts Energy Star CECP ELI Remarks Starting time.
4 secs
1.5 secs 80% in 3 mins stabilized 40 mins
Run-up time
stabilized within 3 mins Min temp to be stated 65 lm/W — 6,000 hours 90% at 1000 hours 80% at 40% life
75% in 100 secs Test under development
Low temperature starting Efficacy (Applicable to 15 & 20 W 2700K lamps) Luminous flux distribution Lamp life
60 lm/W (being raised to 65) — 5,000 hours 80% at 2000 hours
65 lm/W — 6,000 hours 80% at 2000 hours
60 lm/W
To 50% failures
Lumen maintenance
Switching withstand
≥ 1 per hour of rated life
Under consideration
1 per 2 hours of rated life
—
—
Europe is moving to a requirement of ≥ 1 per hour of rated life
Colour appearance
—
Tolerance 5 SDCM
Was 27003000K, will be 2700K, 3000K, 3500K, 4100K, 5000K, or 6500K, tolerance 7-step MacAdam ellipse 80 80 80 Procel requires 0.92
and colour rendering index Power factor apparent) (real and
80
82
0.5 / 0.9
0.9 / 0.9
0.5 / — 32% total harmonic distortion recommended
0.5 / —
0.5 / —
Harmonics
IEC61000-3-2
IEC61000-3-2
IEC61000-3-2
IEC61000-3-24
4
Required by Latvia, Hungary and Czech Republic
450019-1
Technical specifications for CFLs
APEC (proposed basic) Interference with infra-red devices Mercury content Lifetime claims No requirement 5 mg
APEC (proposed enhanced) Complies
Energy Star
CECP
ELI
Remarks Test under consideration
—
—
—
5 mg
—
—
—
GLS equivalence Lamp position
No requirements
To be based on 3 hours use per day — See Figure 1 — Advise if any position results in >5% decrease of output from base-up position
Tested base-up
5.
Selection of requirements for New Zealand CFLs In devising a specification for CFLs to feature in household lighting energy efficiency programmes in New Zealand, the prime aspects to be considered are: i) The programme objectives; ii) The special or unique features of the New Zealand electricity system; and iii) The usage patterns and habits of consumers..
5.1 Objectives
5.2 Situation New Zealand, like many other jurisdictions, is likely to experience constraints in both generating capacity and, in certain areas, transmission capacity at peak times. A reduction of household electricity use for lighting would provide peak load reduction as well as overall energy savings. To achieve “market transformation” that will permanently change the prime household lighting source from GLS incandescent to CFL, it is essential that consumers are satisfied with the product and the service it provides, and also that the power quality attributes of CFLs used in New Zealand are acceptable to the electricity supply companies. 5.3 Features of technical specification As New Zealand consumers are accustomed to GLS incandescent lamps, and as CFLs are intended to supplant them, the lamp selected must mimic an incandescent lamp in terms of size, shape, light output, lamp colour, colour rendering, start time and run-up time. These parameters must be specified appropriately. To avoid longer-term consumer disappointment, the light output from CFLs nearing their end of life must still have a light output similar to the incandescent lamps replaced. Since it is proposed that significant numbers of CFLs are to be installed in households across New Zealand, the quality of the electricity supply could be affected unless the power quality attributes of harmonics and power factor are carefully managed. A slack specification of these parameters would result in the full benefits not being attained,
450019-1
Technical specifications for CFLs
and increase the likelihood of problems such as interference with the ripple control system5 or causing problems to occur in electronic equipment. 5.4 Recommended specification For the New Zealand situation, the following features are recommended for a technical specification covering compact fluorescent lamps with integral ballast to be supplied for major household lighting energy efficiency campaigns. 5.4.1 Basic requirements The CFLs must comply with the safety requirements of IEC 60968 or AS/NZS 60968 (which is the local version) and with the EMC requirements from AS/NZS CISPR 15. 5.4.2 Starting time To mimic GLS lamps, which produce light virtually instantaneously, the fastest starting time reasonably possible is required. A start time of 1 second is the most stringent currently specified, but suppliers ought to be asked if they can improve on this. 5.4.3 Run-up time Again, the fastest time that is reasonable to specify is recommended. This would at present be the Energy Star requirement of being fully stabilized within 3 minutes. 5.4.4 Low temperature starting This requirement is appropriate for New Zealand, as some areas of a significant number of household dwellings are effectively unheated, and while one would hope that initially CFLs will be used in the areas of main use which are more likely to be heated, it is to be hoped that CFL use will become general and thus they will penetrate to unheated areas as well. However, the requirement is not yet in any published standard, so the most that can reasonably be asked for is for the supplier to state the minimum temperature at which the lamp will start (in line with Energy Star requirements) and for preference to be given to those lamps that have a minimum starting temperature of 6°C or less. 5.4.5 Efficacy While high lamp efficacy is desirable, it ought not to be demanded to the extent of jeopardising other performance parameters. However, efficacy is not closely coupled with other parameters, apart from the possibility of too high a specified value restricting the choice of CFL. While a value of 65 lumens/watt is possibly obtainable, a specified value of 60 lumens per watt in line with ELI and present Energy Star requirements is reasonable. 5.4.6 Luminous flux distribution While the test for luminous flux distribution is well established, the metrics for determining whether a particular distribution is acceptable or not have not been developed. It is probably best to cover the requirement of having a light output distribution similar to that from a GLS incandescent lamp by specifying a spiral tube configuration. This also covers some of the dimensional requirements for using the CFL in existing light fittings.
Having CFLs that regularly turn off electric water heaters may help to conserve electricity, but is hardly likely to be appreciated by consumers!
5
450019-1
Technical specifications for CFLs
5.4.7 Lamp life The “normal” lamp life expected, based on a 3-hour switching cycle, is 6,000 hours of operation. CFLs with longer lamp life are available, and providing the lamp can meet life-related requirements, such as lumen maintenance at 40% of lamp life, then obviously “more is better”. Manufacturers have 10,000 hours as a common rated value, but to what extent this is proven is not entirely certain. Therefore specifying a lamp life between 6,000 and 10,000 hours, say 8,000 hours, as a minimum is reasonable. 5.4.8 Lumen maintenance Assuming a specified lamp life of more than 6,000 hours, the ELI and CECP requirement of lumen maintenance being better than 80% at 2,000 hours could result in unsatisfactory performance towards end of lamp life. Therefore, the Energy Star requirement, for a lumen maintenance of at least 80% at 40% life, is to be preferred. 5.4.9 Switching withstand Given the way that people in developed economies tend to switch lights frequently per hour of use, this parameter is important. However, lamps may not yet have been tested to more than the one switching for each two hours of rated operational life as required by Energy Star. There is also some debate over the switching cycle to be used for the test. Energy Star specifies five minutes on and five minutes off, while others are advocating a cycle of fifteen seconds on and four minutes forty-five seconds off. The latter test is, of course, faster to complete. At present, it is reasonable to require compliance with the Energy Star requirement of surviving a test of one switching cycle per two hours of rated life, but to indicate that a better performance would be preferred. 5.4.10 Colour appearance and colour rendering index New Zealanders are used to GLS incandescent lamps, and the appropriate colour CFL to be used is therefore one with a colour appearance of 2700 K and a colour rendering index of at least 80. 5.4.11 Power factor (real and apparent) As discussed in 5.3 above, a high power factor is important in most local situations, so the appropriate value to be specified is a minimum of 0.9 (real). The displacement power factor needs to be in line with this requirement, so a value of 0.9 is appropriate for that as well (although in practice the displacement power factor is higher than the true power factor). In some regions, it may be that local conditions allow the use of lower power factor lamps, in which case a minimum true power factor of 0.5 could be specified. But care would need to be taken to avoid having lamps with low power factor sold where a high power factor is required. And while lamps with low power factor probably have a slightly lower cost, the differential is probably less than the expense to suppliers of holding two ranges of lamps. 5.4.12 Harmonics If CFLs are to be introduced to New Zealand in large quantities, then it is essential that they do not produce harmful amounts of harmonics that may interfere with the ripple controls that still operate or that may introduce additional line losses. As a minimum the lamps need to comply with IEC 61000-3-2 or the Australian and New Zealand
450019-1
Technical specifications for CFLs
equivalent. In addition, it would be prudent to place a limit on the crest factor; a limit of 1.7 would be in line with the other requirements for harmonics. 5.4.13 Interference with infra-red devices No test is currently available for this, but as the problem is presumably connected with the presence of harmonics, the requirements recommended for harmonics and power factor will probably minimise the risk of this rare phenomenon occurring. 5.4.14 Mercury content A limit of 5 mg of mercury per lamp is recommended. However, as this requirement is not included in other specifications, it is more reasonable to ask the manufacturer for a statement of whether the amount of mercury per lamp is more or less than 5 mg, and to give preference to lamps that have less than that value.
450019-1
Technical specifications for CFLs
APPENDIX — Recommended technical specification This outline specification is based on existing and planned international specifications. Where there is a choice of values for a particular performance parameter, the value selected is that which is more suitable for conditions in New Zealand. In some cases, where the performance of available CFLs can be better than specified, suppliers are encouraged to offer lamps with an improved performance. This specification thus requires CFLs that are both available on the international market and appropriate for New Zealand conditions.
Safety The CFLs must comply with the safety requirements of IEC 60968 or AS/NZS 60968 The CFLs must comply with AS/NZS CISPR 15. 1 second The CFLs shall be fully stabilized within 3 minutes Suppliers are to state the minimum temperature at which the CFL will start. (If this is less than 6°C, suppliers may state “less than 6°C”) Lamp efficacy measured after running for 100 hours shall be a minimum of 60 lumens per watt. Spiral tube
Electro-magnetic compatibility Starting time Run-up time Low temperature starting
Suppliers are encouraged to offer a shorter starting time. Energy Star ® test
Efficacy
Lamp configuration
Lamp life
Lamp life shall be a minimum of 8,000 hours
Suppliers are to state leading dimensions (Lamp length and maximum diameter) Testing may be to either the IEC 60969 regime (2 hours 45 minutes on, 15 minutes off) or to the Energy Star® regime of 3 hours on, 20 minutes off. Suppliers may offer a longer rated life.
Lumen maintenance
Switching withstand
Colour appearance Colour rendering index Power factor (real)
Lumen maintenance shall be better than 80% of initial (100 hour) output at 40% of rated lamp life. Lamps shall survive a rapid switching test of 5 minutes on five minutes off with the number of cycles equal to half the rated life in hours. 2700 K 80 or better 0.9 or better
Refer Energy Star® for test Suppliers may state ability to withstand a test with a greater number of cycles. Tolerance to be stated, either 5 SDCM
or 7-step MacAdam ellipse
Displacement power factor
0.9 or better
May consider 0.5 or better at the discretion of the local electricity lines company May consider 0.5 or better at the discretion of the local electricity lines company
Harmonics Crest factor Mercury content
CFLs shall comply with IEC 61000-3-2 Maximum of 1.7 5mg per lamp
Suppliers to state the maximum amount of mercury per lamp. (Preference shall be given to lamps with 5mg of mercury or less.)
450019-1
Technical specifications for CFLs
6.
References and sources i) AS/NZS 4782.1:2004 Double-capped fluorescent lamps - Performance specifications - General (IEC 60081:2000, MOD), and AS/NZS 4782.2:2004 Double-capped fluorescent lamps - Performance specifications - Minimum Energy Performance Standard (MEPS) (While this standard relates to tubular fluorescent lamps, certain of the tests and requirements are also relevant to compact fluorescent lamps) ii) European Harmonised Standard EN 60064:1995 Tungsten filament lamps for domestic and similar general lighting purposes iii) AS/NZS 60968:2001Australian/New Zealand Standard™ Self ballasted lamps for general lighting services—Safety requirements (IEC 60968:1988, MOD) iv) IEC 60969 Ed. 1.2 (Bilingual 2001): Self-ballasted lamps for general lighting services - Performance requirements See also:— AS/NZS 60969:2001: Self ballasted lamps for general lighting services Performance requirements v) IEC 61000-3-2 Amd.1 Ed. 2.0 (Bilingual 2001)Amendment 1 - Electromagnetic compatibility (EMC) - Part 3-2: Limits - Limits for harmonic current emissions (equipment input current <= 16 A per phase) See also:— AS/NZS 61000.3.2:2003 Electromagnetic compatibility (EMC) - Limits Limits for harmonic current emissions (equipment input current less than or equal to 16 A per phase) vi) CISPR 15 Ed. 7.0 (Bilingual 2005): Limits and methods of measurement of radio disturbance characteristics of electrical lighting and similar equipment. See also:— AS/NZS CISPR 15:2002: Limits and methods of measurement of radio disturbance characteristics of electrical lighting and similar equipment (CISPR 15:2000, MOD) vii) EECA Energy Database http://www.eeca.govt.nz/enduse/index.aspx accessed on 30 November 2005 viii) ENERGY STAR® Program Requirements for CFLs — Partner Commitments; First Draft Revision dated 20050830 ix) European Compact Fluorescent Lamps Quality Charter, 25 February 2005 accessed from http://energyefficiency.jrc.cec.eu.int/CFL/ x) Fridley, David et al: “Harmonization of energy-efficiency labeling of compact fluorescent lamps in the United States, China, Brazil and ELI member countries” Paper presented at the RightLights6 conference, Shanghai, May 2005 xi) IFC/GEF Efficient Lighting Initiative Voluntary Technical Specification Compact Fluorescent Lamps (Revised 10 July 2002) accessed from http://www.efficientlighting.net/docs/products/tecn_esp/ELICFLSpec.pdf xii) The International CFL Harmonization Initiative, details of which are at the website http://www.apec-esis.org/cfl/www/. xiii) Isaacs N., Amitrano L., Camilleri M., Pollard A. & Stoecklein A 2002 ‘Energy Use in New Zealand Households, Report on the Year 6 Analysis for the Household Energy End-use Project (HEEP)’, BRANZ Ltd: Judgeford, November 2002.
450019-1