LED lifetime is a statistical projection, not a measured age: it states how many operating hours pass before a population of luminaires falls to a set share of its initial light output (Lx), and what share stops working altogether along the way (the abrupt failure value).
Last updated: 11 September 2026 · Written for distributors, OEM brands, contractors and specifiers buying commercial LED luminaires for UK and EU projects.
Most supplier datasheets blur the two, and many quote a figure measured on the LED package rather than on the finished luminaire. That is where lifetime disputes start: a panel sold as “50,000 hours” dims early in a warm ceiling void, or its driver fails in year three while the LEDs are fine.
The short answer: in the UK and EU, the lifetime claim that belongs on a luminaire datasheet is the rated median useful life Lx, paired with an abrupt failure value and declared at a stated ambient temperature (tq), as set out in IEC 62722-2-1. LM-80 and TM-21 are North American IES methods for the LED package. They are useful evidence behind that claim, never a substitute for it.
Below: what L70B50 means, how IEC 62717, IEC 62722-2-1 and Ecodesign Regulation (EU) 2019/2020 treat lifetime, how the IES reports map onto them, why the driver often decides real service life, and a checklist for testing any supplier’s claim.
LED lifetime terms at a glance
Seven terms carry almost every lifetime claim you will see on a European datasheet or tender. The first four are European metrics; the last three are IES reports that supply evidence for them.
| Term | What it states | Defined in | Applies to |
|---|---|---|---|
| Lx (median useful life) | Hours until 50% of operating luminaires fall below x% of initial flux | IEC 62717 / IEC 62722-2-1 | LED module or luminaire |
| LxBy (useful life) | Hours until y% of operating luminaires fall below x% of initial flux | IEC 62717 / IEC 62722-2-1 | LED module or luminaire |
| Cy / AFV | Hours until y% fail completely / percentage failed at Lx | IEC 62717 / IEC 62722-2-1 | Luminaire (mostly the driver) |
| L70B50 (Ecodesign) | Declared lifetime that sets the endurance-test pass mark | Regulation (EU) 2019/2020 | Light source |
| LM-80 data | Measured flux of LED packages over at least 6,000 h at set case temperatures | ANSI/IES LM-80 | LED package, array or module |
| TM-21 projection | Lx extrapolated from LM-80 data, capped at six times the test duration | ANSI/IES TM-21 | LED package |
| LM-79 report | Initial flux, power, efficacy, CCT, CRI and light distribution | ANSI/IES LM-79 | Complete luminaire |
What does L70B50 actually mean?
L70B50 is a useful-life rating that says: after the stated number of hours, 50% of the operating luminaires in a population still give at least 70% of their initial luminous flux, and 50% have dropped below it. “L70B50 50,000 h” therefore describes the median luminaire at 50,000 hours, not a guaranteed minimum for every unit.
Two points are often misread. First, L70 is not the moment a luminaire “dies”; it is a design threshold at which the light output is considered too low for the original lighting design. Second, the B figure counts only luminaires that still work. Units that have failed outright are handled by a separate metric, the abrupt failure value.
Lx: the median useful life
Median useful life (Lx) is the operating time after which 50% of a population of operating LED luminaires of the same type have degraded below the lumen maintenance factor x. Because B50 is built into the definition, IEC writes it simply as Lx, so “L80 50,000 h” and “L80B50 50,000 h” mean the same thing.
A higher x is the stricter statement. L90 at 50,000 h promises more maintained light than L80 at 50,000 h, which in turn beats L70 at 50,000 h. Compare claims at the same hour value, never an L70 figure against an L90 figure.
By: the share of luminaires allowed below x
By is the percentage of operating luminaires allowed to fall below x at the stated time. L80B10 50,000 h allows only 10% of units below 80%; L80B50 50,000 h allows half. B10 is therefore a tighter statement about the same population, not a sign of a different technology.
How much does B10 add in practice? LightingEurope’s analysis of manufacturer data found the flux difference between B10 and B50 to be around 1% at projected lives of up to 100,000 hours, well inside the typical ±10% tolerance on rated flux. For that reason it recommends declaring the median value Lx and treating B10 or B0 headlines with caution.
Cy and the abrupt failure value (AFV)
Time to abrupt failure (Cy) is the operating time at which y% of initially operating luminaires stop producing light entirely. The abrupt failure value (AFV) is the percentage failed at the median useful life Lx: an AFV of 10% at 50,000 h means one luminaire in ten is expected to be dark by then.
AFV is where the driver, connectors and solder joints show up. A luminaire lasts only as long as its shortest-lived critical component, so a datasheet that gives Lx without AFV tells you about the LEDs and nothing about the product. Why does that gap exist at all? Because the two numbers come from different standards families, and only one of them was written for luminaires.
How European standards define LED luminaire lifetime
European lifetime data rests on IEC performance standards, adopted in the UK and EU as EN or BS EN documents, plus the Ecodesign regulation for light sources. Together they define what a manufacturer must declare and under which conditions.
| Standard | Scope | What it asks the manufacturer to declare |
|---|---|---|
| IEC 62717 (EN 62717) | Performance of LED modules for general lighting | Rated flux, power, efficacy, CCT, CRI, lumen maintenance code, Lx and failure data for the module |
| IEC 62722-2-1 (EN IEC 62722-2-1:2023) | Performance of LED luminaires | Initial data plus rated median useful life Lx with its x and the abrupt failure value, all at a stated tq |
| IEC 63013 (EN IEC 63013) | Long-term flux maintenance projection for LED packages | Method for extrapolating package test data; the European counterpart to TM-21 |
| EN 13032-4 | Photometric measurement of LED lamps, modules and luminaires | Test conditions behind flux, intensity distribution and LDT files |
| Regulation (EU) 2019/2020 | Ecodesign for light sources and separate control gear | Declared L70B50 lifetime, endurance-test pass marks, product information and EPREL registration |
Note what IEC 62722-2-1 does not do: it defines the metrics but not how a manufacturer must calculate them. LightingEurope points out that the quality of lifetime predictions therefore varies widely between suppliers. The standard tells you what to ask for; the evidence behind it tells you whether to believe it.
The photometric code on EU datasheets
Many EU datasheets carry a short photometric code from IEC 62717, for example 840/349. The first three digits give the color rendering class (8 = Ra 80–89) and the CCT (40 = 4,000 K). After the slash come the initial color spread in MacAdam steps (3), the maintained color spread (4) and the lumen maintenance code (9, meaning at least 90% of initial flux at the end of the test period of 25% of rated life, capped at 6,000 hours). A code ending in 7 on a product sold as long-life deserves a question.
Why the tq value changes every lifetime figure
tq is the rated ambient performance temperature: the highest ambient temperature around the luminaire at which the declared performance applies. IEC 62722-2-1 requires tq to be reported even when it is 25 °C, and LightingEurope recommends data sets at a tq of at least 25 °C.
This matters in real ceilings, where a recessed panel in a closed plenum can run well above 25 °C. LightingEurope’s own example datasheet shows the same 48 W luminaire at L80 50,000 h when tq = 25 °C, but only L75 50,000 h at tq = 40 °C. If your project is warm, ask for the Lx at your tq rather than the laboratory figure.
What Ecodesign Regulation (EU) 2019/2020 requires
Regulation (EU) 2019/2020, the Single Lighting Regulation, sets ecodesign requirements for light sources and separate control gear placed on the EU market, including LED light sources built into luminaires. Great Britain applies equivalent lighting ecodesign rules under its own regulations, which sit in the UKCA technical file in the same way the EU rules sit in the CE technical file.
The 3,000-hour endurance test
Market surveillance authorities verify LED lifetime claims with an endurance test on 10 units: 1,200 switching cycles of 150 minutes on and 30 minutes off, which adds up to 3,000 operating hours at an ambient of 25 ± 10 °C. At least 9 of the 10 units must still work afterwards; that is the survival factor requirement.
The minimum lumen maintenance factor
The surviving units must also keep enough light. Their average lumen maintenance factor XLMF after 3,000 hours must reach at least XLMF,MIN, calculated from the declared L70B50 lifetime:
XLMF,MIN = 100 × exp[(3,000 × ln 0.7) ÷ L70B50] %, capped at 96.0%
| Declared L70B50 | Minimum lumen maintenance after 3,000 h |
|---|---|
| 15,000 h | 93.1% |
| 25,000 h | 95.8% |
| About 26,200 h or more | 96.0% (cap applies) |
| 50,000 h | 96.0% (cap applies) |
| 100,000 h | 96.0% (cap applies) |
The cap is the useful insight for buyers. Any declared lifetime above roughly 26,000 hours faces the same 96% pass mark, so the Ecodesign test cannot tell a genuine 50,000-hour product from an optimistic 100,000-hour claim. Compliance proves the light source is not grossly overstated; it does not prove the headline hours. For that you still need package data, a valid projection and in-situ temperatures.
Where LM-79, LM-80 and TM-21 fit for European buyers
LM-79, LM-80 and TM-21 are ANSI/IES methods from North America. European tenders and importers still ask for them because LED package makers publish LM-80 data and most projections are run with TM-21. The practical skill is knowing which European metric each report supports.
| IES document | What it covers | European counterpart | What it proves for your luminaire |
|---|---|---|---|
| LM-79 | Initial photometric and electrical test of the complete luminaire | EN 13032-4; initial values in IEC 62722-2-1 | Flux, power, efficacy and distribution at time zero |
| LM-80 | Flux maintenance of LED packages, arrays or modules at set case temperatures | Lumen maintenance testing in IEC 62717; input data for IEC 63013 | How the LED package ages, not the luminaire |
| TM-21 | Extrapolation of LM-80 data to a projected Lx | IEC 63013 | A package-level Lx, valid only at the tested case temperature |
| TM-28 | Projection of flux maintenance measured on complete lamps or luminaires | Lx in IEC 62722-2-1 | A luminaire-level Lx, where luminaire test data exist |
| No IES equivalent | Complete failures of drivers and other components | Cy and AFV in IEC 62717 / IEC 62722-2-1 | Whether the luminaire keeps working |
LM-79: initial photometry of the complete luminaire
LM-79 is the test method for measuring the initial performance of a complete LED luminaire: total luminous flux, input power, luminous efficacy, luminous intensity distribution, CCT, CRI and chromaticity, measured with an integrating sphere or goniophotometer at a stabilized ambient of 25 ± 1 °C. It produces the IES photometric file; European designers usually work from the equivalent LDT (EULUMDAT) file in DIALux or Relux. LM-79 says nothing about lifetime, because it is measured at the start of life.
LM-80: lumen maintenance of the LED package
LM-80 is the method for measuring how LED packages, arrays and modules maintain luminous flux and chromaticity over time. Samples run at controlled case temperatures (Ts), typically 55 °C and 85 °C plus a third temperature chosen by the LED maker, for at least 6,000 hours, with readings at intervals of no more than 1,000 hours.
Two limits matter. LM-80 applies to the LED package, so it tells you nothing about the optic, the driver or the heat path inside the luminaire. And an LM-80 report does not contain a lifetime: it reports measured data only. Any hour figure comes from a projection.
TM-21: the projection and its six-times limit
TM-21 is the method for projecting LM-80 data forward to a lumen maintenance life such as L70. It fits an exponential decay curve, Φ(t) = B·e−αt, to the later part of the test data: the last 5,000 hours for tests of 6,000–10,000 hours, or the last half of the data for longer tests.
Its most important rule is the projection limit. With 20 or more samples, a projection may extend to six times the test duration; with 10 to 19 samples, 5.5 times. A 6,000-hour LM-80 test therefore supports a reported life of 36,000 hours at most. When the calculated value runs past that limit, TM-21 requires it to be written as a floor, for example “L70(6k) > 36,000 h”, not as a larger number.
Projections are also tied to temperature. A TM-21 value calculated at a Ts of 55 °C only applies if the LED really runs at or below 55 °C inside your luminaire. Between tested temperatures TM-21 allows interpolation; above the highest tested temperature it gives you nothing.
TM-28: projecting at luminaire level
TM-28 projects flux maintenance from long-term tests on complete lamps or luminaires, measured to LM-84, rather than on bare packages. Where it exists it is closer to the IEC 62722-2-1 idea of a luminaire Lx, because optics, heat and drive current are already included. If package data are the only evidence, the luminaire claim depends on the next question: how hot do the LEDs and the driver actually run?
Why the driver and the Ts point usually decide service life
In service, most dark luminaires are caused by the control gear, not the LEDs. LightingEurope states that most abrupt failures in practice relate to the LED driver, and recommends that manufacturers declare the expected driver failure rate at the same hours as the Lx.
Driver life is dominated by temperature. Electrolytic capacitors, the usual weak point, roughly halve their life for every 10 °C rise in operating temperature. That is why a driver datasheet states its rated life at a specific Tc point, the marked spot on the driver case. If the Tc measured in your luminaire at its tq is hotter than the datasheet condition, the rated driver life no longer applies.
The same logic applies to the LED board. The Ts point on the package, measured in the finished luminaire, decides which LM-80 temperature and which TM-21 projection are relevant. Ask for both in-situ temperatures: Ts on the hottest LED and Tc on the driver, at the declared tq.
Dimming protocols add their own failure modes. For DALI-2 projects, see our DALI-2 LED panel light dimming guide; for driver sourcing, compare the leading LED driver brands and the main components of an LED luminaire. Color shift over life is a separate aging mode, covered in why white LED panel lights turn yellow.
How to check a lifetime claim: a buyer’s checklist
Use this table when a quotation, datasheet or tender response states a lifetime. Each row names the document to request, what to check inside it, and the red flag that should stop the order.
| Evidence | What to check | Red flag |
|---|---|---|
| Luminaire datasheet | Lx with its x, AFV at the same hours, and tq stated | Hours only (“50,000 h”) with no x, no AFV and no tq |
| LM-80 report | Package part number matches the bill of materials; drive current and Ts at or above in-situ values; sample count; test hours | Report for a different package, a lower current or a lower Ts than your luminaire |
| TM-21 calculation | Projection within six times the test duration (5.5 times for 10–19 samples); “L70(6k) > …” notation where capped | 100,000 h or more from 6,000–10,000 h of data |
| In-situ temperature test | Measured Ts of the hottest LED and Tc of the driver at the declared tq | No thermal measurement, or measured only in open air at 25 °C |
| Driver datasheet | Rated life and failure rate at a stated Tc; Tc margin inside the luminaire | Driver life shorter than the luminaire Lx with no explanation |
| Ecodesign file / EPREL | Light source registration, declared L70B50, CE or UKCA technical file | Lifetime on the datasheet differs from the declared product information |
| Warranty terms | Annual operating hours assumed; failure definition (dark unit or lumen loss) | Warranty longer than the evidence supports, or silent on hours per year |
Seven ways LED lifetime claims get inflated
Most inflated claims are not invented numbers; they are real numbers used for the wrong thing. These are the patterns to recognize:
- Package life sold as luminaire life: an LM-80/TM-21 figure presented as if it covered the finished product.
- Projection past the TM-21 limit: 100,000 hours “calculated” from a 6,000-hour test that supports 36,000 hours at most.
- The wrong temperature: a projection at Ts 55 °C for a luminaire whose LEDs run at 85 °C.
- The missing driver: a long Lx with no AFV, and a driver rated for far fewer hours at its measured Tc.
- Module data instead of luminaire data: flux or efficacy of the bare LED module stated for the luminaire, inflating both.
- A hidden tq: data at 25 °C for a product that will sit in a warm plenum or a high-bay roof space.
- B10 sold as a breakthrough: B10 or B0 used as a headline when, at the same Lx, it adds around 1% of flux.
Worked example: a 24/7 hospital corridor refit
Consider a contractor pricing a corridor refit for a UK hospital, where lighting runs 24 hours a day: 8,760 hours a year. The tender asks for EN 12464-1 maintained illuminance and a lifetime declaration. Two panels are offered:
| Panel A | Panel B | |
|---|---|---|
| Declared lifetime | L80 at 50,000 h, tq 25 °C | “LED lifetime 100,000 h” |
| Abrupt failure value | 10% at 50,000 h | Not declared |
| Time to 50,000 h at 8,760 h a year | 5.7 years | 5.7 years |
| Evidence supplied | LM-80, TM-21 within six times, in-situ Ts and driver Tc | LED package datasheet only |
Panel B looks better on paper, yet only Panel A’s claim can be checked. The designer can also use Panel A’s data directly in the maintenance factor method of ISO/CIE TS 22012 and CIE 97, where the lamp lumen maintenance factor (LLMF) comes from Lx and the lamp survival factor (LSF) from the AFV. At 50,000 h, LLMF = 0.80 and, without spot replacement, LSF = 0.90. That tells the designer how much to over-light on day one to still meet maintained lux at year 5.7.
Now contrast an office at 3,000 hours a year: the same 50,000 hours takes almost 17 years, beyond the usual refurbishment cycle. LightingEurope concludes that most indoor installations never exceed 50,000 operating hours, so for offices the maintained flux at your real hours matters more than a bigger hour figure. See how these specs fit together in our EN 12464-1 office lighting overview.
What lifetime data means for distributors, OEM brands and contractors
Distributors, wholesalers and importers. When you sell under your own label, the lifetime claim and its warranty become yours. Hold the evidence for every SKU before it enters your catalog: the LM-80 report for the exact package, the TM-21 projection, in-situ temperatures and the driver datasheet. It protects your margin when a customer disputes a failure in year four. OLAMLED supports distributors with regional protection, orders from 1 piece and SKD kits for local assembly (SKD LED panel light supply and SKD LED linear light supply). If you assemble SKD kits into your own housing, the in-situ temperatures change, so re-measure Ts and Tc.
OEM and ODM brands. Your datasheet should follow IEC 62722-2-1: Lx with its x, the AFV and the tq. Ask your manufacturer for numbers in that format, not a single hour figure. Our OEM/ODM services and the OEM LED panel light buying guide for EU distributors cover private-label documentation.
Contractors and project managers. On hospital, school and public-sector jobs the lifetime declaration is often a tender line item. Use the checklist above to reject claims that cannot be evidenced, and match annual operating hours to the warranty terms.
Lighting designers and engineers. Work from Lx at the tq of your space and from the AFV, feed both into the maintenance factor, and request LDT files alongside the lifetime data. The LED panel light buying guide for Germany shows how size, UGR and lifetime sit together in a commercial specification.
OLAMLED has manufactured commercial LED luminaires since 2009, 17 years in the trade. Our 5-year warranty page sets out the lifetime figures behind each product family. Samples ship in 3–5 working days and production runs take 20–25 working days. Browse the LED panel lights, LED linear lights, LED modular panel lights and LED tri-proof lights.
Request the lifetime evidence for your model
Tell us the luminaire and the application, for example a 600 × 600 panel for a 24/7 hospital corridor, and we will send the LM-80 and TM-21 lifetime data, the LM-79 report with IES and LDT files, the driver datasheet and the CE declaration for that model. Ask at the same time for in-situ temperature data for your configuration.
LED lifetime standards: frequently asked questions
What does L70B50 mean?
L70B50 means that after the stated hours, 50% of operating luminaires still deliver at least 70% of their initial flux and 50% have fallen below it. It describes the median unit; complete failures are covered separately by the abrupt failure value.
What is the difference between LM-80 and TM-21?
LM-80 is the test: measured flux maintenance of LED packages over at least 6,000 hours at set case temperatures. TM-21 is the calculation that projects those data to a life such as L70, limited to six times the test duration.
Is L90B10 better than L70B50?
At the same hours, yes: L90 allows 10% flux loss instead of 30%, and B10 lets fewer units fall below that level. Always compare at the same hour value, and note that the B10 versus B50 difference is typically around 1% of flux.
Which lifetime standard applies to LED luminaires in the UK and EU?
IEC 62722-2-1 for luminaire performance, IEC 62717 for LED modules, and Ecodesign Regulation (EU) 2019/2020, or its GB equivalent, for light sources. A compliant datasheet declares Lx, the abrupt failure value and tq.
Can a supplier claim 100,000 hours from a 6,000-hour LM-80 test?
No. TM-21 limits the projection to six times the test duration with 20 or more samples, so 6,000 hours supports 36,000 hours at most, reported as “L70(6k) > 36,000 h”.
Why do LED luminaires fail before their rated life?
Usually the driver fails, or the LEDs run hotter than the conditions behind the claim. Check the driver’s rated life at its measured Tc, and the LED Ts inside the luminaire at the declared tq.
Related specification guides
- The main components of an LED luminaire: chips, driver, optics and housing
- Leading LED driver brands: where most abrupt failures originate
- DALI-2 LED panel light dimming guide: control gear choices for EU projects
- Why white LED panel lights turn yellow: color shift as a second aging mode
