“50,000 hours to L70” is one of the most familiar lines on an LED cut sheet. It is useful, but it is easy to read too much into it. L70 describes a light-output-maintenance threshold: the point at which a population of LED light sources is projected to retain 70% of its initial luminous flux. It does not, by itself, establish that every part of a complete luminaire will operate for that long.
For a commercial lighting submittal, the practical question is not whether a brochure contains an L70 number. It is whether the proposed fixture, in its intended configuration and environment, has evidence appropriate to the project’s maintenance plan and performance requirements. The IES position on LED product lifetime prediction makes the distinction clearly: LM-80 and TM-21 address LED components, while a complete luminaire also includes drivers, thermal construction, optics, and other potential failure mechanisms.
This review sequence helps a specifier use L70 as evidence without turning it into a warranty or an all-purpose lifetime claim.
Start with what L70 actually says
The number after the L is a retained-light-output percentage. L70 therefore refers to 70% of initial luminous flux; L80 and L90 are higher retained-output thresholds. A published time to L70 is a projection or rating associated with that threshold, not a guarantee that a fixture will suddenly stop working at that hour.
That distinction matters in applications where maintained illuminance is important. A high-bay installation may be operating in a warehouse that cannot easily close an aisle for relamping or lift access. An office LED panel may need to preserve a calculated workplane level through a lease cycle. In both cases, start the design with the required maintained light level and the actual calculation, then decide what light-loss allowance and maintenance interval the project needs. Do not use an L70 headline as a substitute for either exercise.
L70 also does not describe every form of change a user may notice. Lumen maintenance concerns light output. Color shift, driver behavior, individual component failures, dirt accumulation, and an altered optical appearance need their own review. The U.S. Department of Energy’s metrics and test-methods overview lists distinct IES methods for source maintenance, lamp and luminaire maintenance, chromaticity shift, and other solid-state-lighting questions; they are not interchangeable labels.
Separate LM-80 testing from the complete luminaire
LM-80 is the IES method for measuring luminous-flux and color maintenance of LED packages, arrays, and modules. It is deliberately a source-level test. The DOE’s LM-80 data study describes its analysis as using data from LM-80 testing of representative LED packages; it is not a test of every driver, heat sink, lens, or enclosure in a finished fixture.
That is still valuable evidence. A sound LM-80 report identifies the LED source, test conditions, measurement duration, and maintenance behavior used as the basis for a projection. But a fixture changes the LED’s operating context. Drive current, board temperature, heat transfer, optic material, ambient conditions, and the selected output option all affect whether source data are relevant to a specific product configuration.
Ask the manufacturer to connect the evidence rather than leaving that link implicit:
- Which LED package, array, or module is installed in the submitted configuration?
- Which LM-80 report and test condition support the stated life rating?
- What in-situ temperature or thermal measurement connects the fixture’s LED operating point to that source data?
- Does the claim apply to the selected lumen package, CCT, optic, driver, and ambient-temperature range?
- What separate evidence supports driver, optical, and complete-luminaire reliability?
Those questions are especially important for an alternate. A series family may include different outputs, distributions, and thermal loads. The same care used to match an LM-79 report and IES file to the offered configuration should be used for the lifetime evidence.
Use TM-21 as a projection method, not a marketing multiplier
TM-21 is the IES technical memorandum used to project long-term luminous-flux maintenance from LM-80 data. The IES description of TM-21 explains that it updates the recommended projection method using LM-80 test data. The calculation gives a disciplined way to extend measured behavior, but it still has defined limits.
The important review flag is the TM-21 6× limit. IES states that the method limits a reported projection to six times the duration of the collected LM-80 data. Its example is straightforward: 10,000 hours of LM-80 data cannot support a TM-21 claim beyond 60,000 hours. The limit is there because an increasingly long extrapolation has less statistical confidence; it is not an invitation to select the biggest number on a brochure.
When a cut sheet says “L70 >60,000 h,” request the supporting report or a clear statement of the underlying data duration and case temperature. A greater-than symbol can be legitimate, but it should not conceal whether the document reports a calculated value, a permitted projection, or an unsupported extrapolation. The IES position statement specifically cautions against lifetime marketing claims that exceed the method’s projection limit.
Review the installed system and maintenance plan
An LED source that maintains output well is only one part of a durable lighting system. The IES notes that complete-luminaire reliability depends on mechanisms beyond LED-component maintenance, including drive circuitry and secondary optics. For that reason, a submittal review should keep the L70 evidence alongside—not instead of—the rest of the product record.
| Review item | What to verify | Why it belongs beside L70 |
|---|---|---|
| Exact configuration | Catalog number, output setting, CCT, optic, driver, and mounting. | A family claim may not describe every offered combination. |
| Thermal and ambient conditions | Stated operating range, enclosure or insulation restrictions, and installation details. | Source maintenance data must be relevant to the installed thermal condition. |
| Driver and system support | Driver identity, control compatibility, warranty terms, and field-replacement path where relevant. | The driver and controls are separate from LED source lumen maintenance. |
| Photometry and calculations | Matching LM-79 report, IES file, and maintained-light calculation. | Initial delivered lumens and distribution determine the starting design result. |
| Maintenance record | Cleaning, inspection, replacement access, and any required emergency or controls testing. | Dirt and service realities affect the maintained installation, not only the LED source. |
For a continuous linear-light system, include remote-driver access, circuiting, and commissioning information. For an industrial installation, include the actual ambient and contamination conditions rather than assuming a clean, conditioned room. If the project requires a particular warranty, state the required term and remedy in the specification; do not infer it from an L70 rating.
Put a clear request in the submittal checklist
For a scheduled luminaire or alternate, ask for the following before treating a life claim as comparable:
- The complete submitted catalog number and selected options.
- The stated L70 (or other Lp) rating, its reported operating condition, and the supporting LM-80/TM-21 basis where the claim relies on those methods.
- A statement that identifies the LED source and ties the submitted configuration to the cited source data.
- Thermal, ambient, driver, control, and warranty information relevant to the installation.
- The matching LM-79 photometric report, IES file, and project calculation used to verify maintained light levels.
- A replacement and maintenance plan that reflects access, operating hours, and the building’s tolerance for light loss or outage.
An ENERGY STAR downlight criterion illustrates the difference between a threshold and substantiation: it defines an in-situ L70 requirement and says higher claims must be supported by a TM-21 calculator projection report. That is a useful model for submittal review even when ENERGY STAR is not the project requirement—ask both what is claimed and how it is substantiated.
The takeaway
L70 is a useful planning metric for LED lumen maintenance, not a universal promise of luminaire life. Read it with the supporting LM-80 data and TM-21 projection, respect the method’s reporting limit, and confirm that the evidence matches the selected fixture and its installed conditions. Then review drivers, optics, controls, warranty, photometry, and maintenance separately. That gives a project team a defensible basis for deciding whether a stated LED life claim belongs in the specification.