What do L70 and L90 mean?

The L-value — a lumen maintenance threshold

The notation L70, L80 and L90 identifies luminous-flux-maintenance thresholds. In a time-based declaration, L70 sets the threshold at 70% of initial luminous flux and L90 sets it at 90%. An L-value does not by itself state the survival rate, abrupt-failure rate or service life of the complete luminaire.

The required maintained performance depends on the application, the selected lighting class and the project-specific maintenance calculation. EN 13201 does not prescribe L70 or L90 as a universal source threshold: M classes use luminance-based requirements, while C and P classes use illuminance-based requirements.

For road lighting, assess the declared luminous-flux maintenance together with the installation’s complete maintenance factor and the maintained requirements of the selected class; do not treat an L-value as a standalone compliance result.

In practice

A bare claim such as L70 > 100,000 h is incomplete unless its scope and convention are stated. Under the IEC median-useful-life convention, L70 means L70B50; that convention does not replace supporting evidence or a separate abrupt-failure declaration.
L70B10 at a stated time denotes that no more than 10% of the operating population has fallen below L70 by that time. It is not automatically a more useful procurement metric than B50; the basis used to substantiate By must be stated.
Specify the required Lx factor, time, temperature, product scope and population convention. If By is omitted, state whether the declaration uses the IEC median-useful-life convention, in which B50 is implicit.
For LED packages, arrays and modules, LM-80 data and a TM-21 calculation can support an LED-source flux-maintenance projection. LM-84 measures maintenance for LED lamps, light engines and luminaires, while TM-28 projects maintenance for LED lamps and luminaires. None of these methods, by itself, establishes complete-product reliability or a By population distribution. See the 100,000 hours article for the evidence that should accompany a declaration.

The B-value — the population dimension

By is an IEC population percentile for gradual flux degradation among operating LED modules or luminaires of the same type. It is not generated automatically by an LM-80/TM-21 projection, and the method used to substantiate a non-median By value should be identified.

B10 identifies a 10% share of the operating population below the Lx threshold at the stated time. B50 identifies the median of that population; LxB50 at a stated time is a median useful-life declaration. Products that no longer operate are excluded from this By population and belong in a separate abrupt-failure assessment, such as a Cy declaration.

Do not assume that B10 is inherently the decisive maintenance-planning value. Compare declarations on the same Lx, time, temperature, scope and evidence basis, and assess expected abrupt failures separately.

Reading a complete notation

A useful-life declaration may be written as LxBy; median useful life is commonly written as Lx with B50 implicit. The following are hypothetical LxB10 examples for an operating population, not measured values for VALDUR:

L₇₀B₁₀ > 60,000 h At 60,000 h, at least 90% of the operating population remains at or above 70% of initial luminous flux

L₈₀B₁₀ > 50,000 h At 50,000 h, at least 90% of the operating population remains at or above 80% of initial luminous flux

L₉₀B₁₀ > 36,000 h At 36,000 h, at least 90% of the operating population remains at or above 90% of initial luminous flux

These three claims describe the same hypothetical LED source at different threshold levels. Note that as the L-threshold rises (i.e., less depreciation is allowed), the hour figure falls — because it takes fewer hours to reach a smaller depreciation margin.

Which measurement and projection method applies?

ANSI/IES LM-80-21 measures flux- and colour-maintenance characteristics of LED packages, arrays and modules under documented test conditions; ANSI/IES TM-21-21 uses applicable LM-80 data to calculate and report an LED-source flux-maintenance projection. ANSI/IES LM-84-20+E1 measures optical-radiation maintenance of LED lamps, light engines and luminaires. ANSI/IES TM-28-20 provides direct and combined projection methods for LED lamps and luminaires, using LM-84 data and, in the combined method, applicable LM-80 data. These methods do not, by themselves, determine complete-product reliability or abrupt-failure life.

A TM-21 reported result is capped by the method’s applicable sample-size-dependent reporting limit, commonly 6 times the test duration for at least 20 units and 5.5 times for 10–19 units. Consequently, 10,000 hours of qualifying LM-80 data can support a reported result only to the applicable 55,000- or 60,000-hour limit, not to 100,000 h. An uncapped calculated or projected value beyond that limit is not equivalent to the TM-21 reported value, and TM-21 does not by itself establish B10.

What this means for procurement

A practical procurement checklist for lumen maintenance:

Define the required Lx factor, time, operating temperature, product scope and population convention; do not compare declarations made on different bases.
For an LED-source claim, request the LM-80 report for the exact package, array or module and the corresponding TM-21 report. Check the source identity, test duration, sample size, failures, test temperatures, drive current and selected in-situ operating point. For a lamp or luminaire claim, request the applicable LM-84/TM-28 evidence for the exact product configuration.
Use the method’s reported result, not an uncapped calculated or projected value. Verify the applicable sample-size-dependent reporting limit in the actual report; under the commonly applied 6×/5.5× limits, a reported value above 100,000 h requires at least about 16,667 h for 20 or more units or about 18,182 h for 10–19 units, together with all other method requirements.
Use source or product luminous-flux maintenance only as one input to the project’s overall maintenance factor; also document luminaire maintenance, environment, cleaning and component-reliability assumptions.
For EN 13201 road-lighting design, verify the maintained requirements of the selected class using the project-specific maintenance factor; EN 13201 does not itself make L70 or L90 the required LED-source threshold.

The full methodology behind the projection

How far can you substantiate a 100,000-hour claim? Check the exact product scope, test data, projection report and applicable reporting limit.

LM-80/TM-21 evaluates LED-source flux maintenance. LM-84 measures maintenance for LED lamps, light engines and luminaires, while TM-28 projects maintenance for LED lamps and luminaires. None of these methods, by itself, establishes complete-product service life.

Service Life & Reliability

100,000 hours — what it actually means

Source and product scope, test conditions, projection limits, and the evidence needed to assess long-hour luminous-flux-maintenance claims.

→

Summary

Lx is the luminous-flux-maintenance threshold used for gradual degradation. In an LxBy declaration, By is the percentage of the operating population allowed to fall below that threshold at the stated time; non-operating products are excluded. For example, L70B10 > 60,000 h means that, within the stated conditions and evidence basis, no more than 10% of the operating population is projected to fall below 70% of initial luminous flux by 60,000 h.

Match the evidence route to the declared scope: LM-80/TM-21 for the identified LED package, array or module, and LM-84/TM-28 for an LED lamp or luminaire. Compare the exact product identity, operating conditions, sample size, test duration and reported projection limit, and assess abrupt failures and other reliability mechanisms separately.