Lumen Depreciation: Projecting L70 and L90 LED Lifetimes accurately

Analyze L70 and L90 lumen depreciation metrics. Use TM-21 reports to accurately project the useful lifespan of commercial LED luminaires in harsh environments

Illumination Pros Editorial

May 1, 2024 16 min read

Lumen depreciation is an inevitable reality in all solid-state lighting systems. While LEDs do not typically experience catastrophic failure like traditional high-intensity discharge (HID) or incandescent sources, their light output gradually diminishes over time due to the degradation of the semiconductor die, the phosphor coating, and the encapsulant materials. This gradual reduction in luminous flux fundamentally dictates the useful lifespan of a luminaire, necessitating precise calculation methods to ensure that target illuminance levels are maintained throughout a facility’s operational lifecycle.

In commercial and industrial environments, understanding the exact rate of lumen depreciation is critical for establishing maintenance schedules, calculating total cost of ownership, and ensuring compliance with occupational safety standards. The industry relies on standardized metrics, primarily L70 and L90, to define the point at which a luminaire’s output has degraded to 70% and 90% of its initial value, respectively. Projecting these figures accurately requires a deep understanding of the underlying thermal dynamics and the mathematical extrapolation methods defined by standard engineering practices.

The transition from traditional lighting to solid-state lighting introduced a paradigm shift in how lifespan is measured and predicted. Because an LED might continue emitting light for hundreds of thousands of hours, a new definition of “end of life” was required. The establishment of lumen maintenance standards by organizations such as the Illuminating Engineering Society (IES) provided the framework necessary for engineers to compare products objectively. By applying rigorous test data and projection models, designers can confidently specify luminaires that will satisfy strict photometric requirements years or decades after installation, even in the most thermally challenging environments.

Core Concept Definitions

The terminology surrounding LED lifetime and lumen depreciation is precise and governed by industry standards. The two fundamental metrics, L70 and L90, represent the elapsed operating time at which a light source maintains a specific percentage of its initial lumen output. L70 indicates the time required for the luminous flux to depreciate to 70% of its initial value, a threshold historically established because a 30% reduction in light output is generally the point at which the human eye can perceive a decrease in illumination, and falls below acceptable minimums for many tasks.

Conversely, L90 represents the point at which the luminaire delivers 90% of its initial output. This stricter metric is frequently utilized in applications where stringent light levels must be maintained, such as art galleries, inspection stations, and professional sports arenas. Achieving a long L90 lifespan requires superior thermal management and high-quality LED packaging, as the margin for degradation is significantly tighter than that of L70.

Underpinning these metrics are the primary testing and projection methodologies: ANSI/IES LM-80-20 and ANSI/IES TM-21-21. LM-80 is the approved method for measuring the lumen maintenance of LED packages, arrays, and modules over an extended period—typically a minimum of 6,000 hours, though 10,000 hours is preferred. However, because commercial luminaires are expected to last 50,000 to 100,000 hours or more, it is impractical to test them for their entire lifespan. TM-21 provides the mathematical framework for extrapolating the LM-80 empirical data to project long-term lumen maintenance, standardizing the prediction curve to ensure consistency across manufacturers.

The Mechanisms of Lumen Depreciation

To project L70 and L90 accurately, one must first understand the physical phenomena that cause an LED to lose output over time. Unlike an incandescent filament that simply burns out, an LED degrades due to complex interactions at the molecular and structural levels. These mechanisms are primarily driven by heat and drive current.

Thermal Degradation of the Junction

The junction temperature (Tj) is the most critical factor influencing the lifespan of an LED. The semiconductor die, typically made of Indium Gallium Nitride (InGaN) for blue pump LEDs, contains microscopic defects. Over time, high temperatures and continuous electrical current cause these defects to multiply and migrate. This non-radiative recombination means that electrical energy is increasingly converted into heat rather than light, accelerating the depreciation cycle in a self-compounding manner.

Phosphor Degradation

White LEDs are predominantly constructed using a blue LED die covered by a yellow phosphor coating (such as Yttrium Aluminum Garnet, or YAG). The phosphor absorbs a portion of the blue photons and down-converts them into longer wavelengths (yellow, red, and green) to produce broad-spectrum white light. High junction temperatures and intense photon flux cause thermal quenching and chemical degradation of the phosphor layer over time. This not only reduces the total lumen output but often causes a phenomenon known as color shift, where the light source drifts toward a cooler (bluer) color temperature as the phosphor becomes less effective at down-converting.

Encapsulant and Lens Yellowing

The optical materials used to encapsulate the LED die and form the primary lens—typically silicone or epoxy—are subject to degradation from both thermal stress and the high-energy blue photons they transmit. Epoxy resins, common in early generations of LEDs, are particularly susceptible to yellowing over time, which severely limits light transmission. Modern high-power LEDs utilize advanced silicones that offer superior stability, but even these materials will slowly darken or become cloudy when exposed to sustained high temperatures, contributing directly to the overall lumen depreciation of the system.

The Mathematics of TM-21 Projections

The ANSI/IES TM-21-21 standard utilizes an exponential decay model to extrapolate LM-80 test data. The foundation of this model is the understanding that lumen maintenance follows an exponential curve after an initial stabilization period.

The Exponential Decay Formula

The core equation used in TM-21 to predict the lumen maintenance (Φ) at a given time (t) is expressed as:

Φ(t) = B × e^(-αt)

Where:

Φ(t) is the projected lumen maintenance expressed as a decimal (e.g., 0.70 for L70).
B is the projected initial constant, derived from a curve fit of the LM-80 data.
α (alpha) is the decay rate constant, representing the rate at which the light output is degrading.
t is the operating time in hours.

To solve for the time (t) required to reach a specific lumen maintenance threshold (such as L70), the equation is algebraically rearranged:

L70 Time = ln(B / 0.70) / α

Similarly, for L90:

L90 Time = ln(B / 0.90) / α

Deriving the Constants from LM-80 Data

The TM-21 methodology requires a minimum of 6,000 hours of LM-80 test data, taken at multiple temperatures (typically 55°C, 85°C, and a third temperature chosen by the manufacturer). Measurements must be taken at least every 1,000 hours. The constants B and α are determined using a least-squares curve fit of the final 5,000 hours of data (e.g., from hour 1,000 to hour 6,000). The first 1,000 hours are omitted from the calculation to account for the initial “burn-in” period where lumen output might actually increase slightly or fluctuate before settling into a predictable decay curve.

If the test duration extends to 10,000 hours, the curve fit utilizes the final 5,000 hours (hours 5,000 to 10,000). Using the most recent 5,000 hours provides a more accurate representation of the long-term decay rate, minimizing the impact of early-life anomalies.

Interpolating for In-Situ Junction Temperatures

A fundamental challenge in projecting real-world lumen depreciation is that the junction temperature of an LED in a commercial luminaire (the In-Situ Temperature Measurement Point, or ISTMT) rarely matches the exact case temperatures used in the LM-80 tests (e.g., exactly 55°C or 85°C). Therefore, engineers must interpolate between the two closest LM-80 test temperatures to find the decay rate corresponding to the specific luminaire’s thermal performance.

The Arrhenius Equation Application

The interpolation method defined by TM-21 relies on the Arrhenius equation, which describes the temperature dependence of reaction rates. In this context, it models how the decay rate constant (α) changes with absolute temperature (Kelvin).

The activation energy (E_a) for the degradation process is first calculated using the decay rates (α1 and α2) from the two nearest LM-80 test temperatures (T1 and T2, in Kelvin):

E_a / k = [ln(α1) - ln(α2)] / [(1 / T2) - (1 / T1)]

Where:

k is the Boltzmann constant (8.6173 × 10^-5 eV/K).
T1 and T2 are the absolute temperatures of the LM-80 tests (K).

Once the activation energy factor (E_a / k) is determined, the specific decay rate (α_i) for the in-situ junction temperature (T_i) can be calculated:

α_i = α1 × exp [ (E_a / k) × ( (1 / T1) - (1 / T_i) ) ]

Finally, the pre-exponential factor (B_0) is calculated using the B constant from the lower temperature test (B1):

B_0 = B1 × exp(E_a / (k × T1))

The interpolated initial constant (B_i) is then:

B_i = B_0 × exp(-E_a / (k × T_i))

With the new interpolated constants (α_i and B_i) specific to the luminaire’s thermal profile, the standard exponential decay formula can be used to accurately calculate the projected L70 or L90 lifespan for that specific fixture.

Evaluating Drive Current Variables

While temperature is the dominant factor, the electrical drive current supplied to the LED array also significantly impacts lumen depreciation. LM-80 tests are conducted at specific drive currents. If a luminaire utilizes a drive current that falls between the tested values, engineers must account for this variable.

It is a common misconception that one can simply interpolate drive currents in the same manner as temperatures. However, TM-21 strictly prohibits interpolating between different drive currents. The standard dictates that projections must be based on the LM-80 data from a drive current that is equal to or greater than the drive current used in the actual luminaire. Utilizing data from a lower drive current will result in falsely optimistic projections, as higher current density exponentially increases both thermal load and non-radiative recombination rates within the semiconductor die.

When a manufacturer seeks to optimize efficiency by under-driving high-power LEDs (a practice known as “derating”), they may utilize LM-80 data from a higher current. This provides a conservative estimate of the lifespan, ensuring that the published L70 or L90 figures represent a worst-case scenario regarding electrical stress.

Advanced Considerations: L90 and High-Fidelity Environments

Calculating L90 projections requires meticulous attention to detail because the threshold for failure is remarkably slim. A 10% reduction in luminous flux can occur rapidly if the luminaire’s thermal management system is inadequate or if ambient temperatures spike unexpectedly.

Industrial Applications and Ambient Temperature Ratings

In heavy industrial settings, such as steel mills or foundry floors, ambient temperatures can regularly exceed 50°C. Standard commercial luminaires tested at a 25°C ambient will experience catastrophic early failure in these environments. When specifying fixtures for high-ambient applications, the engineer must request an ISTMT report conducted at the maximum expected ambient temperature. The resulting high junction temperature is then used in the Arrhenius interpolation. In such scenarios, even a premium luminaire might only achieve an L70 of 50,000 hours, whereas it might be rated for >100,000 hours in an office environment.

Color Maintenance and ANSI/IES TM-30

It is crucial to recognize that L70 and L90 metrics measure only the quantity of light (luminous flux), not the quality of light. As the phosphor layer degrades, the color temperature and color rendering capabilities of the luminaire can shift dramatically. In applications where L90 is specified (such as retail or automotive showrooms), maintaining precise color rendering is often just as critical as maintaining brightness.

Engineers must pair their lumen depreciation projections with an analysis of color shift, often utilizing data derived from LM-80 testing regarding chromaticity coordinates (Δu’v’). A luminaire might technically maintain 90% of its initial lumens for 50,000 hours, but if the color temperature has shifted by 500 Kelvin, the visual consistency of the space will be compromised, effectively rendering the luminaire failed from an aesthetic standpoint.

Reference Tables for Lumen Depreciation Projection

LM-80 Test Duration (Hours)	Maximum TM-21 Projection Limit	Sample L70 Extrapolation (Standard LED)	Sample L90 Extrapolation (Standard LED)
6,000	36,000 Hours (6x rule)	>36,000 Hours	22,000 Hours
8,000	48,000 Hours (6x rule)	>48,000 Hours	31,000 Hours
10,000	60,000 Hours (6x rule)	>60,000 Hours	42,000 Hours
15,000	90,000 Hours (6x rule)	85,000 Hours	55,000 Hours
20,000	120,000 Hours (6x rule)	115,000 Hours	68,000 Hours

Note: The actual projected hours within the 6x limit depend entirely on the decay rate constant derived from the specific thermal performance of the luminaire.

Real-World Application Examples

To illustrate the practical application of these metrics, consider the design of a manufacturing facility requiring an average maintained illuminance of 50 footcandles (fc). The facility operates 24 hours a day, 365 days a year (8,760 hours annually).

Scenario 1: Using L70 as the Baseline

The engineer selects a high-bay luminaire with an initial output of 24,000 lumens. The manufacturer provides a TM-21 report indicating an L70 projection of 80,000 hours at the facility’s ambient temperature of 35°C.

To maintain the 50 fc target at the end of the luminaire’s life, the initial design must produce significantly more light. The Lamp Lumen Depreciation (LLD) factor used in the lighting calculation software is set to 0.70. Consequently, the initial illuminance of the facility is designed to be approximately 71 fc (50 fc / 0.70).

At 8,760 hours per year, the 80,000-hour L70 lifespan equates to approximately 9.1 years of continuous operation. At the 9-year mark, the light levels will have depreciated to exactly 50 fc, necessitating a facility-wide retrofit to maintain safety standards.

Scenario 2: Utilizing L90 for Critical Inspection

A different zone within the same facility is used for critical surface inspection of machined parts. The requirement is a strict 100 fc, and large fluctuations in brightness cannot be tolerated. The engineer opts for a premium fixture and utilizes an L90 standard to minimize visual degradation.

The selected luminaire boasts an L90 projection of 60,000 hours at 35°C ambient. The LLD factor is set to 0.90. The initial design target is set to 111 fc (100 fc / 0.90), meaning the system only requires 11% more initial light compared to the maintained target, vastly reducing initial energy consumption compared to an L70 design.

The 60,000-hour L90 lifespan equates to approximately 6.8 years of continuous operation. While the maintenance cycle is shorter than the L70 scenario, the quality and consistency of the light during those 6.8 years are significantly higher, satisfying the critical inspection requirements.

Common Mistakes and Troubleshooting

When reviewing lumen depreciation data and specifying luminaires, several common pitfalls can lead to vastly inaccurate lifespan projections and premature system failure.

Accepting Extrapolations Beyond the 6x Limit

It is a surprisingly common marketing tactic for manufacturers to publish “calculated” L70 lifespans of 200,000 hours or more based on only 6,000 hours of LM-80 data. This violates the TM-21 methodology. The exponential decay model becomes increasingly unreliable over vast timeframes because it cannot account for sudden, catastrophic failures of secondary components (like driver failure or lens degradation) that occur late in the product’s life. Engineers must demand strict adherence to the 6x rule and reject non-compliant data sheets.

Ignoring Driver Lifespan

L70 and L90 metrics relate strictly to the LED array and its lumen output. They provide zero information regarding the lifespan of the LED driver. A luminaire might possess an L70 rating of 100,000 hours, but the electrolytic capacitors within its constant-current driver might fail catastrophically at 50,000 hours due to thermal stress. A comprehensive maintenance projection must include the Mean Time Between Failures (MTBF) of the driver, as replacing failed drivers often incurs significant labor costs that undermine the projected ROI of the lighting system.

Misinterpreting Ambient Temperature Ratings

Using data derived from a 25°C ambient test environment for a luminaire installed in a 45°C unconditioned warehouse will result in massive calculation errors. The junction temperature will be significantly higher than the interpolated model predicts, causing the actual decay rate (α) to steepen exponentially. Always demand ISTMT reports and TM-21 projections based on the highest conceivable ambient temperature the fixture will experience in the field.

Failing to Account for Dirt Depreciation

As noted previously, assuming that a high L90 rating guarantees long-term performance is a fallacy if the physical environment is ignored. In a woodworking shop, Luminaire Dirt Depreciation (LDD) might reduce light output by 30% in just two years, completely rendering the L90 lumen maintenance projection moot. Proper specification requires selecting fixtures with appropriately sealed IP ratings (e.g., IP66 or IP67) to mitigate dirt ingress and ensuring that cleaning schedules are integrated into the total Light Loss Factor calculation.

By rigorously applying the principles of LM-80 testing and TM-21 extrapolation, while simultaneously accounting for real-world environmental variables and secondary component lifespans, engineers can project LED lumen depreciation with a high degree of confidence. This precision ensures that commercial lighting systems deliver safe, code-compliant illumination from the day of installation through their final hour of operation.

Expanded Analysis of Thermal Management Strategies

Because junction temperature is the primary driver of lumen depreciation, the structural design of the luminaire is paramount. Heat must be efficiently extracted from the LED die and dissipated into the surrounding environment.

Thermal Interface Materials (TIM)

The efficiency of the thermal path begins immediately beneath the LED package. Thermal Interface Materials (TIMs), such as greases, pads, or phase-change materials, are utilized to eliminate microscopic air gaps between the LED circuit board and the primary heat sink. Because air is a poor conductor of heat, even a sub-millimeter gap can cause a localized temperature spike that severely impacts the L70 projection. High-performance luminaires rely on premium TIMs with high thermal conductivity (measured in W/m·K) that will not pump out or dry up over tens of thousands of operating hours.

Heat Sink Geometry and Convection

Once the heat reaches the aluminum housing, it must be dissipated into the ambient air. Passive convection is the most reliable method, as active cooling (fans) introduces mechanical failure points. The geometry of the heat sink—specifically the surface area and the orientation of the cooling fins—determines the efficiency of this transfer. Vertical fins allow for optimal convective airflow as hot air rises, creating a chimney effect that continually draws cooler air across the surface. Luminaires prone to dust accumulation must balance fin density with the need to shed debris; tightly packed fins might offer superior initial cooling but will quickly clog in an industrial setting, trapping heat and accelerating lumen depreciation.

The Impact of Enclosed Fixtures

Retrofitting LED lamps into existing enclosed fixtures poses a significant threat to lumen maintenance. An enclosed environment prevents convective airflow from rejecting heat into the wider room ambient. As the internal air temperature of the fixture rises, the effective ambient temperature for the LED lamp skyrockets, severely elevating the junction temperature. A lamp rated for an L70 of 50,000 hours in open air might depreciate to 70% output in less than 15,000 hours within an enclosed globe. Dedicated LED luminaires engineered as complete thermal systems offer drastically more reliable L70 and L90 projections than retrofit lamps.

Regulatory and Certification Implications

The accuracy of lumen depreciation projections extends beyond engineering best practices; it is deeply intertwined with energy codes and utility rebate programs.

DesignLights Consortium (DLC) Requirements

The DesignLights Consortium (DLC) establishes strict performance criteria for commercial LED luminaires. To qualify for utility rebates in many jurisdictions, a product must appear on the DLC Qualified Products List (QPL). The DLC mandates a minimum L70 lifespan, typically ≥50,000 hours, and requires manufacturers to submit full LM-80 reports and TM-21 calculators to substantiate these claims. Furthermore, the DLC evaluates L90 metrics for premium classifications, pushing the industry toward higher sustained performance and rigorous thermal management. Understanding these requirements is essential for engineers aiming to maximize return on investment for facility owners.

LEED and Sustainable Design

In sustainable design frameworks such as LEED (Leadership in Energy and Environmental Design), minimizing environmental impact involves extending the functional life of building systems. Specifying luminaires with exceptional L90 projections minimizes the frequency of replacements, reducing the raw materials consumed and the electronic waste generated over the building’s lifecycle. Precise TM-21 projections allow designers to document compliance with long-term performance criteria, demonstrating a commitment to enduring sustainability.

Conclusion and Future Trajectories

Lumen depreciation is a continuous, irreversible process, but it is entirely predictable when standard engineering methodologies are applied correctly. The interplay between thermal dynamics, drive current, and semiconductor physics necessitates precise testing through LM-80 and mathematical extrapolation via TM-21. While L70 remains the historical benchmark for general illumination, the increasing demand for high-fidelity, consistent lighting is driving the industry toward the stricter L90 standard.

As LED technology matures, innovations in semiconductor crystalline structures, more robust phosphor formulations, and advanced silicone encapsulants are pushing the boundaries of lumen maintenance. However, the fundamental laws of thermodynamics dictate that heat will always remain the enemy of solid-state efficiency. The responsibility rests with the lighting professional to critically evaluate manufacturer claims, demand comprehensive ISTMT testing, and calculate accurate Light Loss Factors that reflect the true operating environment. Only through this rigorous approach can lighting systems be designed to achieve their targeted illuminance levels reliably throughout their extended lifespans, ensuring both safety and performance for decades to come.