Introduction

The US Energy Star program has developed the TM-21 and TM-28 calculators, based on the TM-21-2011 and TM-28-2014 standards published by the IES, to support the LED industry in predicting the service life of LED modules and luminaires using LM-80 and LM-84 test data. As both standards were developed by the IES, LED product manufacturers need only use the calculation tools provided by the standards body to avoid misunderstandings regarding calculation methods within the industry, whilst ensuring that LED lifetime predictions are more accurate and comparable across the board. TM-21 and LM-80 are well-established methods for evaluating LED lumen depreciation within the industry; this calculation method primarily relies on LM-80 test reports issued by LEDs or LED module manufacturers; By entering the relevant test data from LM-80 and ISTMT reports into the TM-21 calculation spreadsheet, one can derive L70 or other Lxx LED lifetime parameters. TM-28, on the other hand, uses the light decay test report (What’s light decay?) for the complete luminaire (the LM-84 report) as the basis for calculation; this is the focus of this article, and we shall now proceed to a detailed explanation in the main text.

Theoretical foundation of two methods to calculate the LED lifetime

TM-21 is based on the Arrhenius thermal accelerated aging model, whose core assumption is that the lumen depreciation rate of an LED chip is determined by its junction temperature and is relatively independent of the luminaire’s other components. It involves measuring the lumen maintenance of the LED chip for at least 6,000 hours at specific temperatures (typically 55°C, 85°C, 105°C, etc.) in accordance with the LM-80 standard. Combined with the LED operating temperatures (Ts and Tj) and operating current from ISTMT or LM-98-24 test reports, the L70 LED lifetime is calculated using TM-21’s exponential decay extrapolation formula.

TM-28, on the other hand, adopts a system-level testing approach, recognizing that the light output decay of a complete luminaire is not solely attributable to the LED chips but is also influenced by various factors such as the LED driver (More about LED driver), optical lenses, and thermal management structure. The core standard is LM-84, which directly measures the light output decay curve of the complete luminaire under real operating conditions, and then calculates the LED lifetime using the TM-28 calculator. Additionally, TM-28 allows for a “hybrid extrapolation” approach: if the actual test duration of the complete luminaire is less than 6,000 hours, LM-80 and TM-21 data from the LEDs can be used to assist with the extrapolation. This method more closely reflects real-world usage scenarios, but the testing cycle and costs are significantly higher.

Test subject and data source of two methods

The TM21 test applies to LED chips, packaged modules, or arrays (at the component level). Data sources consist of three parts: first, LM-80 test reports provided by LED chip manufacturers, which typically include decay curves spanning thousands of hours at at least one temperature point (ideally three); second, ISTMT (In-Situ Temperature Measurement) reports completed by the applicant, documenting the actual operating temperature of the LED chips (Working temperature of LEDs and LED driver) within the product; and third, the ISTMT report must include a schematic diagram or photograph showing the location of the temperature measurement point (TMP).

The TM-28 test applies to complete luminaires (finished products). Its core data is derived from the LM-84 test report, which must document the changes in luminous flux of the entire luminaire after at least 6,000 hours of continuous operation in a laboratory environment. If the hybrid extrapolation path is selected, the LM-80 report for the LED chips and the ENERGY STAR TM-21 worksheet must also be submitted. Unlike the TM-21 method, TM-28 does not require the separate submission of ISTMT or TMP photographs, as the actual measurements of the complete luminaire already account for all thermal effects.

Advantages/disadvantages and applications of two methods

The core advantages of TM-21 lie in its low cost and short turnaround time. ISTMT testing typically takes only a few days, and LM-80 data is provided free of charge by LED chip manufacturers, so it takes only about one week to estimate the LED lifetime. The drawback of this method is its reliance on external data, and there are theoretical errors in temperature extrapolation. Since this method tests the temperature at the Ts point, variations in the location of that point can lead to errors in the temperature test results. Additionally, because different LED chips (More about LED chips) are positioned differently within the luminaire structure, heat dissipation may vary slightly. All of these factors cause the TM-21 extrapolated results to deviate from the actual lifespan of the entire luminaire. TM-21 is best suited for LED chips from major manufacturers such as Lumileds, Cree, Nichia, and Osram, as they offer products with mature thermal designs and can provide certified LM-80 test reports.

The greatest advantage of TM-28 is its authenticity and independence. Actual testing of the entire luminaire is not constrained by LED manufacturer data and can reflect the comprehensive aging behavior of all components, including the driver, optics, and heat dissipation (Understanding factors which refer to luminous efficacy). It is particularly suitable for non-standard heat dissipation designs or white-label LED solutions. The disadvantages are equally evident: an LM-84 test for 6,000 hours takes approximately 250 days, and the costs of laboratory rental and testing are far higher than those of ISTMT testing. Although a hybrid extrapolation approach can shorten some of the actual testing time, cross-validation of the data is complex. TM-28 is more suitable for high-end products, custom luminaires, or scenarios where LED chips lack reliable LM-80 data. Of course, for some stringent municipal projects, this method yields relatively more reliable LED lifetime results because it takes a wider range of factors into account.

How to use TM-21 or TM-28 calculator to predict LED lifetime?

In the lighting industry, both the TM-21 and TM-28 calculators are professional tools used to estimate the lifespan of LED luminaires. Simply put, their core function is to estimate the lumen depreciation of LED components or luminaires (ZGSM LED lights) over the long term, based on LM-80 test data or LM-84 test reports. Taking TM-21 as an example, the LED lifetime estimation process is primarily divided into three steps: First, obtain the LM-80 test data; second, determine the luminaire’s operating temperature (Ts or TMPLED) and drive current; third, enter the LM-80 test data along with the luminaire’s operating temperature and current into the TM-21 spreadsheet. Finally, we can derive the luminaire’s lumen depreciation curve and thereby get its projected or reported LED lifetime (i.e., the percentage of light output at a specific time). For TM-28, the estimation method is similar to that of TM-21, but it directly uses the complete LM-84 test report for the luminaire. This report already includes the 6,000-hour light decay curve measured for the luminaire at specific temperatures and currents; in this case, TM-28 can directly extrapolate the LM-84 data to determine the LED luminaire’s lifetime. Below is a list of the required documents for both methods.

Lifetime	TM-21 method	TM-28 method
	LM-80	LM-84
	ISTMT TMPLED (View of In-Situ Point- Ts)	TM-28 calculator
	TM-21 calculator	TM-21 methods (ISTMT and LM80)

As shown in the figure below, the image on the left depicts the TM-21 calculator, and the image on the right depicts the TM-28 calculator. For TM-21, we can enter the lumen maintenance percentage of the LED chips under different Ts conditions (typically three temperatures) and the Ts temperatures in ISTMT report into the calculator; for TM-28, we need to enter the test results of the luminaire at different ambient temperatures—including the Ts temperature (here it’s termed as the tested ambient temperature) and the lumen maintenance percentage—into the calculator. When we enter the luminaire’s actual operating temperature (as found in the ISTMT report) into the calculator, it can use the TM-21 or TM-28-fitted lumen depreciation curve to estimate the LED lifetime.

Which one is mostly used to calculate LED lifetime? TM-21 or TM-28?

TM-21 is the most widely used standard in the industry for estimating LED lifetime; TM-28 is selected only for applications with stringent product performance requirements. Although TM-28 provides a more accurate reflection of the overall degradation of the entire luminaire, it also has significant limitations. TM-21 relies on standard LM-80 module lumen depreciation reports, combined with ISTMT temperature measurement data, to complete lifetime projections. It does not require prolonged aging tests on finished luminaires, resulting in low testing costs and short cycles. The vast majority of new general lighting product development projects and standard North American energy efficiency certifications adopt this approach. In contrast, TM-28—introduced in 2016—requires LM-84 long-term aging tests on the entire luminaire. Due to the high testing costs and lengthy duration, its use is mandated only in specific scenarios such as the DLC’s high-end commercial lighting projects, public lighting projects (ZGSM public lighting solutions) and Amazon U.S. luminaire reviews; over the past decade, its overall adoption rate has remained far lower than that of TM-21. Even if LM-84 and TM-28 were to fully replace LM-80 and TM-21, the inherent flaws could not be avoided; specifically, the mismatch between test conditions and actual operating environments is a problem common to both sets of calculation methods. In standard testing, LEDs operate continuously for 24 hours under constant temperature, humidity, and drive current in a controlled environment; however, luminaires do not operate continuously for 24 hours in real-world use, and temperature, humidity, and current are not static. The key issue is determining the extent to which these operational differences affect LED lifetime estimates, yet relevant research on this topic remains limited.

ZGSM LED lighting solutions

Summary

This paper systematically compares the two major standards in the field of LED lighting lifespan assessment—TM-21 and TM-28—analyzing them from five dimensions: theoretical foundations, test subjects and data sources, advantages, disadvantages, and applicable scenarios, methods for using calculation tools, and the current state of industry applications. TM-21 is based on the Arrhenius thermal acceleration model. It uses LED chips or modules as test subjects and relies on LM-80 reports and ISTMT in-situ temperature measurement data to estimate lifespan through exponential extrapolation. It offers significant advantages, including low cost and a short testing cycle (approximately one week), making it the preferred solution for the R&D of most new general lighting products and for North American energy efficiency certification. However, its drawbacks are equally evident: it relies on data from external LED chip manufacturers, and variations in the location of temperature measurement points (Ts points), as well as differences in heat dissipation depending on the chip’s position within the luminaire, may cause the estimated results to deviate from the actual lifespan of the complete luminaire. TM-28, on the other hand, tests the complete luminaire and estimates LED lifetime based on the LM-84 long-term aging report for the entire luminaire. It comprehensively reflects the degradation behavior of all components—including the driver, optical lens (Lens and light distribution), and heat dissipation structure—and yields results that more closely mirror real-world usage scenarios. It is particularly suitable for non-standard heat dissipation designs or white-label LED chip solutions. However, LM-84 testing takes approximately 250 days and is significantly more expensive than TM-21 testing; therefore, it is mandated only in scenarios such as R&D for high-performance products, high-end commercial lighting under the DLC program, and Amazon U.S. luminaire reviews. In practical applications, TM-21 holds an absolute dominant position, while TM-28 is used far less frequently. It is worth noting that both standards share a common limitation: their test conditions (24-hour continuous operation with constant temperature, humidity, and current) do not align with actual usage environments. Furthermore, there is currently a lack of sufficient research on the specific impact of these differences in test conditions on LED lifetime estimation results.

Related Products

Related Blogs

Related Cases

People also ask

Author introduction