Lumen-maintenance testing for LED lamps, light engines and luminaires (MAGAZINE)

Now that procedures are in place to make long-term lumen-maintenance projections for LED components, efforts are being made to develop methods for testing the lighting system as a whole, as JIANZHONG JIAO explains.


This article was published in the February 2012 issue of LEDs Magazine.

View the Table of Contents and download the PDF file of the complete February 2012 issue.


It is widely understood that measuring lumen maintenance is critically important for determining the life of LED lighting products. In the past few years, the Illuminating Engineering Society of North America (IESNA) has developed two test methods that address the lumen maintenance of LED light sources used in such products. LM-80 is an approved method for measuring lumen depreciation of LED light sources, and TM-21 is a technical memorandum which specifies how to extrapolate the LM-80 data in order to make long-term lumen-maintenance projections.

Beyond the LED, other components at the LED lighting-system level also can impact the long-term lumen maintenance. These components include, but are not limited to, LED lamps, engines, luminaires, drivers, thermal-management devices and optical components. Over time, these components may experience some change or degradation. In particular, the plastic elements used in the optics may change in several ways, including light transmittance, haze and undesired color change. In turn, the overall light output of the LED lighting products will be further reduced beyond the normal lumen degradation of the LED source. Therefore, it may be necessary to conduct lumen-maintenance tests at the LED lighting-system level.

System-level testing

In 2010, the IESNA Testing Procedures Committee (TPC) formed a working group to draft a document to recommend the best methods for testing lumen maintenance of LED lamps, engines and luminaires. Because of the various ways of integrating LEDs into lighting systems, the TPC experts concluded that LED lumen-maintenance information obtained from LM-80 and TM-21 may not be sufficient to reflect the changes in a system’s light-output level over time. This conclusion led to the direction of testing the lighting system as a whole.

Further, the performance of LED lighting systems is typically affected by variables such as operating cycle, ambient temperature, airflow, and orientation, as well as conditions imposed by auxiliary equipment and fixtures. These conditions are not accounted for in the LM-80 testing of LED sources. As a result, it may become necessary for the IES to provide recommendations on these test conditions, and test methods would need to be designed to give comparable results when adopted by various testing laboratories.

A lumen-maintenance test standard for LED lamps, engines and luminaires is still being developed, with multiple drafts created, and some preliminary balloting already completed. Different from the LM-80 test standard, this new document will address the tests that are more uniquely applicable to LED lighting systems, including temperature conditions, operational duty cycle and test duration.

For LED sources, LM-80 recommends three case temperatures at which testing is performed. Often, the LEDs being tested are placed in temperature-controlled thermal chambers with sophisticated active-cooling systems. These precisely set the LED case temperatures to meet test requirements, with each test running for a minimum of 6000 hours at each case temperature.

It can be cost-prohibitive and impractical to test entire LED light systems – such as large luminaires for outdoor roadway lighting fixtures or chandeliers for indoor lighting – in the temperature-controlled chamber for long periods of time. Instead, the TPC recommends LED lighting systems be tested in a condition as close as possible to the installation orientation and mounting method, which in many cases is room temperature (25°C) with tolerances.

The temperature-control points for the room environment where LED lighting systems are placed in the burning cycle, or the period of time when LEDs are turned on, may need to be monitored. This provides a more practical and flexible situation for test labs or manufacturers to conduct tests, either in designated test rooms or within enclosures where the LED lamps, engines and luminaires are mounted. Understanding that ambient temperatures may affect long-term lumen output for the LED lighting systems, the TPC recommends that the tested samples be mounted in accordance with manufacturer recommendations, taking unintended thermal dissipation into consideration.


Also of concern is the operational duty cycle. When testing LED sources per LM-80, the samples are consistently powered on throughout the test. In practice, however, LED lighting systems are cycled on and off in almost all applications for both indoor and outdoor lighting. As has been demonstrated in the past, cycling lamps on and off has an impact on lamp life for other lighting systems such as high-intensity-discharge (HID) and fluorescent lamps.

Although cycling may not have as dramatic an impact on LEDs, the driver electronics can be affected. Therefore, the TPC recommends that the burning cycle during lumen-maintenance testing should not be conducted in a constant on condition. Instead, the TPC specifies an operational cycle with certain periods of on time and off time. This cycling condition is intended to have a closer relationship to real-life LED system applications.

Test duration

The duration of testing is another variable to consider, and the area where experts have struggled the most to establish a standard. It is clearly understood that lumen maintenance over time usually cannot be detected over a short testing period. LM-80 specifies a minimum of 6000 hours of testing.

However, for LED systems, including lamps, engines and luminaires, prolonged testing creates a large burden for manufacturers and the lighting industry. Any recommended testing procedure from the IES must present practical values for the duration of testing. A testing period that is too short may not lead to conclusive characteristics being identified. On the contrary, as stated above, a testing period that is too long may create an unsustainable burden.

Without standardization of the test duration and with incomplete testing methods, the industry could suffer from inconclusive test results. The discussions on testing duration are still taking place in the TPC working group, and there are several specific suggestions currently under consideration.

One suggestion is to specify certain testing durations to the users of the document depending on what information users are looking to determine. In other words, if the intent of the test is to identify LED lumen-maintenance degradation in the early stage of usage, the samples may be tested for 2000 hours. If the intent of the test is to use the collected photometric data to make long-term lumen-maintenance projections, the samples may be tested a minimum of 20% of the specified lumen-maintenance life. In doing so, the test data may apply as with the TM-21 memorandum. If the test is meant to obtain the actual lumen-maintenance life for the LED products, the samples should be tested for the entire claimed lumen-maintenance life.

It is expected that a new testing standard will provide significant value to the SSL industry. In a broader sense, consistent, objective and practical testing methods recommended by the IESNA are necessary for the implementation and adoption of SSL technologies, both for government or consortium specifications, as well as for industry product-performance standards.

More in Indoor Lighting