News

Worldwide comparison of 110 LED testing laboratories: Results Presented

The International Energy Agency’s (IEA) Energy Efficient End-Use Equipment (4E) Annex on SSL (SSL Annex) works to assist governments of member countries in promoting SSL as an effective means to reduce energy consumption worldwide. Starting in 2011, the SSL Annex launched an initiative that sought to address the lack of a global laboratory performance assessment scheme. This initiative was designed to help support harmonisation of SSL testing around the world. It also represents an attempt to establish a common proficiency test for accreditation programmes aimed at different regional test methods. (Proficiency testing is used to establish whether a lab can be accredited for testing against a specific standard.)

“The output of the SSL Annex and the interlaboratory comparison is very useful for governments and market surveillance authorities around the world” said Dr. Peter Bennich, chairman of the SSL Annex’s Management Committee and representative of the Swedish Energy Agency, one of the Annex’s member countries. “In Europe, for instance, the results are particularly relevant for stage 6 of the ecodesign requirements for lighting. After 2016, LED lighting will be even more prominent.” Dr Bennich added: “When compact fluorescents were introduced, we saw a lot of dissatisfied consumers due to the poor performance of some products. Knowing that we can trust the laboratory tests is at the core of the regulatory and market surveillance work we as a governmental Agency are undertaking. SSL products are part of a global supply chain. And this interlaboratory comparison helps to ensure that we can trust the test results from these 110 laboratories around the world.”

The SSL Annex’s Interlaboratory Comparison 2013 (IC 2013) was conducted between October 2012 and August 2013, and involved 110 laboratories comparing measurements of photometric, colorimetric, and electrical quantities of several different types of SSL products. More than 50 laboratories participated in the study directly while others were ‘linked’ to IC 2013 through existing proficiency test schemes in North America and the Asia Pacific region. The sample of products tested by the laboratories varied slightly, taking into account the most important products in the regional markets – but all test samples included both LED retrofit lamps and luminaires.

The comparison looked at the measured values of luminous flux, luminous efficacy, active power, RMS current, power factor, chromaticity x and y, correlated colour temperature, and colour rendering index. “While most of the laboratories were within the expected levels of agreement, a few extreme outliers were observed which must be caused by problems at the participant laboratories in meeting the requirements in the test method” said Dr Yoshi Ohno at the US National Institute of Standards and Technology (NIST), lead author of the IC2013 report. “Identifying these large deviations by a small group of laboratories demonstrates the importance of proficiency testing, as these laboratories would not have been aware of their problems without participating in IC 2013.”

The results of the electrical measurements identified some issues relating to test equipment and the test method used. This is one of the important issues for LED lighting test methods in use today, and future improvements are expected. The uncertainties reported by the participants were found to have a very large range (often more than two orders of magnitude), and some were significantly underestimated. Still other laboratories did not report uncertainties at all or for any colour quantities (i.e., chromaticity x, y, CCT, CRI).

“The 2013 IC helped us to understand that uncertainty evaluation, especially for colour quantities, is still very difficult for the SSL industry, and reported uncertainties are often not reliable” said Dr Ohno. “Practical methods and tools for uncertainty evaluation of measurements, as well as educational documents and training for the SSL industry on practical uncertainty evaluation are urgently needed.“