Avoiding Overrated LED Lifetime Claims Begins with You: How to Add Real Value to Your Client’s New Lighting System

Lighting used to be simple. Install the light fixture, screw in the light bulb, flick on the breaker and the switch and there’s light. Enabled by North America’s relatively cheap electricity, this simplicity includes a massive amount of wasted electricity and money. Standard incandescent bulbs emit less than 10% of input power as light, the other 90% as heat creating more air circulation and cooling needs, and electrical load. Halogen incandescent improves the efficiency a bit. Discharge technologies (e.g. fluorescent, HID, high pressure sodium) are much more efficient and generate less heat, with the downside of both bulbs and ballasts needing periodic replacement, creating an e-waste stream.

Solid-state lighting has ushered in a new era of energy efficiency that Thomas Edison himself surely would have embraced. Well-designed LED bulbs and fixtures (aka LED lamps and luminaires) use less electricity than fluorescent, with much longer life. Whereas incandescent by definition provides light as a byproduct of heat, LEDs generate much less heat as a byproduct of producing light. It is this advantage in thermal management that allows LEDs to last so much longer.

True LED lamp and luminaire life however depends on a number of important factors, thermal performance being one of them, and exaggerated lifetime claims - the kind we’re seeing a lot of often result from misunderstandings about the industry standards underpinning such claims. Unfortunately, many others are the result of willful efforts to misuse the standards in a meaningless game of specsmanship usually taking the form of LED product lifetimes defined in hundreds of thousands of hours. For example: 300,000 hours for a streetlight would be north of 34 years of continuous operation. Are we to believe this fixture will last 68 years at 12 hours on / 12 off? The LED lamp and luminaire market has run amok with overrated lifetime claims. Armed with a simple calculation and a willingness to stubbornly defend common sense, you the informed electrical contractor can help bring these games to an end.

The Illuminating Engineering Society IES) publishes American National Standards (i.e. ANSI) for LED measurements. Established in 1906, the IES is the recognized technical and educational authority on illumination. For over 100 years our objective has been to communicate information on all aspects of good lighting practice to our members, to the lighting community, and to consumers through a variety of programs, publications, and services. IES members include engineers, architects, contractors, designers, educators, students, distributors, utility personnel, manufacturers, and scientists, all contributing to the mission of the Society: to improve the lighted environment by bringing together those with lighting knowledge and by translating that knowledge into actions that benefit the public.

LED bulb or light fixture life is a function of its weakest link. Either the source and/or its ballast or driver will fail first, or second as a result of housing failures. But unlike preceding light sources the LEDs within an LED lamps or luminaire are typically not serviceable. Just as a transistor radio contains transistors, an LED bulb or fixture contains irreplaceable light emitting diodes soldered to a board; often the LED driver powering the LEDs is also not replaceable. Many manufacturers assume that property owners will be inclined to renovate a space around the time its light fixtures reach end of life; that remodel will replace the fixtures anyway. For that reason, they include no provisions for upkeep of the fixture as it is envisioned to become trash after, say, 50,000 hours of operation. Especially for hard-to-reach fixtures with integral LEDs, this lack of maintainability makes a rigorous fixture specification process all the more important. Replacing a transistor radio is easy. Hiring multiple trades to replace a single trimless recessed can that fell far short of its lifetime rating is an expensive nuisance.

Though LED driver failures, broken fasteners and weathered optics are all reasons for replacing a fixture, it’s the LEDs that are most scrutinized, and for good reason. Just like discharge sources, LED output depreciates over time. The metric most commonly referenced when setting LED product life claims is luminous flux maintenance (aka lumen maintenance). And just as 70% of initial flux is a typical threshold for fluorescent group relamping, 70% is an accepted threshold for replacing LED products. ANSI/IES LM-80-15, the IES Approved Method for Measuring Luminous Flux and Color Maintenance of LED Packages, Arrays and Modules is the standard for measuring luminous flux maintenance of LEDs. “LM-80” measures luminous flux of LEDs over time, as a function of temperature. Importantly, what is tested is the LED packages, arrays or modules, the discrete subcomponents within bulbs and fixtures. The resulting LM-80 test reports include luminous flux and color values over time for the LED devices operating at 55°C, 85°C, and often higher temperatures.

Here’s where the problem begins, and where you can add real value for your clients. Before LEDs, lamp lifetime was established by burning a group of bulbs until 50% of them had failed, and that duration informed the product’s lifetime rating. If 100 bulbs were placed on test and the 50th failure occurred at 1,073 hours, the manufacturer rated the bulb for 1,000 hours, as captured in IES standards. With LEDs, testing to 50,000 hours (a common life rating) would take nearly six years, and 100,000 hours, more than a decade. This is obviously impractical.
The solution from the IES Testing Procedures Committee is to measure LED luminous flux at regular intervals over 6,000+ hours per LM-80, and to use a standardized calculation to project to a point in the future when flux levels should reach 70% of initial light output. The calculated time (in hours) until an LED reaches this threshold is called L₇₀, as defined in IES TM-21-11, Projecting Long Term Lumen Maintenance of LED Light Sources. With this standard, a 6,000-hour LM-80 report, and the temperature of the LEDs measured inside a product, a manufacturer can calculate the projected luminous flux maintenance for those LEDs in that application, e.g. 36,000 hours to L₇₀; this is the projected luminous flux maintenance life.

TM-21’s intermediate steps can sometimes produce huge numbers (e.g. 315,518 hours) that aren’t valid. Let’s set aside the LEDs and pick up a ball. Just as you can be confident that a ball dropped from your hand will land in a small circle you’re standing in, we can be confident in the measured LM-80 data. Move that circle 10 feet back your confidence projecting the ball into the circle goes down a bit. Move that circle another 20 feet and you might think you’ll land it (but I don’t). One hundred feet away, no matter how good a game you talk, hitting that circle is a bad bet. Setting aside the lousy comparison of human and semiconductor performance, it’s this same concept of confidence that sets up the one rule in TM-21 that is sacrosanct, called the “6X rule”. Unobserved, the entire TM-21 standard is rendered meaningless.

The 6X rule states “luminous flux values must not be projected beyond 6 times the total test duration (in hours) of measured data”. If an LED manufacturer has reported 6,000 hours of LM-80 data, the longest legitimate projection at the bulb or fixture level is 36,000 hours, even if an intermediate calculation shows 339,304 hours. It doesn’t matter if one is projecting to L₇₀, L₈₀, or L₉₀; regardless, statistical confidence in the projection falls off dramatically beyond 6X. The most relevant output from the TM-21 calculation is called the Reported L70 value, because it is capped at 6X. Calculated L₇₀ and Projected L₇₀ values, if reported, simply aren’t intended for designing a lamp or luminaire, and TM-21 says so. 

I worked for an LED manufacturer for years, involved in LM-80 testing planning and review. The longest credible and reported LM-80 test I’ve heard of is 18,000 hours. For that reason, any projection longer than 110,000 hours doesn’t pass the smell test with me. I round it off to 100,000: anything north of here is suspicious and worthy of inquiry. Those reporting 150,000+ hours to L₇₀ while referencing TM-21 are conveniently ignoring clause 5.2.5, the 6X cap. They are attempting to make money from the uninformed. Unfortunately, many other reputable manufacturers, not content to see competition eat their lunch, feel compelled to play the same statistical games.

Don’t be fooled. Defend your client and her project with a quick calculation. Especially for critical types on the fixture chedule, ask your reps for the IES LM-80 reports for the LED packages, arrays or modules integrated into this fixture model. Your wall pack vendor claims 100,000 hours to L₇₀ but presents you 
a 10,000-hour LM-80 report? Call them on it. Tell them you need the 17,000-hour report and their TM-21 projection. A streetlight manufacturer is hoping you’ll spec their cobrahead rated 250,000 hours to L₉₀? Let them know you’ll review the 42,000-hour LM-80 report. They don’t have it, and they’ll be waiting to see if you cave. Don’t. Even if they had such a report, what relevance would it be if the tested LED parts are five years old and no longer sold?

LED lifetime is unfortunately complex and subject to gaming, but with some basic principles and a back of the envelope calculation you can defend against overrated products. This year IES will publish ANSI/IES TM-21-19, the first revision of this standard, and the accompanying ANSI-approved IES TM-21 Calculator. For more information please visit www.ies.org or email abaker@ies.org.