We recently posted an investigative piece, looking at the provenance of a Philips IP65 batten that scared the contractor so much that he had it removed from the project. In that piece, I flippantly referred to the stated 50,000 life of the batten, forgetting that Richard Hayes, blessed contributor to these pages, has a few things to say about LED life and Maintenance Factor. So I asked Richard what the fuss is all about.
He told us . . .
LED salesmen are always confused when they tell me their LEDs last 50,000 hours and I reply SO WHAT ! End users who only see that headline figure might not be aware that a 50,000 hour LED system might well have to be replaced after only 7,000 hours. Manufacturers are misleading specifiers and end users by making these simplistic statements.
Unlike the previous generations of (filament) light sources, that definitively failed at end of life, LEDs are a bit like a fart at a dinner party; they eventually fade away to nothing, leaving nothing but a bad memory behind them. Trawl the internet and you will generally find statements that LEDs last in excess of 50,000 hours. The industry helpfully (help as in passing a drowning man an anvil) provides a measure of life from LM 80 testing and TM21 data extrapolation.
Some typical statements :-
- L70 = 100,000 hours,
- L90 = 60,000 hours
- 50,000 hours life time (L70/B50)
But do these statements mean that the products last for 100,000 or 60,000 or 50,000 hours?
The answer is Yes . . . and No – because it all depends on how we define Life. L70/50k is taken to mean that the LEDs will be emitting 70% of their initial flux at 50,000 hours burning. From there it gets a bit more complicated but, for the sake of clarity, let’s try to keep it simple.
If we specify an LED product in any commercial project then, as responsible lighting designers, we must calculate the illuminance as a maintained value at the end of the maintenance cycle of the installation.
If I take the standard method of calculating a maintenance factor for an installation and I include a light source depreciation factor (LLMF) of 0.7 then combine this with a Room Surface Degradation Factor (RSMF) and a Luminaire degradation factor (LMF) I will end up at 50,000 hours with a Maintenance factor of between 0.5 and 0.6.
Hands up all those that believe this to be viable situation? By 50,000 hours, the installation will be delivering only 50% – 60% of its original output; either that or else the design requires a starting illuminance of double the actually required level, because of the maintained value at the end of the maintenance cycle of the installation.
To get a Maintenance Factor that is economic (and worthwhile) I will have to reduce the burning time of the LEDs to a point where the LLMF factor is 0.9 or greater. If I had the actual degradation graph for the LED I could look for the burning hours where the Lumen depreciation meets my target and I would make this time the maintenance interval of the installation. In most cases, these graphs and data are not readily available.
The new draft of EN 12464 Pt1 Light and lighting – Lighting of work places – Part 1: Indoor work places refers us to a new document on calculating Maintenance Factor – PD ISO/CIE TS 22012:2019 Light and lighting – Maintenance factor determination – Way of working. The document contains a lot of new information, but of particular interest are Tables B.1, 2 and 3 which take the TM21 standard calculation and apply it at 5000 hour intervals to the common LM80 life declarations we see from manufacturers
Note CIE 97 terminology calls the factors LLMF, RSMF and LMF as I’ve used above, but the new document changes the nomenclature to fLF , though it is careful to point out that it means the same as LLMF
For Example (N.B. all values taken from PD ISO/CIE TS 22012:2019 Tables B1, B2 and B3)
- L70 =50,000 hours reaches 0.9 Lumen Depreciation factor at 15000 hours
- L70=50,000 hours reaches 0.95 Lumen Depreciation factor at 6700 hours (by interpolation from Table B1)
- L90=50,000 hours reaches 0.95 degradation factor at 25,000 hours
- L90=70,000 reaches 0.95 degradation factor at 35,000 hours
So we can interpolate that L90=60,000 hours would have a usable LLMF of 0.95 at 30,000 hours
- L70 = 100,000 reaches 0.95 degradation factor at 15,000 hours
- L70 = 100,000 reaches 0.90 degradation factor at 30,000 hours
If I define life as offering an economic period over which I can practically use the LEDs then life could be as low as 7,000 hours, but might stretch up to 30,000 hours. Just a thought: didn’t the last generation of HPS lamps hit 30,000 hours life ? T5 fluorescent lamps hit 25,000 hour life ? And LEDs mean progress ?
As the LEDs will not fail i.e. they will still glow and consume power, there will be no indication to the end user that at somewhere between 7,000 and 30,000 hours, in order to maintain the original illuminance specification, they will have to either relamp – by changing the LED boards or replace, by fitting new luminaires (in the case of “sealed for life” fittings, for example.)
Unless manufacturers make it clear that 50,000 hour life could mean replace at 7,000 hours and that 100,000 hour life means replace at 30,000 hours end users will be left with poorly lit installations that fall below recognised and established norms.
I have only been able to find one manufacturer who makes this apparent on their website with a detailed pdf explaining LED life, Kudos to Whitecroft, I wish everyone was as honest.
(LED & Maintenance Factor – Whitecroft Lighting refers)