Relationship between Lumen Degradation and Ambient Temperature of LED Luminaires.

It is likely to be based on standard statistical calculations for MTTF, modified by an acceleration factor A_f using the Arhenius Equation:




I've not plugged numbers in to check, mind ... you'd need to know, or search up, the activation energy of the type of LED junction you were considering to do this. Basically, this is the forward voltage drop of an individual LED at the point the LED turns on.

Slight divergence, I recall LEDs being advertised with 50,000 hrs and have recently noticed that some are now 30,000 or even 20,000 hrs...............


I put up 11 x 1800 mm "quality" high output LED three years ago in an industrial unit using them 60 hours or so a week, and 5 of them have part failure where a section anything from 100 to 400 mm is now black.


So glad my policy of "customer expectation" meant it is the client's job to whine to everyone else bar me!


Regards


BOD

Realise that a large no. of so called LED failures are actually a failure of the cheaply made driver electronics, though in the case of a stripe of LEDs going, I suspect one open circuit in  series parallel network, so one series stripe has gone dark. It may now be that those remaining are being a bit over-driven.


However, all semiconductors, LEDS or cheap power supplies, benefit greatly from cool running.

When over-voltages have been correctly suppressed (and that is another can of worms) then the life is limited by thermal factors, that for silicon can halve the lifespan for an 8 degree increase at the top end.

Electronics designed for places like down oil wells may run well above soldering temperatures (being brazed) and can have very short lifespan indeed (less than thousand hours )
app note about it.