On Tue, 01 Nov 2016 16:10:25 GMT, Scott Lurndal wrote:

How does the cost of a Kwh contribute to the rated lifetime
of an LED-based illumination device?

Good point. I don't even remember how we got into the costs, other than I
made a joke that I'd love to live in Louisiana where the costs actually were
listed as around that amount.

I smell a red-herring.
Costs of electricity vary widely nationwide. Yes, California
has more expensive electricity (specifically to encourage conservation)
than other parts of the country, but that is not a factor in
rated lifetime.

Agreed.
We're talking actual lifetime of the LED "unit" (which includes whatever can
be replaced, which, for a household unit, is usually the driver and chips
and the housing, all in one package like my Costco setup).

Tj and cycles (on/off) would seem to be the two major controlling
factors, just as they are for incandescents.

Interestingly, the paper Jeff originally cited mentioned three main factors,
all of which exacerbated existing physical cracks between crystals:
1. Cycles
2. Heat
3. Vibration

To that, Jeff noted his bathroom fixture had a few more, mainly:
a. Humidity/moisture
b. Dust & orientation (aka heat retention or dissipation)
c. Voltage variations (e.g., we have *many* power outages per year here)

I'm only about 30 miles from Jeff (give or take) but we lose our power so
often that I don't know of many people who don't have a built-in generator
out here (plus we need the power to pump the water to fill our sinks).

So, overvoltage is key here. Very key. (I have holes blown in some of my
appliances, for example - even though I can't prove what caused it -
certainly I can see the burn marks and the high failure rate of fixed
appliances.)