Early issues concerning light quality have largely been solved in all but the cheapest LED products and the (relatively high) initial purchase cost of LED bulbs has been literally decimated in the last six years. As a result the LED lighting market, which was worth $4.8bn in 2012 is predicted to grow by 45% per year to $42bn by 2019.
Regardless, all technologies will have their detractors. As a recent example, Scott Elder of Linear Technology stated that, although LEDs claim a lifetime of tens of thousands of hours, most will actually fail well before this. This, he claims, is the ‘big lie’ of LED lighting. It’s an alluring argument. However, in my estimation, it just doesn’t add up.
If you look at his working-out Scott assumes that the ‘circuit elements’ in traditional bulbs and LEDs fail at an equivalent rate. More ‘circuit elements’ (such as the separate components of a driver circuit) lead to a higher failure rate, he claims. However, it’s hardly sensible to compare the ‘circuit element’ of a piece of tungsten (heated to the point that the wire evaporates) with relatively cool-running silicon connected by copper circuits. According to this same working out I’m pretty certain that a standard PC would fail almost as soon as you got it home!
The rest of Scott’s ‘proof’ is purely anecdotal. ‘Seven out of twelve’ of his wife’s expensive LEDs, he says, have failed, though he makes no reference to brand, the timeframe over which the LEDs have failed, or the conditions in which they were operated. My own anecdotal riposte is that I have only had one LED out twelve in my house fail over a number of years, and that was from a batch of extremely cheap, low-end LEDs.
However I think Scott’s right on one point: It may be beneficial to look at the ‘total circuit solution’. Perhaps we should ask manufacturers to put their money where their mouth is in terms of guaranteeing an overall package lifetime?