Going off on a Light Tangent

During Bird and Jen's visit, Jen asked a question about lighting and efficiency (incandescents, fluorescents, and halogens). It got me reading through some of the background material on lighting efficiency; I thought it was interesting enough to warrant another energy post.

Jen's original question was, "how efficient are halogen bulbs, and can I feel a little less bad about using them instead of compact fluorescents?" I didn't know the answer offhand--I thought they were a bit more efficient than regular incandescents, but didn't know the number.

First, a bit on the technology of halogens, which I found pretty interesting (you might not find it interesting, but hey, it's my blog). Instead of having a tungsten filament in a vacuum (say, the same old technology from 1906), or in an inert atmosphere, the bulb is filled with a halogen gas, such as bromine or iodine. As a result, when a portion of the filament gets thinner (due to evaporation of the filament material) it is preferentially re-deposited at those thin spots, making the filament 'self healing.' This lets you run the bulb at a much higher temperature (that would destroy normal bulbs too quickly), thus, the nice pretty white light that you get out of halogen lamps. This was gleaned from a Wikipedia page, incidentially.

So, what's the efficiency, then? Well, that's not typically the metric used for lighting. A good explanation-in-a-nutshell is further down that same page (efficiency and efficacy of incandescent light bulbs). It has a table with efficacy of lights, in terms of lumens (visible light energy units) output per watt (energy unit) of input, which is pretty much what we want to know. So, typical values are:

Incandescents: 12-18 lumens/Watt
Halogens: 16-25 lumens/Watt
Fluorescents: 45-60 lumens/Watt
Light-emitting diodes: 60-100 lumens/Watt (prototypes)

So it seems to me that my original guess was right (halogens are a bit more efficient), but that the difference between them and fluorescents is huge.

To look at it another way, incandescent bulbs are turning 2% of the incoming energy into light, and fluorescents are up at 8%. So incandescents are basically space heaters that also produce a little bit of light on the side. This is assuming that the theoretical maximum efficacy is 683 lumens/Watt.

Also, to think of lumens/Watt in another useful sense, the maximum efficacy (according to that Wikipedia page) would be 95 lumens/Watt, given that any real material would radiate in the infrared and UV spectrum, as well as visible light. However, that would require a filament the same temperature as the sun... a slight engineering challenge.

Another problem that I have noted with halogens in use is the fact that they are often used to crank up the lighting to higher levels than used before. Just think about those 200-500 Watt halogen torchieres, and what wattage lighting was in that room before. So if you're a bit more efficient, but you're burning more light, you might end up with the same performance, or even worse. It's a bit like the way that people hear that cream cheese has fewer calories than butter, so as a result, they slather on a 1/2" thick slab of cream cheese on their bagels.

Of course, the problem with compact fluorescents is the light quality; it's sometimes referred to as color temperature (in degrees Kelvin: equivalent temperature of radiating body), or in CRI (color rendering index)--I think the latter is a more useful metric. The CRI is a scale from 0-100; an incandescent bulb is around 95, while a sodium vapor highway lamp is near 20. One noticeable things is that there's more of a range for fluorescent bulbs: see that CRI page--they range from around 60 to close to 90.

That's one important thing to realize when criticizing CFLs: the cheap ones really do suck, trust me. But nicer ones, like GE Genura (for recessed cans, see picture below), and Panasonics are much better than Happy-Elephant-Brand-Made-In-Mainland-China-And-Sold-At-Big-Orange CFLs. I installed these Genuras (82 CRI) in the can lights back at Pemberton Street, and I was very happy with them--pretty similar color to the overhead halogen in the kitchen. I also had the misfortune of installing cheaper CFLs in my current apartment, and the color rendering is just bloody ugly ("daylight bulb" my ass, thank you very much.) Therefore, my kitchen has one incandescent and one CFL bulb, to keep from being completely depressing.



As a final note about CFLs: when trying them out, you should realize that they take a little while to warm up--maybe about 30 seconds to a minute. So the wan initial light is not what you should be evaluating them on.