Friday, 26 June 2009

U.S. Energy Statistics

Energy Statistics is not an exact science and the more detailed information you seek, the more difficult it gets. Making estimates of national energy comsumption, and even energy consumption per sector, based on information from energy importers and distributors, seems to be a fairly straightforward matter of collecting and compiling available data. But when it comes to detailed splits of how this energy is used in each sector it gets a lot more complicated, time consuming and costly, especially in the domestic sector, as this requires surveying thousands of homes, in order to get a decent average of typical household energy use, as described here: Survey Methods

The official source of U.S. energy statistics, the Energy Information Agency has not published any new detailed household data since 2001 because it is so difficult to make correct estimates for such a large country with such varying climate zones, housing types and economic conditions. But as percentages often stay somewhat similar within a decade - unless something unusual occurs - the available data may still give a reasonably correct estimate even part of it is somewhat dated. (I've been informed that it is more important for accuracy to stick to the same source than to the same year, as different agencies may measure in different ways and include different things in their calculations.)

All sectors

- U.S. share of world primary energy use 2006: 21.1% [1]

- Total consumption 2008: 99.3 quadrillion Btu (British thermal units) [2]
End-use sector shares of total consumption 2008 [3]:
- Residential (private homes) 22%
- Commercial 19%
- Industrial 31%
- Transportation 28%

Residential sector

A. Applying the proper hierarchies (see my first post about energy statistics) we get the following picture by combining the available data from EIA:

- Total U.S. energy consumption 2008: 99.3 qBtu [2]
- Residential energy consumption 2008: 21.64 [2]
- Residential electricity consumption 2008: 4.71 [4]
- Lighting part of household electricity 2001: 8.8% [5] = 0.42 qBtu = 1.94% of household energy consumption or 0.42% of total U.S. energy consumption.

B. But not all household lamps are incandescent, as many have already switched to halogen, CFL or LED. In 2001, the incandescent percentage of household lighting was around 95%. Today it is estimated by EIA at around 70% and CFL at around 20%. Plus, many have dimmers and sensors installed and may already have cut their incandescent lighting consumption. But let's say 70% for simplicity's sake = 1.36% of residential energy consumption, or 0.32% of national energy consumption.

C. Of these 1.36%, CFL proponents hope to save 75% = 1.02% of residential energy comsumption. But, as shown under CFL Analysis - Savings Summary, the average CFL savings are realistically probably closer to 50% than 75% (especially if you include the covered CFL bulbs with lower output and all the cheap CFLs that do not meet EnergyStar standards) = 0.68% of household energy consumption, or 0.16%of national.

D. Not all of the incandescents left will be suitable for replacement. This can be due to luminaire restrictions (e.g. downlight or reflector); to the need for perfect colour rendering, immediate switch-on, sparkling effect or a relaxing, attractive light environent; to vision problems, hypersensitivity to UV or other medical reasons; or to plain dislike for the medocre light quality of CFLs.

E. When switching to energy saving lights, it's easy to start using more light instead, or to leave them on for longer, thinking energy savers use so little anyway (see Jevons paradox) or because the recommendation is to not turn CFLs on-and-off too frequently. Article comment to illustrate: "My dad switched to CFLs, but now he just leaves the lights on all the time because he says 'they use so little power, I can’t be bothered to turn them off'." Many CFLs are also supposed to be turned on for 15 minutes to 3 hours at a time in order not to shorten their life dramatically.

So, in reality, the potential savings we're talking about are in the order of up to 0.68% of total residential energy use or 0.16% of total U.S. energy consumption, probably less when taking D into consideration, and possibly none considering E.

Does this strike anyone as chasing very small drops in a very large energy ocean, while the BIG energy consumers: heating & cooling, water heating, transportation, commerce and industry can continue using up the remaining 99+% in peace as everyone thinks they've saved the planet by buying a few CFLs?

Commercial sector

- Total U.S. energy consumption 2008: 99.3 QBtu [2]
- Commercial energy consumption 2008: 18.54 QBtu [2]
- Commercial electricity consumption 2008: 4.61 QBtu [4]
- Lighting* part of commercial electricity 2003: 37.6% [6] = 1.34 Qbtu = 7.22% of commercial energy consumption or 1.35% of national.
*Including public street and highway lighting

There are many ways of calculating the percentage of each lamp type being used in commercial buildings: annual lamp sales; annual luminaire sales; current existence of different luminaire types, lumen-hours per year and lamp type, lit floor space, energy consumption per lamp type in Btu, in kWh etc. Finding recent info that goes into such details is difficult. Here is just one example, which may be still be reasonably representative (though most likely some incandescents are already replaced by CFLs):

Approximate percentage of lit floor space in commercial (non-mall) buildings 2003 [7]:
FL & CFL 74%
HID 9%
Incandescent 10%
Halogen 4%

So, even though lighting uses a larger percentage of the electricity consumption in the commercial sector, the incandescent part is so small - and shrinking - that replacing the few that are left is not going to make much of a difference energy-wise, only ruin the romantic atmosphere in restaurants, clubs, hotels and museums. (Offices rarely use incandescent lamps, except possibly a few in executive, reception or recreation areas, where their softer light may be needed.)

The greatest savings on lighting can probably be done by:

1. Upgrading existing poor quality FL tubes and luminaires in commercial buildings to T8 tri-phosphor tubes with electronic ballaststo metal halide spots/downlights or to LED spots/downlights.

2. Adjusting illuminance levels downwards in offices (as less light is needed when working online, than was needed when most office work consisted of physical paper work) and giving more individual control to users so they can turn down or off lights when not needed.

3. Using occupancy sensors and turning most office lights off at night.

4. Phasing out inefficient and poor-quality high-pressure mercury HID street lamps, replacing them with ceramic metal halide.  

5. Switching to LED traffic signals.

None of which requires banning incandescent lamps.


1. EIA: Annual Energy Review 2008 (p. 348) Figure 11.3
2. EIA: Annual Energy Review 2008 (p. 41) Figure 1.0
3. EIA: Annual Energy Review 2008 (p. 76) Figure 2.1a
4. EIA: Annual Energy Review 2008 (p. 263) Figure 8.0
5. EIA: End-Use Consumption of Electricity 2001
6. EIA: Annual Energy Review 2008 (p. 102) Figure 2.11
7. EIA: Lighting in Commercial Buildings 2003 Table 1

Tuesday, 9 June 2009

Energy Savers Review

Examples of energy saving lamps I've bought to see how they look in a home environment. They all looked good enough in the shop, but it's always hard to tell in the bright and predominantly fluorescent light-mix of a shop. Brief descriptions + my personal, subjective impressions of how they look in my home environment. (Note: lamp pictures do not reflect actual size, and prices are converted from SEK to Euro, include 20% VAT and may vary between countries.)

* 28W Osram E27 clear Halogen Energy Saver A-lamp

Info: CRI 100 (= full colour rendering). Costs about twice as much than its incandescent equivalent, uses about 20% less energy (though advertised as 30% less) and lasts twice as long.

Impression: Looks exactly like the 40W incandescent equivalent it's supposed to replace, though slightly brighter and with a rather glaring light point so best for luminaires with a shade.

* 28W Osram Spot R50 E14 Halogen Energy Saver reflector lamp

Info: CRI 100. Costs only slightly more than its incandescent equivalent, uses 20-30% less energy and lasts twice as long. Price only slightly slightly higher than incandescent reflector lamps.

Impression: Looks exactly like the 40W incandescent it's supposed to replace. And when I say "exactly", that means exactly and not "more or less similar", since halogen is an incandescent light, only concentrated into a smaller inner bulb.

* 30W Philips Master Classic E27 frosted Halogen Energy Saver A-bulb with infra-red coating and integrated transformer

Info: A low-voltage retrofit lamp that can be used in a standard mains-voltage luminaire. CRI 100. Costs over 10 times as much (€13) due to the built-in electronics, but then it lasts 3000 hours, so divide that by 3 and then deduct the 50% electricity savings and it's not so bad.

Impression: This one too gave a nice warm-white incandescent light that looked bright enough to replace a 60W bulb, as it promised. I could not tell it apart from a standard 60W frosted bulb.

* 7W Osram Duluxstar 'warm-white' E14 frosted CFL mini globe

Info: Appearance-wise, one of the most incandescent-like CFLs on the market, with a correlated colour temperature (CCT) at 2700K. CRI around 80 = standard (mediocre) colour rendering capacity. Price: about €10, but if you want a decent-looking (and decent-performing) CFL, be prepared to pay for it.

Impression: Visually, the light looked very soft and incandescent-like in the shop, but at home it still has a touch of that pink shade typical of flourescent light, though less markedly so than its early predecessors, more warm-pink than cool-pink, and admittedly an improvement compared with older CFLs and all the cheap budget lamps on the market.

As for colour rendering capacity, my do-it-yourself-spectral analysis with the back of a DVD shows the spectrum cut up into distinct bands with all the wavelenghts inbetween missing, as is normal for standard-quality FL light.

It does look bright enough to replace the promised 40W bulb (now in the beginning, will fade with age) though it took several minutes to reach full output. And the light was actually nicest before it did. Now it has turned a little more pink-white and makes the room look uniform and sterile. Many may not notice that much of a difference from an incandescent, or care if they did. But as I have a very well-developed sensitivty to such nuances, I could not relax in such a light and would never use it in my home.

* 1.2W Anslut 'warm-white' GU10 20-point LED reflector lamp

Info: 20-diode spotlight. Price was decent for an LED, just over 6€.

Impression: Don't quite know what to make of this one. On the one hand it's impressive to get so much light - at least in one direction - out of what is only 1.2W!

I picked this particular lamp because the light looked more white than the markedly green-white or blue-white I'd seen previously. It seems to have decent colour rendering too, both to the naked eye and in my DVD-test where I could see the full spectrum reflected without any large visible gaps (though no magenta).

Colour: Still slightly green-yellow-white (which is not surprising as 'warm-white' LEDs usually consist of blue diodes with yellow filters). Some may like this slightly cooler light (around 3000K, but gets a little warmer over time) but for my personal taste it still looks too much like FL light and gives my kitchen an industrial feel which dos not complement the warm colours and traditional design in a good way.

I have to say I was disappointed as I prefer mercury-free LED before CFL and would love to find a good enough LED to recommend instead. For commercial purposes fine, but not for home lighting unless that industrial feel is what you prefer. I'll keep looking.

(If you're a producer or retailer and have one you think is good enough, feel free to mail me and send me a sample.)

New (5 Aug):

* 1.8W Kjell & Co 'warm-white' frosted LED E27 mini globe

Info: Price around 12€.

Impression: This lamp is a joke. It's not even remotely warm-white, it's cool-white like a moon-beam, and about as dim. It gives only 65 lumen, less than a 10W incandescent, which is good for absolutely nothing. You certainly can't read in it and it's not warm enough to be used as mood-lighting (except at a Halloween party perahps). And this was the brightest LED globe light I could find in Stockholm retail stores!

Looking at the small print on the back of the package it says this lamp type is recommended "for decoration" or "for dark spaces like the cellar stairs, the attic passage-way, the garage or storage area". But it is not decorative, just dim and generally gloomy, now why would anyone want to put such a light in their cellar stairs and risk breaking their neck, or in spaces that are usually already creepy enough without adding a dim ghost light to it?

The only reasonable application would be as night light, but as this bulb requires a real luminaire with a full E27-socket, which makes it useless as night light too. (Instead, see my Coloured LED Review for a really great LED plug-in nightlight that costs only slightly more.)

New (9 Oct): Now I've sent for some higher watt LEDs from an online store that are supposed to be better quality.

* 4W clear 'warm-white' SMD LED E27 mini globe

Info: Price around 19€. Rated life 50 000 hours. 350 lumen or "about as much light as a 40W incandescent but using 1/10th the energy". Will not get warm, light up 100% in half a second. 

Impression: Yes, like all LEDs it lights up instantly and is luke-warm enough to touch even after being on for a while.

Colour: Warm-pink-white that looks similar to 'warm-white' fluorescent light rather than to golden-white incandescent light.

Brightness: Nowhere near that of a 40W incandescent. The 350 lm may be correct but a 40W incandescent gives 410-505 lumen and visual comparison between an incandescent 40W lamp seems to confirm it, so this seems to be another case of consumer fraud.

At the same time it is too glaring to the naked eye and must be used in a lamp with a thick shade so that the glaring little dots don't shine through. Which reduces its brightness even more as it is designed to throw light to the sides rather than downwards. Tried it in different luminaires. In modern table- & floor luminaires it doesn't work very well: what little light that finds its way out of the shade is very dim and gloomy indeed, and of no use whatsoever. A classic architect luminaire seems to be the only one it works with. The wide shade spreads the light much better than the very directional GU10 spotlight. In this luminaire it reading works if you can ignore the faint light dots reflected on the page.

Light quality: Like the other LEDs, the spectrum of this one is continuous in the warm end of the spectrum but spiky in the blue end, with no magenta. Colour in the room look sort of dampend, as if seen through a grey filter. Whatever room I try it in, it turns all gloomy and depressing. No life.

* 3W Cree 'warm-white' frosted LED E14 mini globe

Info: Price around 24€. 120 lumen or equivalent of a 25W incandescent. 50 000 hr life. Ceramic foot and chromed aluminium house.  

Impression: The frosted glass makes this one easier on the eyes and works well enough to read in. The socket limits its usefulness as its long heat sink makes it stick out too far in all the various E14 reflector luminaires I have. Putting it in a luminaire with a shade will reduce light output too much. The best fit would probably be in a vanity light for those who want a non-glaring white.

Colour: Cool-pink-white. More like fluorescent light and even less incandescent-like than the Osram CFL tested above.  

Brightness: Again erroneous equivalence info. An 25W incandescent lamp gives 215-235 lm so a 120 lm should not be enough to replace it. However, this one actually seems even brighter than a 25W incandescent, though the light itself has a duller quality.

Light quality: Continuous spectrum but with green, violet and magenta missing. Colours in the room tend to look a bit grey and faded and white surfaces look distinctly cool-pink, even though the bulb itself looks more neutral-white.  


More lamp descriptions can be found on this site:

Thursday, 4 June 2009

EUP Elections

Planning to vote in the upcoming EU Parliament elections?

Info on which parties are against the incandescent ban is not easy to find. These are the only ones I know for sure:

* Britain: United Kingdom Independence Party (UKIP), Conservative.

* Germany: Free Democratic Party (Freie Demokratische Partei, FDP), Liberal.

* Sweden: June List (junilistan, jl) cross-political, in EU Independence/Democracy group. Thinks the incandescent ban is a joke, or in Sören Wibe's words, "symbol politics and hypocricy".

* Sweden: Feministiskt initiativ (fi). Don't have a specific policy on such a marginal issue but think EU should concern themselves more with human rights issues than micromanaging the use of consumer products.

The other major parties in Sweden seem to be for the ban, including the few that are usually EU-skeptic, such as the left vänsterpartiet and green party miljöpartiet.

* EU: Libertas (pan-European reform party). Thinks EU interferes too much in general, that it is too far removed from the people, that more decisions should be returned to national governments, and is therefore against legislation such as the incandescent ban.

If anyone knows of more parties against the ban, feel free to mail or post in the comments section.