Yet another geek blog

13 January 2009

Fluorescent with Rage about CFL Bulbs

Incandescent light bulbs increase your heating bill.

My New Years resolution was to stop getting would up about idiotic science stories in the press. But if I read one more time that incandescent light bulbs reduce your heating bills then they will be able to plug me into the national grid.

The Daily Mail has declared war on energy efficient light bulbs and this bizarre myth seems to be everywhere:

It was even reported as a fact in a pro-CFL article in the Guardian, although here it was only thought to be true if no lampshades were used.

So Here Comes the Science Bit

  • Incandescent bulbs are energy inefficient.
  • Therefore they are hot.
  • Therefore they heat your room.
  • Therefore your heating system needs to do less work.
  • Therefore, (if we ignore the cost of running the bulb), your bills are lower.

The bit in brackets you will only read here.

What does 'Efficiency' Actually Mean?

It can mean many things. In this context, it is a measure of how much of what you want you get for your money. So, I have a bar-heater and a light-bulb and I measure how much heat and light they give me for every pound I put in to the meter. My bar-heat gives me 10 units of heat and 1 unit of light and my light bulb that gives me 1 unit of heat and 10 units of light.

What is the cheapest way for me to heat my room to 20°C? Obviously the bar-heater is 10 times cheaper. So, for the very lowest bill possible, 100% of my heat must be bought from the cheapest source. If any of my 20°C is bought from anything but the cheapest source then the cost will be higher. The hotter my bulb becomes, the more of my 20°C will be provided by it and the higher my bill will be.

I need to spell this out:

Heat from anything but the cheapest source can only increase the cost.

So Why the Confusion?

I can only imagine that people somehow believe that the heat from a bulb comes free with the light.

The lamp shade idea is presumably caused by the fact that you cannot directly feel heat being radiated through a shade. But the bulb is heating the shade and the shade is heating the air so all the heat is still there.


posted by Yet Another Geek @ Tuesday, January 13, 2009


  • At 29 January 2009 18:02 , Anonymous Nick said...

    "So, I have a bar-heater and a light-bulb and I measure how much heat and light they give me for every pound I put in to the meter. My bar-heat gives me 10 units of heat and 1 unit of light and my light bulb that gives me 1 unit of heat and 10 units of light."

    Actually, if for every unit of energy you put in, your bar heater will give you 9 units of heat and 1 unit of light, and an incandescent bulb gives you 1 units of heat and 9 units of light, you will end up with exactly the same amount of heat in your room.

    When light hits anything that doesn't completely reflect it, a portion of that light energy is absorbed (converted to heat). The remaining light bounces off and hits something else. This continues until 100% of the energy has become heat.

    The only exception is any light which leaks out of windows, etc.

    Seeing as incandescent bulbs are only about 3% efficient to begin with, this effect is negligible.

    Of course, any time you wouldn't have the electric heat turned on, like in the summer, or if you live someplace hot, this heat is waste.

    If you want to get really picky, the fact that CFLs are not a linear load, leads to increased distribution losses in the power grid. In the case where the heating effect of the bulb is useful (my apartment on a cold winter day), the actual lighting+heating efficiency of a CFL is lower than that of an incandescent. You don't directly pay for this lost power, since it's before your meter, but the environmental impact is there, and the extra expense on the electric company's end will surely be passed down in the form of increased rates.

  • At 30 January 2009 11:18 , Blogger Yet Another Geek said...

    Yes, I know that you end up with the same amount of heat. That is why I included the thermostat in my example to make that clear. I have had e-mails from a couple of people explaining that light bulbs produce heat.

    I am not denying that incandescent light bulbs (or any other energy using entity) produce heat. The point is that this heat is not helpful. It increases the heating bill, can only increase the heating bill, and more heat makes it worse not better.

    I will go through my imaginary example again. One unit of heat bought via the bar heater costs me 10p (£1 ÷ 10). One unit of heat bought via by bulb costs me £1 (£1 ÷ 1). Now I heat up my room to 100 units of heat.

    If I only use my bar heater, this costs me (10p × 100) = £10.

    If I only use my light bulb (I have magic insulation), this costs me (£1 × 100) = £100.

    All combinations of the two fall between these extremes.

    So, yes I know that the heat is the same heat but the money is not the same money. The most cost effective contribution of heat from by light bulb is zero. The more heat I get from my bulb the further from this optimum I get and the more I pay.

    An Analogy
    I eat like to eat ten apples a week. I can buy these cheaply from Adli on my weekly shop but sometimes I buy a more expensive one with my paper on the way to work. It is simply not true that buying the expensive apples helps me with my weekly shop. Arguing that the apples are all the same does not affect this. Producing an Aldi till receipt for nine apples does not affect this.

    Another Analogy
    I need to get by boat across a lake as quickly as possible. I can use my outboard motor or row. The fastest way of doing this is to use just the outboard motor. It is quite true that rowing moves me closer to my destination. It is also true that rowing more moves me closer still. It is indeed counter intuitive that more rowing slows me down more than less rowing. Nevertheless, this is indeed the case because any time at all spent rowing is time not spent motoring.

    Non-Linear Loads
    I take the point about non-linear loads. But as Nick said, that cost is not borne by me and so is outside the point I am making. Also, even if you add in the power lost by the distribution network you should still be using less power than using the incandescent light. The issue is summarised in Power Quality Implications of Fluorescent Lamps in Residences which finds no evidence of a problem.

  • At 31 January 2009 03:25 , Blogger nwimpney said...

    I understand your original explanation completely.

    The point that your missing is that a 100W lightbulb puts out 100W puts out about 97W of heat, and 3W of light. That 3W of light, as it's absorbed by objects in the room (except for a small amount which goes through your windows, etc) will convert into 3W of heat.

    a 100W electric heater will put out about 99.9W of heat and .1W of light

    That .1W of light will be absorbed by objects in the room, and convert into .1W of heat.

    in both cases you have exactly the same heating efficiency. 100% There's no such thing as an inefficient electric heat source. Watts power in = watts heat out.

    There really is absolutely no difference.

    Regarding the non-linear loads. What I said before is accurate only when it's cold any you would otherwise be running electric heat. In other conditions the large power savings more than make up for the distribution losses.

    I use a variety of lighting in my apartment. I have a CFL in my bedside lamp, just because I had one handy last time the incandescent burned out. I have one in my kitchen, which would be an appropriate place if it wasn't an enclosed light. As is, it'll probably die prematurely.
    I use incandescents in my hall and entryway, since they're usually only run 2 minutes at a time when I'm entering or leaving.
    In my living room and work area I have a 2 bulb fluorescent, and a 175W metal halide.

    I often work on things where I need bright high quality light. CFLs just aren't there yet. I haven't seen one yet that doesn't distort some colours quite badly.

    I guess the point is, CFLs aren't bad, but there are places where incandescent sources make more sense.

    If I switched the few incandescents I do use to CFL there would be absolutely no benefit. The lifespan probably wouldn't be extended a whole lot, and possibly even reduced. (Their rated lifespans are tested with 2 hour run cycles)

    The very small amount of power saved would be offset by an equivalent increase in my heating bill (electric heat)

    And then add in the environmental impact, both from manufacturing and disposal.

    Not much harm, but absolutely no benefit. So why ban incandescents?

    The lamp shade, BTW, makes absolutely no difference. You could paint the light bulb black so thick you can't tell if it's off or on, and it will still convert 100% of the energy consumed into heat.

    Seriously, I believe very strongly in conserving energy.
    I also understand the physics well enough to know that any time I need my electric heat running I can get the exact same heat at the exact same cost by running any electrical devices in the room, and gain the additional benefits of music, light, or whatever.


  • At 4 February 2009 17:47 , Blogger Yet Another Geek said...

    Indeed, the ultimate efficiency of any heat generating device is 100%. And I know that lampshades do not affect this, see the last paragraph of my post. It was the person I was linking to who thought that they did.


    But, saying that all rivers flow to the sea is not the same thing as saying that all paths to the sea are of equal value (however 'value' is defined).

    For example, I am standing out side on a freezing day and somebody offers to sell me some ice-cream; should I buy it? Yes I should. True it will give me a sensation of coldness, but its calorific value more than compensates for that. Will I buy it? No. It makes me feel cold. I cannot feel the calories. What is worse, I will happily accept a nip or whisky to 'keep the cold out'. Does the whisky warm me? No. But it feels like it does! Similarly, I will pay extra for carpets in my lounge because they 'make it warmer' than lino. Neither carpet nor lino have the slightest effect on a thermometer, but the value that I am putting on them is represented by money that I am paying and is not defined by a reading on a thermometer.

    If this sounds defensive then I don't mean it to. My original posting was fired off in anger, uses imprecise terms and units I made up for illustration. I am grateful for being forced to examine it properly.

    For example, I said "What is the cheapest way for me to heat my room to 20°C?". What I mean is "how can I spend the least amount of money to feel comfortably warm?". This 'feeling comfortably warm' involves, but is not restricted to, the amount of heat in the room. For truly, I do not care about the temperature in degrees centigrade of my house; that is why I turn the heating off when I go out; that is why I would spend more on heating if I could not have carpets.

    The ultimate ability of something to turn electricity in to heat is defined by the wattage it consumes. But it does not follow that the heating engineers of the world have been involved in a mass conspiracy to hide this fact and make us buy new heaters. Because they know that there is more to heating than heaing.

    Imagine, for example, a room with a radiant heat source that was uncomfortably hot at one end and uncomfortably cold at the other. Next to it is a room with three such heaters spaced out for comfort. Thermodynamically, they are both 100% efficient. You could even make a case for the second room being slightly less efficient because of the extra wiring. But they are not the same according to my definition of efficiency which is the money I spend buying a comfortable room. The money I spend on electricity in the first room buys me one third of a comfortable room. The same money spent on the second room buys me a whole comfortable room.

    This is why I said that efficiency "can mean many things. In this context, it is a measure of how much of what you want you get for your money." I am not engaging in linguistic trickery. I am trying to make it clear that I am not using the word 'efficiency' in the same way that it is used in thermodynamics.

    The phrase "heating a room" is also imprecise. I place more value on the air being heated than anything else in the room. The Romans heated their stone floors.  These floors heated the air above them. Modern convection heaters heat air and then drive the warm air around a room. The second is more efficient for me because I am quite happy to just have warm air. If my floors, walls and ceilings are poor conductors then I care very little for their absolute temperatures.

    What about Bulbs?

    My central point is that not all ways of heating a room are the same. They may all have the same thermodynamic efficiency, but they are unequal in terms of the money that we are prepared to spend on them. In short, there is a reason that we do not use arrays of light bulbs as heaters- nobody would buy them. There are better ways to spend your money on keeping warm.

    This (after a long journey) returns me to my point. When there is more than one source of something (anything), and these sources are of unequal value (however you choose to define 'value') then the best thing to do is to get all of the something from the best source.

  • At 6 February 2009 03:58 , Blogger nwimpney said...

    Well, you wouldn't heat your house exclusively with lightbulbs strictly for aesthetic reasons (who wants 10 light bulbs heating their room) And that clearly isn't the point that I'm trying to make. What I'm saying is that your electric bill will not go down if you switch from incandescent to CFL in a room which is electrically heated (during times of year when it's being heated)

    I understand your point about efficacy of heating, and your analogy about the ice-cream in the winter.

    If you have carpet instead of lino you will be warmer. It's not about feeling warmer, it's about the fact that you are applying a heat load through your feet, and lino will drain that heat more than carpet, reaching a lower long term temperature. This is the basis of insulation.

    Your point comparing one concentrated heat source vs well distributed heat sources is also technically true (to what degree depends on the heat loss through your walls/windows etc).
    However it doesn't really say much to bolster your viewpoint. If anything the light bulb is an example of that distributed heat source which you find more comfortable.

    Certainly you can find heating scenarios which are extremely impractical while still putting the same amount of heat into the room, but lightbulbs in normal use aren't really one of them.

    Seriously, how is the heat from a lightbulb any less effective? I'm not saying everyone should switch to heating entirely from lightbulbs. What I'm saying is just that there's no savings.

    You set your thermostat to 20^C because you find that comforable.

    You then change your lightbulbs to CFL, save a couple hundred watts of power. The heater's just going to run more often to compensate for the reduced heating in the room.

    So, to your final point: "When there's more than one source of something...."

    Well, perhaps one reason is that you're defining "value" in a prejudiced way.

    The way I see it:

    An incandescent bulb is a good heater and inefficient but good quality lightsource (good CRI, little flicker, etc.) with relatively energy efficient construction.

    A CFL is a mediocre heater, and efficient, but slightly lesser quality light source (worse CRI, sometimes more flicker/noise) with a much less efficient construction, some (possibly minor) environmental/safety issues.

    End scenarios:

    Incandescent: electric bill unchanged, extra heat from lights is cancelled by reduced electric bill. I've got a nice light which I can quite clearly see colours under. Energy use of production. approx 1kWh per bulb. (doesn't cost me directly, but I care about the environment)

    Fluorescent: Electric bill unchanged. Lights waste less power, but also contribute less to heating causing heaters to run more. Colour of light is generally acceptable in most circumstances, but can sometimes be a problem. Subjectively, the light just seems a bit unnatural. Flicker hasn't really bothered me that much, but some people notice it. Energy to produce is about 10kWh (I expect this is a pretty conservative number though) They do last longer though, I'll divide by 5 (probably a bit generous IME) so 2kWh for the same amount of light-hours. Mercury? I dunno on this one.. It's a very small amount, but worth considering nonetheless.

    I will gladly tolerate all of the minor disadvantages of a CFL when there's something gained in return

    However, when I add it all up, I'm using twice as much power for production, the end product is a lot less environmentally friendly (not just mercury, but the electronics in the ballast, the plastic in the casing, etc), gives me a poorer quality light, and saves me nothing on my electric bill.

    all bad, no good.

    Use them where they're worth it. Locations where they're not enclosed (so the ballasts don't die early), where they're generally run for longer time periods, and where their efficiency is not entirely offset by the need for electrical heating.

    To anyone who doesn't know any better, use them anywhere they're not enclosed, and you're fine.

    But if you understand how it all works you can gain the benefits of an incandescent bulb where they're appropriate, and there's absolutely no downside.


Post a Comment

Links to this post:

Create a Link


If you use Firefox, and install Adblock, you miss out on this stuff: