Published April 2009. See related articles:
Update to CFL article: June 2009
More CFL experiments: Sept 2009

My thoughts on Compact Fluorescent lamps (CFL’s)
(also called "Energy Saver" lamps)


I am pleased the NZ government reversed the proposals to ban incandescent lamps from sale. Not because I am against reducing domestic power consumption -far from it. I am against the ban because it would have achieved very little national power saving, if any. In addition, there are a lot of reasons why CFL’s are not suitable to replace incandescent and few reasons why they are.

Why did the previous government want to do this. Well, politics aside, if someone told you that you could replace all the light bulbs in every house with bulbs that used only 1/5th of the power and lasted 5 times as long, wouldn’t you be interested? A country could avoid building new power stations for another ten years…

The government was targeting only domestic users in the ban because industrial and commercial lighting is already using fluorescent and other high-efficiency light sources. So what proportion of the country total electricity consumption are we talking here. The MED papers on this subject say that domestic users account for about one-third of the total national power consumption. The rest is commerce and industry. Thus, any potential power saving is a maximum of 1/3 of the national consumption.

Estimating domestic lighting usage

Now, how much does lighting in your house account for in your total bill? No doubt this varies depending on how many people in the house and how proficient they are in turning off lights that are not required. We need an average. In winter you might need lights for 5 hours each night. In summer, perhaps 3 hours. The summary of lighting in a house might look like this:

Lounge 5 x 75 watt running 5 hours = 1875Wh per day
Kitchen 2 x 100 watt running 3 hours = 600Wh per day
Hall 2 x 60 watt running 5 hours = 720Wh per day
Bathroom 1 x 100 watt running 1 hour = 100Wh per day
Bedrooms 6 x 75 watt running 2 hours = 900Wh per day

There are no doubt other lights like garage lights, outdoor lights, but we will offset them against the calculation above based on winter months, knowing that in summer we operate for much less time.
Based on the above, household lighting consumption is then 1530kWh per year. This is about 17% of a typical household annual total consumption. Yes, this is a simplification and I know your mileage will vary but we need an average. In fact, many studies put lighting at between 11% and 18% of total household consumption over a year. That is, if you are using all incandescent lighting.

Now the previous government didn’t expect you to replace EVERY incandescent bulb with a CFL. You can’t replace the outdoor flood lamps, nor any that are frequently switched on and off, nor the specialist halogen types. Let’s say you replaced half your old bulbs with eco bulbs. Then your savings would be a maximum of 8.5% of your annual bill. Of course, savings will be less than that, because the CFL’s do use some power, and it is more than they state, as will be discussed further.

Now some technical stuff.

Pros and Cons of CFL’s

1. They last longer.

Well, usually, unless they are switched on and off a lot, or are used in a fitting which doesn’t let the heat escape. I unwisely used one which was controlled by a heat sensor in a doorway. It stopped after 3 months. So did the next one. I went back to incandescent.
I have seen CFL ads claiming that they will last 10,000 hours. That is over 5 years life, if used for 5 hours a day. How are you going to check that? What are you going to do if it stops after 2 years? Find your old supermarket receipt and take it back? I don’t think so. Actually I have some ordinary incandescents that have lasted much longer than 2 years. Most last at least a year.

2. They use less power.

This is the biggie of course. Suppliers claim you can replace a 100 watt incandescent with a 20 watt CFL and get the same light output. Well, you can’t. I have made some brief measurements with a light meter comparing a 60 watt incandescent to an 18 watt CFL. The CFL only produced 85% of the light output of the incandescent and that was only after it had warmed up. Yes, the CFL only produced 37% of its light output initially, finally reaching full brightness after some minutes. If you need to replace a 100 watt bulb, you will need at least a 23 watt CFL and a 26 watt one is closer to the truth.


Warm Up time

Several minutes to produce their normal light output. This makes them unsuitable for halls and stairwells, where you want the light straight away.

Purchase Cost

Higher than incandescent bulb. Between 5x and 8x the price.

Energy used in manufacture

Higher than incandescent. I have read a couple of papers about this and the suggestion is that the energy used for production and disposal is 3 times that to make an incandescent bulb.

Power factor and harmonics generated.

Some brands have a power factor of only 0.5 and less. This means that while you only pay for the wattage stated on the bulb, the country has to generate and be able to distribute twice as much power. However I did find one brand that had a power factor of near 1 and also had negligible harmonics. The measurements below are using a WT210 AC power meter.
Bulb, type, watts, VA, brand
60W, inc, 55.2, 55.2, generic
26W, cfl, 22.3, 44.6, not known
11W, cfl, 9.5, 19, not known
13W, cfl, 8.3, 24, radiant
20W, cfl, 18.1, 18.3, eco

The ‘eco’ brand which was heavily promoted and subsidized, via supermarkets about 2 years ago is actually very good. I don’t see these available now.
The harmonics of the current totaled about 2.5% in all cases but the ‘eco’, which had 0.2% thd.

Not suitable for dimmers

Actually, some versions are now made for dimmers, but make sure you have that sort. They will no doubt be even more expensive. Don’t put a standard CFL into a dimmer controlled circuit.

Not suitable for some lamp fittings

Even though they don’t produce much heat, what there is has to escape. Tight, close fittings, especially some down-light fittings are not a good idea. CFL’s have electronic circuitry in the base. Like all electronic circuitry, heat is the enemy. Get appropriate fittings for your CFL’s. Of course, that pushes up the cost even more.

More Cons
There are other issues to put you off. For a sensible, comprehensive discussion see Rod Elliotts article at

The national saving in power

Finally, did we decide just how much power would be saved by the country. Shall we say half the households replace half their incandescent bulbs with suitable CFL’s. Most of the CFL’s will be the cheapo brands with a power factor of 0.5.
Each new CFL should be marked as 1/4 the power of the incandescent. e.g. to replace 100W, use a 26W CFL. Actual generated power needed is twice this(due to the power factor).
Total percentage power saved is then 50% of 8.5% per household. Half of all households do this, so total is 50% x 8.5% x 50% x 33% of total power generated in the country. Overall a 0.7% saving. Seems quite small considering all the down-sides and I haven’t even mentioned the small amount of mercury to be disposed, which will be a lot when thousands of CFL’s are dumped over time.
Shall we wait for LED technology to mature.