Saturday, March 6, 2010

Measuring Household Lighting: Survey Design Issues

Lighting: Credit NRECA
In developed countries we take electric lights for granted. We throw a switch, and presto the room lights up from ceiling to floor. But we all have experienced life without electricity during periodic brownouts and blackouts when we get out the candles for several hours. Unfortunately for the approximately 1.5 billion people without electricity in developing countries, there is no switch to throw.

One of the major benefits of rural electrification in developing countries is household lighting. In fact, people often ask my advice on how to measure such benefits. The issues can range from sampling to questionnaire design to sampling techniques. Thus, this is the first in a series of blogs which I will call Survey Design. For the first in this series, I am going to focus on how to measure household lighting. Stay tuned for more on surveys.

Lighting Efficinecy: Credit D. Barnes
People without electricity in developing countries generally light their homes with candles or kerosene wick lamps. The light given from a kerosene lamp is very dim and not great for reading and there is a good reason for this. For measuring lighting in households it is essential to understand that light is measured in lumens. A lumen is a unit used for measuring brightness. A kerosene lamp or a candle gives off about 11 lumens (see chart). By comparison a 100 watt incandescent light bulb provides 1,600 lumens, which is labeled on most packaging. The figure demonstrates that electric lights are much more efficient in converting energy into light. In a measurement study on lighting output it was found that a 100 watt incandescent light bulb typically found in developing countries actually turned out to provide 12.8 kilolumens of light per kilowatt hour compared to 0.1 kilolumens per kilowatt hour for a kerosene lamp. Fluorescent lights are even more efficient.

The reason for the difference between the lumen levels listed on the label on a light bulb package and the testing results has to do with the measurement environment. Actual lumen levels vary based on factors such as reflectors, lenses and location of the light in a room.

Efficiency of Lamps: Credit D. Barnes
How does that have an impact on household lighting? I will demonstrate this by using the results of a household energy survey from Peru. These results are based on actual use of various lighting sources for a national rural sample of 6000 households. As can be seen, the candle and kerosene lamps provide barely enough light to walk around the house. Car Batteries are used for lighting by households with higher incomes but without grid electricity and they provide more light. Due to the efficiency of converting energy into light, electricity from the grid is the best source for household lighting for households who participated in the survey in rural Peru. One may wonder why these household without electricity do not use household photovoltaic systems, but that will be the subject of another post.

For more resources on this topic click the link below. 

With these technical facts, we are now ready to turn to how these measurements were taken for households with and without electricity. For this, it is necessary to have a survey and set of questions on lighting use. Interviewers ask questions on the number of lights in the household and their daily use. I have included the actual questions from the Peru survey completed in 2005 that were used for preparing the figure in this blog along with the energy to light conversions factors. The questions look complicated but they are actually pretty easy to ask because rural people generally do not have many light bulbs in their homes. However, it is essential to have each commonly used lighting fixture listed in the survey. It takes quite a bit of space, but this makes it easy for the interviewers to ask and record each response. Thus for household with kerosene lamps you ask how long they used a particular lamp and from that you can calculate the amount of light that a household has available from all lamps by type. This is repeated for every lighting lamp and then totaled for the household. These lighting figures are essential for understanding the benefits of household lighting which is pretty evident to most of us only when the power goes off.

The studies upon which the conversons factors are based are published in several reports, and they include one in a research note on energy issues and another more technical study published in 1998 called Rural Lighitng Services:  A Comparison of Lamps for Domestic Lighting in Developing Countries.  There are some more recent studies than these but I include these articles because they were the original references for much of this work. 

Finally there are two studies in which I was involved and they include Energy Policies and Multitopic Household Surveys: Guidelines for Questionnaire Design in Living Standards Measurement StudiesFor an earlier version of this report that can be printed use this link,  but there are differences between this and the final version. 

The other study is cited in a the previous blog on the benefits of rural electrification. Finally, there is quite a bit of innovative information at the website of a program that is called Lighting Africa.


Maurice said...

I would be very interested in seeing the energy-to-light conversion comparison extended to LEDs. Having just been working on the Lighting Africa Policy study in East Africa, it seems that LEDs are quickly gaining a monopoly in small-scale off-grid lighting.


Douglas F. Barnes..... said...

I asked a reliable source and he indicated that the new lights put out about 75 lumens per watt compared to 60 for fluorescents. Another consideration might be length of service under real world conditions.