---

Global Lighting: 1000 Power Plants

---

From space you can see the world's cities lit up at night, but how much electricity is being used? And, what about the fuel-based lighting demand for the two billion people who don't have electricity?

In a special IAEEL project, we developed the first global estimate of energy demand and CO2 emissions for lighting. In doing so, we created a new database of electric lighting energy use estimates at the national level, and a separate model for fuel-based lighting energy.

Based on an extensive review of the literature, our database currently contains electric lighting energy use estimates for 38 countries representing 3.7 billion people, 63% of the world's population and 81% of the PPP-corrected world GDP in 1997. (PPP stands for "purchasing power parity" and is a way of correcting cross-country currencies for buying power.) We found 36 national estimates of lighting in the residential sector, 27 in the service sector, 28 in the industrial sector, and 15 in the street- and other lighting sectors.

For the industrialized countries with available data, national lighting electricity use ranges from 5% (Belgium, Luxembourg) to 19% (Israel) of total electricity use, while in developing countries the value is as high as 86% (Tanzania). Few studies distinguish between urban and rural lighting electricity, although significant differences can be expected in the developing world. For example, households in urban Thailand use approximately 380 kWh/year of lighting electricity versus 110 kWh/year for rural homes. A similar ratio has been noted in Ghana.

Based on the country estimates in our database, we developed a simple model for predicting sector-by-sector lighting electricity use for countries where national estimates are not available. Using this method, we constructed a global estimate (based on 178 countries).

The corresponding lighting-related electricity production for the year 1997 is 2 016 TWh (21 103 Petajoules, PJ), equal to the output of about 1000 electric power plants and valued at about $200 billion per year. About half of this energy is used in IEA (International Energy Agency) countries, covering most of the industrialized world. Global lighting electricity use is distributed approximately 28% to the residential sector, 48% to the service sector, 16% to the industrial sector, and 8% to street and other lighting sectors. The corresponding CO2 emissions are 2 893 million metric tons (MT) per year, of which approximately 645 MT (22%) is attributable to the IEA member countries.

Two technical conservatisms in our analysis should be noted. We have not included estimates for the effect of lighting on air conditioning energy use, and we have assumed very conservative transmission and distribution losses of 10%, although the values in developing countries tend to be substantially higher.

$ERVICE $ECTOR $AVINGS

We examined service sector lighting in IEA countries in some detail. IEA service sector lighting energy represents 6% of total national electricity use on average and from 26% to 60% of electricity used within the service sector.

We identified a few studies that estimated the lighting savings potential for individual IEA countries. Among these, most focused either on a specific technology (e.g., CFLs) and/or on a specific policy option (e.g., ballast standards). The studies also differ in whether they provide a technical potential (with no moderating assumptions for partial penetration or cost-effectiveness) or a potential taking into account market or economic constraints.

For our assumptions, we have used a number of prior studies showing a conservative commercial sector lighting savings potential in the range of 25% to 40%. This represents a hypothetical policy pathway that includes a combination of modest standards and aggressive voluntary programs promoting cost-effective lighting efficiency improvements with today's technologies. In practice, savings will vary by country, depending on existing baseline conditions, etc.

As an illustration of the greater potential that may be achieved by considering the above-mentioned factors, the National Swedish Board for Industrial and Technical Development (NUTEK) developed a 64% high-efficiency lighting savings potential for the service sector. This number is particularly notable given Sweden's historical efforts and reputation as a leader in the field of energy-efficient buildings.

We identified an IEA-wide service sector lighting electricity savings potential of 133 to 212 TWh/year, corresponding to approximately 86 to 137 tons of reductions in CO2 emissions. The upper end of the savings range presented here is greater than the total individual national electricity use of Australia, Austria, Belgium, Denmark, Finland, Greece, Hungary, Ireland, Luxembourg, The Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, or Turkey. The lighting end use within the service sector offers greater carbon savings potential than other lighting sectors.

Assuming a representative electricity price of $0.10/kWh, the annual savings would be valued at $13 to $21 billion. Note that these rough estimates are "overnight" savings, i.e., based on today's consumption levels. Re-computed for a future date based on a growing "business-as-usual" reference case, the absolute value of the savings would of course be greater. The lighting end use within the service sector offers greater CO2 emissions reduction potential than other lighting sectors.

Another way to examine the savings potential is to compare the current-day commercial lighting energy intensities among countries. For a given level of gross national product, we can observe a factor of two (or more) variability in lighting energy intensities.

FORGOTTEN FUEL

Two billion people light their homes each day with fuel-based light sources, and in some regions this number is rising as population growth outpaces electrification (see IAEEL 2/99). Fuel-based lighting in the workplaces also appears to be very common, but the magnitude of this energy use is yet to be determined.

Our analysis of global fuel-based household lighting demand suggests that it represents an amount of primary energy of 2 400 PJ, equal to 43% of that used to provide household electric lighting around the world (ranging from 20% to 88%, depending on assumptions made, and $32 billion/year or 65% of the cost of household electric lighting (and 107% the cost of household electric lighting in IEA countries).

Combining fuel and electricity used to provide lighting raises the global total to 23500PJ, or about $232 billion per year, in lighting energy costs.
Due to low lamp efficiencies, per-household fuel-based lighting expenditures in poor households shockingly rival those seen by affluent households who enjoy the vastly higher levels of quality, safety, and services provided by electric light. Electrified households enjoy 300 times higher energy-service levels (measured in lumen-hours per capita) than households dependent on fuel-based lighting. Moreover, the cost per useful lighting energy services ($/lumen-hour of light) for kerosene lighting is 300 times higher than "inefficient" incandescent lighting and 1500 times higher than compact-fluorescent lighting. If the people currently using fuel-based lighting were to shift to electric lighting, with consumption levels equal to those in electrified households today, global household electric lighting energy use would triple.

DATA DAZE

Efforts to formulate global lighting energy use assessments are complicated by data and methodological uncertainties. For example, lighting energy use estimates can vary widely for a given country. This is strikingly evidenced by the 245 TWh and 340 TWh estimated by two studies of the U.S. service sector. The large difference between these two studies depends more on methodology than on the quality of data. The quality of data we find from country sources varies over a wide range. Some estimates are developed based on a simple "residual" analysis, i.e., allocating unidentified parts of the electricity balance to lighting, while others are based on extensive measurement and statistically validated surveys or careful bottom-up modeling efforts. Given the large known potential for lighting energy savings, it is remarkable how little effort has been expended by most nations to quantify the electricity used for lighting. While the collection of end-use energy data is arguably not a high national priority in most countries, this lack of attention is particularly problematic in this instance given that lighting is usually an early and high-visibility target for energy savings campaigns and policies.

It is equally remarkable how little data have been collected in the public domain on the lighting markets themselves (e.g., shares, performance, and utilization of specific types of lighting components in the stock and in new sales). The absence of such information severely limits our ability to formulate precise scenarios of future lighting electricity demand and to identify the savings that could be captured by new policies.

Evan Mills

IAEEL's global lighting database is an ongoing project. Individuals with country-specific lighting energy use estimates or associated data are invited to submit them to emills@lbl.gov.

Thanks to Benoit Lebot of the International Energy Agency for sponsoring the work upon which this article is based.

Top of page