Selected Stories Current Story Figure List Table List Home Page Order a Publication Previous Next

ost effectiveness of efficient magnetic ballasts There are several additional inputs required to assess cost effectiveness beyond those shown in Tables 1 and 2. We also need to specify operating hours, electricity prices, wattage savings, and ballast lifetimes. We describe these input parameters below, and then conclude these section with a discussion of the results shown in Table 3. Operating hours and electricity prices Two key parameters for any cost-effectiveness calculation are the hours of operation and the appropriate electricity prices. Many previous analyses have used averages for these parameters based on estimates of typical building characteristics. For this analysis, we rely on a survey of commercial building characteristics conducted by the US Department of Energy's Energy Information Administration. The Commercial Buildings Energy Consumption Survey or CBECS (US DOE 1991a, US DOE 1991b) contains survey data collected from a statistically representative sample of more than 5500 commercial buildings throughout the US. We began with the electronic CBECS sample, and extracted operating hours per week, floor area per building, total electric consumption, total electric expenditures and other relevant parameters for our analysis. We then dropped out about ten percent of the records. Most (about 80 percent) of these were dropped because the records did not have data in the field for "percent of lights that are fluorescent". The rest were dropped because they had no data for weekly operating hours or total electricity consumption (needed to derive average electricity prices), or were not operated during the past twelve months. Survey records for more than 5000 buildings remained after we cleaned the data.

We calculated implied electricity prices by dividing total electric expenditures by total electricity consumption.1 We then dropped out about 80 records that had implied electricity prices in 1989 dollars of greater than $0.20/kWh because such high prices are likely the result of data problems (the US average price was around $0.07/kWh in 1989). The weighted average prices from CBECS were then uniformly scaled down by 10 percent to make the CBECS average reflect the overall US average price in 1989, from US DOE (1991c). We adjusted the prices to 1993 dollars using the fixed-weight price index for consumer expenditures (Census 1994).

Our calculations of cost effectiveness assume that electricity prices remain constant. This assumption reflects the expectations of forecasters in the late 1980s, who expected little change in real electricity prices over the following ten to fifteen years (US DOE 1989).

Wattage savings per ballast

Estimated savings per ballast are 10 watts for the F40 ballasts and 15 watts for the F96 ballasts (Freegard 1988). We use the savings from two-lamp ballasts, which comprise the overwhelming majority of ballasts sold in the US.

Lifetimes

Ballast lifetimes are 33,000 hours for F40 ballasts and 36,000 hours for F96 ballasts, taken from surveys of ballast manufacturers, luminaire manufacturers, and lighting management companies with vast experience in replacing ballasts (LRI 1995b). The average lifetime implied by these figures and the CBECS hourly usage distribution is eight to nine years. For buildings where the implied lifetime would exceed 20 years (very low usage buildings) we arbitrarily assigned lifetimes of 20 years.

We use the same lifetimes for standard and efficient ballasts, even though Freegard (1988) states that the efficient ballasts last "twice as long" as the inefficient type. This improved lifetime is mainly the result of lower heat dissipation inside the ballast. The longer lifetime of the efficient ballasts would improve the cost effectiveness of these devices, though, as we discuss below, the internal rate of return calculations are somewhat insensitive to the assumed lifetime.

Economic characteristics of efficient magnetic ballasts

We show two measures of cost-effectiveness in Table 3: cost of conserved energy ($/kWh) and the real internal rate of return (IRR) for an investor choosing to purchase the efficient magnetic ballast. We begin by sorting the 5000 remaining CBECS buildings from lowest to highest operating hours. We then combine the data into three operating hour bins (low hours, medium hours, high hours). The first bin comprises the lowest 30 percent of the buildings, the second bin comprises the middle 40 percent of the buildings, and the third bin comprises the highest 30 percent of the buildings. As shown in Table 3, buildings with higher operating hours generally have lower electricity prices than buildings with lower operating hours, mainly because intensively-operated larger buildings are given favorable utility rates.

The cost of conserved energy (CCE) using a 6 percent real discount rate is at least 70 percent lower than the price of electricity in all cases. Using a 20 percent real discount rate, the CCE is at least 40 percent lower than the electricity price for each building cohort. The IRRs for our building cohorts using the appropriate operating hours and electricity prices range from almost 40 percent real for F96 ballasts in the low operating hours case to about 200 percent real in the high operating hours case for F40 ballasts (Figure 3). By either of these two measures, the choice of efficient magnetic ballasts is quite cost effective.

1) Our calculation of electricity price from the CBECS utility billing data implicitly includes the demand charges common in commercial buildings. Back to text  Last Updated On: 8/19/04