The varying energy use levels of different market segments is indicated by the array of regression coefficients shown in Table 3.3. For each additional square foot of conditioned space, NAC increases by 4.65 kWh. An increase in household size of one occupant corresponds to an increase in NAC of 1,430 kWh. For each degree Fahrenheit increase in thermostat setting when a household member is at home and awake, NAC increases by 150 kWh. Location in climate zone 1 (the mildest weather) decreases the average NAC by 3,000 kWh. Similarly, the presence of a basement decreases the average NAC by 1,930 kWh. In contrast, the presence of air conditioning increases NAC by 1,810 kWh, and the presence of a forced-air heating system increases NAC by 1,490 kWh. All of these regression coefficients have the expected signs and are significant.
Thus, the fact that the EA and SGC programs have penetrated different market niches has significant energy-consumption implications. Within the various single-family market segments, however, the two programs have achieved comparable levels of energy consumption.
The results indicate that the energy efficiency of residential construction practices in Bonneville's region was significantly improved by its two MCS programs. The penetration rate of both programs in 1987 was 4% to 5% of new electrically heated single-family homes. However, the SGC program had a bias toward more "upscale" households, while EA homes include a more representative cross-section of the region's new home construction. After controlling for differences in the occupant and general housing characteristics of participants in the two programs, the annual savings of both programs is 1,900 kwh per home, or 0.9 kWh/ft2 per home. This is less than the Northwest Power Planning Council's predicted annual space-heat savings of 3.1 kWh/ft2 per home. The shortfall is due in part to the greater-than-anticipated energy efficiency of non-MCS homes.
The fact that the two MCS programs serve different market niches and have diverse energy impacts can be illustrated by the following analysis. In the absence of the EA program, it could be assumed that the twenty-eight EA jurisdictions would have been served by SGC utilities in the same way as the rest of the Bonneville region (i.e., with a 5% penetration rate of SGC homes). This would have resulted in thirty-four MCS homes rather than 671, totalling a loss of potential energy savings of 1.2 million kWh (637 homes at 1,900 kWh/home).