Simple Regression Models for Fully Air-conditioned Public Sector Office Buildings in Subtropical Climates

Abstract

A total of 20 fully air-conditioned office buildings with centralised HVAC systems in the public sector were selected for this study. Electricity use per unit gross floor area ranged from 163 to 416 kWh/m², with an average of 270 kWh/m². Monthly electricity consumption data over a five-year period from 1996 to 2000 were gathered and analysed. These data were correlated with the corresponding monthly ambient temperature and cooling degree-hours using simply linear regression technique. Regression analysis revealed that electricity use showed a strong correlation with both the mean monthly ambient temperature and cooling degree hours, though the latter performed slightly better than the former. The average coefficient of determination (R²) was 0.87 and 0.88 for ambient temperature and cooling degree-hours, respectively. This is only a first investigation examining the correlation between energy consumption and the corresponding outdoor temperatures, and thus has limitations. To have a better understanding of the energy use characteristics of public sector office buildings in subtropical climates, more work is required.