A procedure to group residential air-conditioning load profiles during the hottest days in summer

A methodology is proposed whereby residential customers can be grouped by their diurnal air-conditioner use profiles during the hottest days in summer. The procedure, illustrated with data from a load experiment study, involves three phases. The first is to form a subset of electricity use data corresponding solely to days when utilities are likely to face summer peaking problems. The second phase is to identify a characteristic diurnal A/C load profile for each customer which adequately represents his day-to-day behavior during these peak days. The final phase is to cluster these diurnal A/C load profiles into physically consistent discrete groups. The eventual practical applications of this methodology are that it could assist utilities in better planning and implementing cost-effective peak shaving strategies.