Date: May 2025

Utilities often struggle to identify when customers are using major end-use devices like HVAC systems and EV chargers—essential data for flexible load management programs. Without this visibility, demand-side management and marketing teams cannot target customers, limiting program effectiveness. To solve this problem, Alabama Power Company partnered with E Source to apply machine learning to 15-minute interval kWh AMI data. Our approach detects HVAC and EV usage patterns, classifies equipment types, and scores customers based on HVAC usage intensity compared to peers. The results are helping demand-side management teams refine their program strategies, focusing on more efficient targeting and resource allocation. In this session, we’ll walk through the process behind this data-driven solution and share lessons learned, including: using AMI data to detect major end uses; helping EV customers optimize charging based on time of day for maximum savings; and refining ongoing strategies to improve marketing and grid management.

Speakers:
Hayley Burns, TRC Companies
Will Gifford, E Source
Joyce Solomon, Southern Company