Load Shape Segmentation for Better Grid Stability and Increased Customer Benefit
Reducing greenhouse gas emissions through beneficial electrification and increased renewable energy production necessitates improvements in energy load management. Utilizing advanced metering infrastructure (AMI) data, this research project used a random sample of 50,000 residential customer meters to demonstrate the use of load clustering to identify typical load patterns within a population. The study identified load patterns that coincided with grid peaks and evaluated opportunities for technologies to reduce or shift usage out of key times of day. The research team also summarized load patterns across a sample of potentially low-income utility accounts to identify specific energy usage patterns within this sub-sector. The team also assessed an AMI features-based approach to summarizing energy use data, finding that specific features (such as baseload, heating and cooling degree day model slopes, and seasonal peak demand) can help identify use patterns well-suited for certain efficiency measures or programs. When comparing AMI data to nearly 2,000 previously completed efficiency projects (related to heating, cooling, and domestic hot water), the team determined AMI features can identify projects that are more likely to result in positive savings. The research team concluded that a combination of AMI filters and additional equity indicators could point out savings opportunities not just in the highest users but equitably across a population, improving targeted customer outreach and program design as well as further advancing electrification and managing the resulting increased load on the grid.