Vernon Electric Cooperative (VEC) in Wisconsin serves 13,500 members with power from Dairyland Power Company (DPC). Since deploying Advanced Metering Infrastructure (AMI) in 2016, VEC has improved reliability, power quality, and billing accuracy through 15-minute interval meter readings. This study evaluates VEC’s load management strategies. 

The research aims to validate current load control devices and includes objectives such as understanding data flow between third-party applications, setting up a database with recurring data imports, creating an entity relationship diagram (ERD), and utilizing 15-minute load data for machine learning forecasting. 

Data sources include DPC’s member site (SSO) and Mosaic, integrating data from the Meter Data Management (MDM) system and iVue. The project involves creating and populating a Microsoft Azure database, analyzing load management event shedding, and forecasting future load using Python. 

Key findings reveal significant seasonal variations in AMI values, with higher electricity consumption in winter and summer. Anomaly detection using the Isolation Forest algorithm identified unusual consumption patterns, crucial for improving forecasting accuracy. Analysis of load types such as air conditioning, commercial and industrial accounts, grain drying, and irrigation, showed effective load control during peak periods. 

Correlation and peak load analysis clarified consumption patterns and load management effectiveness. The load forecasting model predicted an increase in AMI values from 2025 to 2028, highlighting the need for grid planning. This study offers a framework for enhancing load management and forecasting, ensuring efficient electricity usage and improved grid reliability for VEC and its members.