AI for Public Power Utilities

AI is revolutionizing public power utilities, as demonstrated by CDE Lightband and the City of Santa Clara. These entities are leveraging AI to enhance efficiency, improve decision-making, and streamline operations. The integration of AI in utilities is not just a trend but a necessity for modernizing infrastructure and meeting growing demands.

At APPA’s National Conference in New Orleans, Brian Taylor, General Manager of CDE Lightband, and Lance Haynie, Government Affairs Director for the City of Santa Clara, Utah, showcased practical applications of AI for public utilities. Their presentations highlighted how AI is being used to address critical challenges in the industry, from load forecasting to customer segmentation.

AI Applications at CDE Lightband

CDE Lightband has adopted AI as a core component of its operations. The utility started by developing an AI policy, ensuring that AI applications are mission-critical or essential to the business. Using Google Gemini as its platform, CDE Lightband has implemented AI in various areas, including communications and marketing.

One of the key applications is knowledge retrieval. AI helps the utility access and analyze historical data, such as board minutes dating back to 1938. This allows CDE Lightband to research prior board actions, including approval dates and board member votes, enhancing decision-making and operational transparency.

CDE Lightband also uses AI for advanced metering infrastructure (AMI). By analyzing AMI data, the utility can cluster customers into similar usage patterns and segment them for targeted marketing. This approach was recently applied to a smart thermostat program, where AI helped identify 20,000 ideal customers out of a total of 86,000.

AI in Load Forecasting

Load forecasting is another area where AI has proven invaluable. The City of Santa Clara, Utah, has achieved 92% accuracy in load forecasting using AI. This high accuracy is crucial for determining whether to generate power in-house or purchase excess power from the market, especially in a region with predictable weather conditions like Utah.

The transition from a manual system to an AI-driven approach has resulted in significant cost savings. Previously, a scheduler would input weather data from sources like weather.com and load data from various sources. The new AI system pulls all data directly, iteratively sending alerts and fixing errors, leading to more efficient and accurate forecasting.

Enhancing Efficiency and Decision-Making

AI’s impact on administrative efficiency is substantial. The City of Santa Clara employs AI to condense lengthy meeting minutes into concise summaries, extract key information from documents, and reduce hours of administrative work. Tools like ChatGPT are particularly useful, requiring minimal technical expertise and providing immediate benefits.

The city follows a "Human in the Loop" philosophy, ensuring that a person is always involved in decision-making processes. This approach balances the efficiency of AI with the need for human oversight, ensuring that AI enhancements are aligned with the city’s goals and values.

Conclusion

The integration of AI in public power utilities is transforming the industry. From load forecasting to customer segmentation and administrative efficiency, AI applications are enhancing operations and decision-making. As more utilities adopt AI, the future of public power management looks increasingly data-driven and efficient.