Too often, distribution grid operators are forced to be detectives . They must use limited tools, visual line inspections, and incomplete data to hunt down the source of anomalies and disturbances that lead to outages on the worst performing feeders. They are often reacting to problems rather than pre-  emptively solving them. Grid sensors, AI (Artificial Intelligence) and new analytics models  are changing the game. AI driven analytics make it possible to identify potential problems that stem from vegetation and equipment    failure before they cause an outage. It is estimated that there are 5.5 million miles of local distribution grid lines in the U.S. , with several hundred million people and countless businesses relying on power that is always available. And, with the average age of many of these lines reaching 50+ years, utilities are searching for new ways to keep the grid functioning effectively. What’s more, the U.S. Department of Energy estimates that power outages cost the U.S. economy $150 billion annually. 

Many utilities are rapidly adopting artificial intelligence with data from advanced, high resolution line sensors,    to analyze and  proactively mitigate potential outages.  Predictive analytics technologies are among the most promising, and valuable advancements deployed to solve  the monumental challenge of meeting the nation’s ever-growing power needs driven by the electrification of everything. 

AI Can Detect Grid Anomalies and Pre-empt Faults & Power Outages

Sophisticated line sensors gather extremely accurate feeder data at targeted locations which is passed to AI-driven analytics models which then accurately identify, categorize, and report precursor anomalies that indicate impending faults.  An easy-to-use interface helps  operators identify feeder segments with high rates of these precursor anomalies, giving them the precise information needed to detect and prevent line failures and power outages.  Some of the advantages of this new approach to preventing outages are:

• More complete visibility of feeder system activity by location
• Greater reliability through predictive analytic indicators
• The ability to validate predictions with field outage results
• Predictive analytics capabilities with lower upfront investment
• Flexible deployment capabilities for simplified integration of line sensor data with OT systems and analytics modules
• Simplified grid management and monitoring at scale

These advancements are important because they enable utilities to more quickly locate and react to unplanned outages, reduce outage duration and frequency and improve reliability metrics. The combination of machine learning techniques, software and intelligent line sensors  can now provide indicators of impending outages giving distribution operators the ability to reduce the frequency of outages, providing their customers with what they need, affordable and reliable power.  The power of AI when combined with high resolution sensors is to detect subtle changes in the waveform that might not be visible to the naked eye yet still gives leading indicators of impending faults. 

Operators can access the data, reports and alerts that help them pinpoint anomalous activity by feeder segment so that they can proactively inspect lines, at the right location and with probable fault codes. AI ingests this inspection and validation data, in addition to waveform data, refining the analytic models, and improving accuracy over time.  The result is a learning system that improves grid reliability and saves utilities millions, improving safety and customer satisfaction.

The Forefront of Research and Solutions

Grid operators are at the forefront of research, strategy and technology advancements that can help reduce outages, modernize the grid, and improve service and performance for both customers and the utility alike. Working with operators of all sizes, Sentient Energy is dedicated to advancing data-driven solutions for our customers, and their customers. To learn more about our strategies, please visit our website.

Learn about Sentient Energy MM3ai System