Utilities need to monitor the output of their photo voltaic (PV) arrays across the entire grid to detect and alert on equipment and production problems.
The output needs to be checked against equipment ratings and then correlate the output over all the PV arrays to account for weather, particularly cloud patterns.
Fractal developed a concept called the “virtual meter” which allows for the easy definition of PV groups.
Daily and interval data is analyzed – breaking out reporting by equipment type and program participation.
Machine learning was implemented to generate patterns and detect and categorize equipment issues.
Automated alerting and alarming at scale was then applied to the entire customer base.
This Fractal system was built and deployed in less than 90 days.