The Challenge
A facility's 400 kWp solar array was generating expected output at the system level. No performance issues were apparent. What wasn't visible: the performance distribution across the eight strings that comprised the array.
What Became Visible
String-level monitoring revealed that four of eight strings were underperforming by 18–22% compared to the others. The performance difference wasn't random — two strings were affected by shading from a nearby structure (discovered after checking historical photos and noting shade progression), one string had a wiring connection issue (loose terminal), and one had a failed micro-inverter operating at 40% efficiency. At the array level, these underperforming strings dragged down overall efficiency by approximately 8–10%, an effect that was masked by the four healthy strings pulling the average up.
What Changed
String-level monitoring with per-string performance tracking and daily comparison to benchmarks. Automatic alerts when individual string output deviated by more than 10% from peer strings.
How it worked: Each performance issue was identified and corrected. Vegetation was trimmed to eliminate shading. Electrical connections were inspected and tightened. The failed micro-inverter was replaced. String performance normalized to within 2–3% of each other.
Results
each 18–22% below average
after vegetation trimming
loose terminals and connections
from string-level interventions
Array-level monitoring hides individual string problems by averaging them across all strings. When overall output is acceptable, underperforming strings go undetected. String-level monitoring makes every string visible — and problems become obvious.
Operational Reality
Most solar arrays operate with 5–15% performance variation between strings. The variation goes undetected until string-level data exists.