n8n Solar Production Loss Diagnosis: SolarEdge + Gemini + Slack

Diagnose Solar production losses automatically—using SolarEdge, weather data, Gemini, and Slack

This n8n workflow compares SolarEdge energy production against weather-adjusted expectations, identifies anomalous performance, and then uses Google Gemini to attribute likely causes—creating a Jira work order and sending alerts to Slack.

What this workflow does

• Runs monthly (scheduled) or on demand, then loads site settings like the date range, expected kWh, and threshold rules.
• Collects performance data by pulling daily production from the SolarEdge Monitoring API for the configured period.
• Adds weather context by fetching hourly radiation and cloud cover data from Open-Meteo.
• Builds diagnostic history by querying Postgres for the site’s last 30 days of performance and the most recent incidents.
• Calculates metrics including actual vs expected production, weather-adjusted losses, performance ratios, anomaly flags, and historical baselines—then upserts results into a Postgres solar performance table.
• AI root-cause attribution (when anomalous): sends metrics and historical context to Google Gemini to generate structured diagnosis details such as root cause, confidence, priority, and recommended actions.
• Operationalizes the diagnosis: stores the AI output as an open incident in Postgres, creates a Jira task, links the Jira key back to the incident, and posts an alert to Slack.
• No anomaly: posts a daily/period summary message to Slack.

Use cases

• Spotting monthly underperformance and reducing time-to-root-cause for SolarEdge sites.
• Turning abnormal production dips into Jira work orders with AI-generated recommended actions.
• Providing consistent, weather-aware production diagnostics across multiple solar assets.

Technical details

• n8n nodes/logic: if, set, code, jira, slack, postgres
• Integrations: SolarEdge Monitoring API (OAuth2), Open-Meteo, Google Gemini, Postgres, Jira, Slack

Video demo: https://youtu.be/7lmQu9wd46w