A local Streamlit dashboard for visualizing and analyzing historical data from a Steca TR A503 TTR solar thermal controller.

Daily View — temperature curves, pump activity, solar irradiance overlay

Yield Tracking — monthly energy yield with irradiance correlation

Degradation Signals — heat exchanger thermal resistance trend

• Daily View — Temperature curves, pump speed, power output, and flow rate for any selected day
• Yield Tracking — Daily / monthly / yearly energy yield (kWh) with lifetime totals
• Degradation Signals — Three long-term indicators:
  • Flow rate trend (pump/piping blockage)
  • Heat exchanger thermal resistance (scaling/fouling)
  • Collector year-over-year performance on clear days (soiling/degradation)

10 years of 1-minute data (~5.26 M rows, ~840 MB) never loads fully into RAM. DuckDB reads CSV files directly via glob patterns and returns only small aggregated DataFrames to Plotly. Single-day views (1,440 rows) use pandas directly.

pip install -r requirements.txt

Place CSV files exported from the Steca controller in the directory configured in config.toml (default: ./data_files). The expected structure matches the Steca SD card export:

data_files/
└── YYYY/
    └── MM/
        └── YYYYMMDD.csv

A flat layout (data_files/YYYYMMDD.csv) is also supported as a fallback.

Edit config.toml before first run:

[data]
directory = "./data_files"   # path to CSV root

[location]
latitude  = 48.14            # your coordinates (for weather API)
longitude = 11.58
timezone  = "Europe/Berlin"

[app]
clear_day_cloud_cover_max = 20   # % cloud cover threshold for "clear day"

streamlit run app.py

The Steca TR A503 TTR exports one CSV per day with 1-minute intervals:

Err values (sensor faults) are treated as NULL. The European comma decimal separator ("1,6" → 1.6) is handled automatically by both DuckDB and pandas.