Sub-minute UH50 heat meter monitoring

The problem

E.ON Värme needed to see district heat behaviour between normal hourly interval readings, where fast control changes in buildings were hidden.

What Henrik owned

Henrik implemented the M-Bus protocol handling, built the Raspberry Pi edge setup and developed the Flask backend for ingestion into InfluxDB and display in Grafana.

Work delivered

• Read Landis+Gyr UH50 heat meters over M-Bus
• Built a Raspberry Pi setup with a USB M-Bus converter and Teltonika RUT950 modem
• Collected sub-minute heat meter values
• Built a custom Flask backend for ingestion into InfluxDB
• Displayed the measurements in Grafana
• Identified buildings where district heat flow control oscillated between full flow and zero flow

Technical approach

The edge setup read the meter locally over M-Bus and sent high-resolution values to a small backend, making short control cycles visible in InfluxDB and Grafana.

Where it stands

The solution ran for a period and showed district heat control behaviour that was invisible in normal hourly values. It was later dismantled.