Close

Background Info

A project log for Steam Heat System Temperature Monitor

Hardware and software system to monitor various temperature metrics related to a building's steam heat system.

synnoacksynnoack 12/12/2023 at 11:210 Comments

The origin of this project comes from a successful experiment to determine if it was possible to monitor water usage in the building having suffered a expensive leak that went unnoticed until the utility bill came. Using off-the shelf parts and software (rtlamr, RTL-SDR dongle, MQTT, InfluxDB, Grafana), we now have a close-to-real-time dashboard and alarming for anomalous events. In discussing some of the challenges of operating a steam heat system of this age, the current expert mentioned that if anyone notices this specific pipe being too hot to touch, we should call him to troubleshoot. That specific pipe was a steam condensate return line that should never be as hot as the steam supply line (~215 ℉). That condition is a symptom of one of the float and thermostat steam valves being stuck open which is simply a waste of energy. Our thought was that if we can monitor water usage, maybe we can monitor the temperature of that pipe. 

We have a robust WiFi network throughout the buildings, so monitoring that condensate return and using WiFi to collect the data seemed to make sense. With an ESP8266 board, a DS18P20 temperature sensor, and a little ESPHomeYAML, the remote temperature monitor was created. We were able to get the MQTT messages with the temperature readings into the InfluxDB and create Grafana dashboards. The monitor was powered by a single 18650 LiPo cell because of a lack of mains power in the area where the return pipe is located. The ESP8266 board was not optimized for low power operation and monitoring ceased after about 10 days. We did discover a steam trap that was defective, so the monitoring experiment was a success. A second revision included two DS18P20 sensors to allow direct monitoring of the supply and return connections to a steam trap. The strategy being that a functioning trap should have a temperature delta between the supply and return. Most commonly these temperature deltas are checked with a IR Thermometer, but unless the heat zone is calling for heat, a novice might not be able to detect a real problem. The investigation of a more long lasting remote solution began. The scope was also expanded to monitor the ambient air temperature in all areas because only a few of the thermostats are connected so that the temperature can be pulled into a monitoring system. Knowing how adjustments or repairs impacted comfort levels seemed very important.

One of the biggest challenges is the scale of the steam heat system, floor space involved, and the construction materials used. There are two not-so-recent natural gas fired low pressure steam boiler systems (one from the 1940's and the other 1960's vintage) which heat a three floor complex with a total combined floor space of 40,000 sq. ft. comprised of 46 rooms. The wide array of construction materials used in the various buildings is also a complication. Older portions are brick, concrete block or masonry. Path loss of the various wireless technologies in the environment is a major concern. An additional consideration is the availability of mains power for units. Since this is a multi-use building, it would be likely that a device would be randomly unplugged for some reason or that plugging something in permanently would not be allowed due to safety considerations.

Discussions