I have a thermal imaging camera that attaches to my phone (toodon) and thought I’d point it at the hot water coming out of a tap.
I nearly collapsed when it was reading 56-57°C my hot water is set to 48°C.
My next reaction was to wonder if this was what was causing my heat pump issues completing hot water cycles in colder weather.
Something doesn’t add up. I have 4 sensors on my hot water tank two of which are in proper sensor pockets at the same height in the tank. One connected to my Ecodan FTC controller and the other connected to an Eddi. They both measure more or less the same temperature and the Ecodan one is what’s used to control when the heat pump knows the tank is up to temperature.
I removed some insulation from one of the higher up sensor pockets and pointed the camera down that and it was also reading 56°C.
My next thought was maybe the camera was not measuring water temperature well. So I boiled a kettle and pointed the camera at the pouring water, which read at just over 100°C.
Any thoughts? Maybe thermal cameras are not good at reading water.
My water temperature is typically around 5°C hotter leaving the top of the tank compared to my DHW sensor that’s in the middle. My setpoint is 47°C and I typically reach 53 at the top, but of course the heat pump is controlled by the center sensor. Generally those IR cams are great as long as you don’t point them at metal. Here’s an image of my own Topdon cam. The DS18B20 at the tank outlet pipe reads 50.1°C, so that matches quite well. Current center temperature is at 44.6°C.
The water at the top of the tank will often be a bit hotter due to stratification etc, as the flow temperature through the coil will be higher than the cylinder target temperature at the sensor. There also may be a bit of IR camera calibration error due to the emissivity setting.
For reference here’s how top- and center sensor vary over time for me. They can be massively apart when you quickly use a lot of got water and the center sensor only sees fresh cold water.
Looks like the emissivity setting might have been wrong.
This was last night before I changed the setting to 0.97 (think it might have been 0.75 before):
I’ve verified with a kitchen probe thermometer and that’s reading the same.
Unless for some reason last night the temperature was higher but my cylinder sensors are reading similar.
My cylinder setup is a bit weird.
I have a 250L cylinder that came with two sensor pockets mid way up where the coil is and another not far off the top.
I found that if I had my heat pump reading the temperature in the lower ones it would trigger every time we used much hot water (so during every shower) ao I tried moving it to the top pocket but this was too near the top to allow for enough hot water above it to complete a shower before it went cold.
So I drilled a hole a bit lower down though the outer casing of the cylinder and removed insulation so I could put a probe against the outside of the actual cylinder.
I have set the DIP switch in the FTC6 controller to make it think it has a heat exchanger for hot water instead of a coil, the only thing this setting seems to do is enable it to be controlled by 2 temperature sensors. You can then tell it to either use the bottom one to keep the whole cylinder hot after any use or to use the top one to just reheat when the cylinder is getting low on hot water.
This works well as the reheat is triggered by the top sensor but the bottom one is used by the heat pump to work out when its up to temperature.
The top sensor reads the temperature a bit low as its not in a pocket but my temperature drop setting allows for this.
My Eddi has a temperature sensor in the other bottom pocket and a second one that I’ve pocked through the insulation down the side of the unused secondary return take off.
I’ve just checked some boiling water with my thermal imaging camera with the new setting.
Although it starts saying 100C right near the spout of the kettle it falls to the low 90s quickly whereas I think that was when it was still measuring 100-102C last night.
So I think the emissivity setting was the problem.
