Underfloor heating on the upper floor - one room always remains too cold

And if you close them via KNX, does the flow indicator also change (to zero)?
Or does it stay there?

Goes to 0

I suspect you don’t mean SMT but MDT?
Exactly that gauge I have installed myself.
Some of them need a 1-degree correction!
So measure again with a precise thermometer.

Yes sorry, I meant MDT. As I said, we all even checked with a calibrated thermometer. I had borrowed it from work.
Rooms showing 21.5 degrees also feel comfortable, the one room with 19.9 is noticeably colder when entering. From the “feeling” the switching of the actuators corresponds to the measured temperature. As said, the measurement showed no anomalies here.

You won’t have any chance with an expert if you don’t increase the supply temperature and show that it still won’t get warm...

You’re probably right, that was more of an emotional reaction on my part out of desperation.

Basically, I have the feeling you’re looking here for the one, all-solving answer.
But a hydraulic balancing is tedious work over several weeks.
So proceed as follows:
- Calibrate all thermometers or use exact external ones
- Ensure that the Taco setter valves show correctly
- Disconnect KNX control or permanently switch all actuators on
- then start throttling the too warm room a bit
- continue this until the temperatures are right everywhere.

What’s already noticeable is that the smaller purple room has more flow than the significantly larger red one.
That can’t be right

Yes, I had hoped to have overlooked something in the matrix but apparently that’s not the case.

I can’t do the hydraulic balancing again at the moment...
Shortly after the heating season started, I set everything to manual at full flow. Waited two days. Noticed all rooms warmed up except “red”. Then started throttling the first room. That took another three days. Then the second room, which was okay after the first adjustment, and so on. That cost me over two weeks and almost drove me crazy. After that came thermal imaging camera, flow tester and we even adjusted the ventilation system up and down...
The result was: all rooms as I want them, except that one room just wouldn’t get warm.
We also checked the insulation in the cold attic above the concrete ceiling for gaps.

Honestly, I don’t know what hydraulic balancing is supposed to bring in this case? The fact that heat energy doesn’t get into the room is contradicted by the visibly fully open valves. About 3L/min arrive at both valves. You can read that off. The flow corresponds to that of the neighboring room. Heating circuit lengths and number are identical. So where does the 1.5-degree difference come from? I can’t reconcile that with a possibly missing hydraulic balancing.

To produce a 1.5-degree difference in room temperature under the same conditions, the flow would have to be 20-30% lower, which you would immediately see.

I’m now setting the valves to manual and 100% again and raising the supply temperature to 29 degrees to see what happens to that room; I can’t think of anything better right now.

What did the thermal imaging camera show from the floor? Everything normally warm? Do you have a thermometer to check the supply and return temperatures of the individual rooms?

Otherwise, I would really do the extreme test for a few hours. So turn off all other rooms (on the floor) and see how quickly the temperature reverses. So it is warmer there than in the other rooms. If that hasn't happened after 2 or 3 hours, then something is wrong with the room allocation.

If everything fits, then the only thing left is that the heating load of this room is simply higher than that of other rooms! So what is different about the room than the 17m² room? More exterior walls? Worse or more windows, skylights? Fewer heat sources (missing people, computers, TVs, ...)?

You can read it off. The flow rate corresponds to that of the neighboring room. Heating circuit lengths and number are identical. So where does the 1.5-degree difference come from?

Do you still have pictures from the construction phase? Maybe the circuits were actually installed differently than planned – for example, a 15 cm distance. It could be a construction defect that no one noticed. I don't have any more ideas either.

Thermal image of the floor was like in the other rooms, no noticeable difference. However, what I do notice is that the control valve in the office no longer closes. I had to turn the flow to 0 in the heating circuit distributor box. I slowly suspect defective/incorrectly connected actuators! Same phenomenon in the ground floor bathroom. In the KNX I set the motor to close and it does not close. The electrician probably has to check the signals before the heating engineer comes again. Maybe the motors receive wrong values due to incorrect address assignments and therefore regulate down in auto mode from time to time (at night?). During the day today, only the bathroom on the ground floor and upper floor caught my attention here. But there is also a wall heating connected there and maybe that is why they never close?! I have a phone call with the electrician soon, maybe I will be wiser after that.

The rooms are really mirrored including identical arrangement of windows. Heating load is also identical. I really can't explain it.

Photos from the shell construction are available without gaps (more than 6,000 pieces). The piping is correctly assigned. Also checked again. Retrofitting a reconnection would have been impossible anyway due to the tightness in the box, especially with these stiff pipes and short distances from valve to floor. Furthermore, it would not be possible because sleeves are over the pipes.

In any case, thanks for your numerous ideas, I will now make sure that the electrician checks his work again and if that still does not lead to the desired result, I will have the heating engineer come again.

Do you still have pictures from the construction phase? Maybe the circles were actually laid differently in reality than planned – for example, 15 cm spacing. It could be a construction defect that no one noticed. I don't have any other ideas either.

No, they are even partly only 6-8 cm (bathroom). To achieve almost identical pipe lengths as much as possible.

Remove ERR and perform a proper thermal balancing.

Calculated heating load and reality are two different things. Solar/internal heat gains are not considered in the heating load.