MZSA: Temperature control goes crazy at night

[MaPaKo]

I ran the MZSA model for the first five days of the year with fixed temperature setpoints for all zones of 20 °C for heating and 24 °C for cooling. Since the weather is cold, we are clearly in heating mode. During the day, the desired temperature of 20 °C is held, but during the night, the control seems to go crazy. Something similar happens during the 4th day with fast oscillations of the HP power but less impact on the temperature.

Can you tell me why this happens?

[dhblum]

Thanks @MaPaKo for reporting. Some thoughts:

• Night time: The behavior with your constant set points at night has to do with the system switching between "Unoccupied, night setback" mode and "Unoccupied off" mode as described in Table 2 in the test case docs.
• Fourth day: in the morning the AHU heating coil valve is modulated on a small amount but with some dampening oscillation to maintain the supply air temperature set point. The small oscillations in supply air temperature influence the zone air temperatures a small amount and also impose an oscillating load on the heat pump. See the plot below for illustration between 4.40 and 4.55 days.

[MaPaKo]

OK, thanks. Is there a way to override the occupation status, or would that require modifying the model? If the latter, would you consider making in an input in a future release? Not having control over the building half of the time makes it unusable for many types of research.