Yesterday we had a power cut for a little over 2 hours from about 11:00 to 13:10. The inside temperature fell during the time of the power cut and rose again as soon as the power came back.
The central heating had not been on, so the lack of it during a power cut is not an issue. That part of the house gets some sun from south-facing windows, and a little bit of heat from the gas Aga which remains on during a power cut, but that is in a kitchen about 50 feet (two rooms) away.
I don't remember the sun going behind the clouds during that time. The weather station is positioned so, as much as possible, it is shaded from the direct heat of the sun all day.
And yet the inside temperature (green line on graph) recorded by my weather station fell gradually by about 4 deg C between the start of the power cut (about 11:00) and the time when the power came back (13:10). The time when the power went off is a bit vague: from memory it was around 11:00. The time it came back is easy to determine because the line after the power was restored and my Raspberry Pi running weather station software began logging every minute becomes more jagged due to readings every minute, whereas all those readings that were recovered later from the weather station's internal memory are at 10-minute intervals, smoothing out minute-by-minute variations
If it *wasn't* a spurious result due to the station losing mains power, can anyone think why a room that was not being heated by central heating would cool when the mains went off, and start to warm up as soon as the power came back.
The only thing that would not be on during a power cut is a couple of freezers, but they are two rooms away from the one where the weather station is. Would freezers really give out enough heat to warm a room about 50 feet away by 4 deg C?