This issue will be resolved properly with a calibration. I couldn't resist noodling with the data I now have to come up with an estimate. A very rough estimate, let me repeat. I would not take this too seriously. This sort of thing should be measured, not estimated or guess-timated.
The fact that the entire reactor plus Dewar cools down when the pump is running proves that however much heat the pump adds to the water, the losses are greater, at least at temperatures 2 or 3 deg C or more over ambient. I have not found a place to sample a 1 deg C gap. You should be able to use Newton's law of cooling to figure this out for any temperature but I find the calculator gives a wide range of coefficients for this data. There are too many fluctuations and too much noise at lower temperature differences I think. In some cases it is clear to me the coefficient is different because there is a lingering source of anomalous heat. That is another use for the calculator: to look for heat. - Jed
