Well, I just discovered that the Plot.ly export function did not work. Does anyone have the Google drive link to the raw data?
The raw data must be analyzed as can has done and double checked. However it looks like the energy balance says that the experiment was null regardless of what the rolling average COP appeared on average. The rolling average COP was added at the last minute. I am not sure it was calculated correctly. If you look at the rolling average COP and its area above 1.0 and its area below 1.0, it looks like the energy balance should have been positive.