Hmmmm . . . I screwed up this analysis, I think. Let me try again. First, I refer to this graph:
http://a2.sphotos.ak.fbcdn.net/hphotos-ak-ash4/304196_10150844451570375_818270374_20774905_1010742682_n.jpg I confused the issue by thinking about where the output line should be, given the likely recovery rate and the problems with the TC placement: Also look at the response during the initial phase when there was 2.8 kW of > electric power being input. I think it is almost certain there was excess > power during this segment. Maybe not as much as shown here, given the low > likely recovery rate, but there must have been some. . . . Forget all of those considerations. We know that excess heat always proceeds heat after death. We know that Rossi would not turn off the power unless there was clear excess power already. Let's take those to facts alone. Perhaps it is a coincidence that the recovery rate and the placement of the TC's and various other factors happen to compensate for one another, but for whatever reason the output power is shown here to be somewhat above input. Let us take this graph as a reasonable guess as to where it was. It was either a little above input or a lot above input. It could not have been below. This graph shows it being a little above, at around 3 kW. If that is correct, then after the power was cut, it went up to ~6 kW, as shown, later peaking at 8 kW. If that is an underestimate, it must have gone above 6 kW. It could not have been less than 3 kW, whatever it was, because as I said Rossi would not have turned off input power. The test would have been abandoned, as others have been in the past. - Jed
