At one point in the 18 hour test the device was putting close to the same power as a compact car at full throttle. Even the most inaccurate measurement can't misreport that kind of power.
Of course it can. Let's keep this as simple as possible with a thought experiment - imagine a simple setup with just a water pipe and a heater. The pipe is connected to a tap at one end, the heater heats a section in the middle, and the temperature of the water is measured at the other end. The equation for this will look something along the lines of:
[latex]$$ T_1 = T_0 + \frac{P}{CF} $$[/latex]
Where T is temperature, P is power, C is heat capacity and F is flow rate = V/t. Obviously this is the idealised case with no losses, etc..
So if the final temperature increases, either the starting temperature has increased, the heater power has increased or the flow rate has decreased. For now, lets assume T
0 is constant and just look at ΔT. If ΔT doubles, either the heater power has doubled or the flow rate has halved.
That gets the maths part nicely out of the way. Now lets look at the setup in a little more detail. The water pipe is flexible rubber, the heater has a dial that can be set between 0 and 10W, and the flow rate isn't actually measured at all. We start it all running with the heater set to 5W, and measure 1K increase in temperature. Great. Now all this has taken lots of time so we go home for the weekend. The next Monday everything is right where we left it so we carry straight on with the next experiment with the heater now set to 10W. We measure T
1 to be 5K higher than T
0 was yesterday. Now, what should our conclusion be here:
1) The dial on the heater is not actually accurate, and since we're not actively monitoring the power it could be much higher than we think?
2) The flow rate has reduced, maybe due to the pipe bending or getting squashed?
3) It's been a really hot weekend and T
0 is actually 3K higher than last week?
4) Someone sneaked in over the weekend and hid an extra heat source inside the apparatus?
5) Cold fusion?
OK, thought experiment over. Now, consider Rossi's tests in the same way:
1) As already discussed, they didn't monitor heater power at all. If the heater didn't have the specifications expected, or if they deliberately used a different one as part of a con, then the actual power could be massively different for a given voltage. Multi-kilowatt heaters run happily off regular household power supply, so there wouldn't necessarily be any noticeable problem other than the higher than expected power.
2) As already discussed, it's a well known fact that flexible pipes do, in fact, flex, especially when under heating. It's also a well known fact that turbulence and cavitation can greatly reduce flow, especially when there is significant heating and vaporisation. I can entirely block off a normal rubber tube with my bare hand, so a reduction in flow even of 1000 times or more would easily be possible.
3) Probably not a significant factor here, but note that there was no monitoring of the input temperature, they just assumed it must be constant. Not enough to account for the results by itself, but certainly an entirely unnecessary source of error.
4) Again, already discussed at length. There's easily enough space to hide a battery or combustion powered heater inside the apparatus, as well as an additional heater whose use is not adequately described in the reports.
5) Hmm.
So, what do we have? Three obvious sources of error, two of which could easily account for the results by themselves or in combination, and all of which could easily be eliminated by any vaguely competent experimenter. In addition, two very obvious places (2 and 4) where fraud could easily be taking place, although only one (4) absolutely requires fraud. And again, both could easily be eliminated by any competent experimenter.
The only explanation for these observations is fraud on the part of Rossi and Prof. Levi. Inaccurate measurements and one con man is not enough to explain the results so far. Proof of fraud is required.
Wrong.
Evidence of fraud is required. We have plenty of that. It is, of course, not 100% conclusive proof, but it's easily good enough for us to say it's probably a fraud, and for virtually no-one in the scientific community to take it the slightest bit seriously.
However, see my points above. It's far from certain that Rossi is a fraud, he could just be incredibly incompetent. However, since all the problems could easily be eliminated by simply getting someone competent involved, fraud seems the most likely explanation. After all, even if he believes he's competent and that his tests actually prove anything, why would he refuse to address all the obvious criticisms in order to prove it to everyone else?
