Regarding the graphs here: http://newenergytimes.com/v2/news/2010/SRI-Expt-HH/SRI-Experiment-HH.shtml
With data from this document, pdf p. 165: http://newenergytimes.com/v2/archives/1998epri/TR-107843-V1.PDF I don't see any problem or confusion with this. Maybe I am missing something. The bottom graph (Krivit's) is a little confusing. It would be improved with: 1. The error bars. 2. The power on the right y-axis starting at zero. With the power starting at zero it is apparent that the power is almost the same in every case, and the helium fluctuates from 1 (or possibly 0) to 3.5. It only starts to correlate with the heat in the last, highest power level. The text on p. 166 explains why. It seems clear to me: "These results while not always showing helium with the appearance of excess power, in NOT CASE show helium in the ABSENCE of excess power. Thus artifacts such as from air inleakage are not observed within the limits of He-4 detection of 1 ppb by volume. The absence of helium in 3 of the 5 samplings with the palladium alloy deserve some comment. The appearance of helium in the vapor phase of the cell implies a reaction at or extremely near the cathode surface. If not right at the surface, the absence of microcracks could be expected to trap produced helium in the metal phase. It has been observed that the alloy of Pd*Ce*Sm exhibits far fewer microcracks than the nominally pure palladium of the two Englehardt batches studied here. Hence depending upon the active cathode regions juxtaposition relative to microcracks could determine capture or release of the helium produced . . ." In other words it is so close to the noise it is more of a presence or absence test than anything else, and the likely reason why the helium is all-but-absent in some cases is clear. It is trapped in the palladium. In other studies they went to great lengths to release it from the palladium, by dissolving the metal for example, and they usually recovered the expected amount. This is not a study of cumulative helium like the Case cell. Based on Miles, my guess is that for this technique to produce a clear correlation you need 100 or 200 mW. Otherwise it is just a presence or absence test. The absolute amounts are so small that contamination from the atmosphere is ruled out, as Miles often points out. It would be far above this level. (Note: You cannot copy the text directly from the EPRI copy of TR-107843. It comes out garbled. I'll fix that tomorrow.) - Jed
