Miles 2016 CF Data - An analysis

  • Jed posted some links earlier to some ACTUAL DATA obtained by Mel Miles in a typical F&P CF experiment. http://www.mail-archive.com/vortex-[email protected]/msg112745.html This data in in an Excel spreadsheet and was constructed from copies of Miles’ lab notebook by a Coolescence employee (Cantwell I presume). Since there is real data to look at, I have done so, piddling off and on for a few days now, and I have some comments on it.


    First, the data does not show Pxs (excess power). Apparently Miles calculated that elsewhere (not an issue), but we are not told what Pxs Miles is claiming for this data, just that it ‘shows excess’. Miles uses the ‘standard’ F&P calorimetric method (which I have commented on in my whitepaper, it can’t model a CCS), and in that method, the Pout is given by a “k” multiplied times the temperature difference between the cell and a reference (usually a constant temp bath, as is done here). Miles does calculate the “k” for the two different cell temp measurements he records (T2 and T5 vs T3, T3 is bath temp, T2 and T5 are cell temp (T1, T4, plus 1 unlabeled T are all room temp – not used in calcs)).


    Cantwell says that Miles considers T5 to be ‘Bad’. I have looked at T2 and T5 as a function of time and have no idea why Miles concludes this. There are no indications to my eyes that T5 is ‘Bad’. It does tend to have a slightly lower value than T2, but that is common and calibration should cancel the difference out. But I will only use T2 here per Miles’ comments.


    So, that leaves us with two temps and the cell voltage and current, plus some calibration info, that allows us to calculate the k(Tc-Tb) = Pout. Given the cell voltage and current we can compute Pin total, but we technically need to account for the power going into electrolysis, which is simply lost (if there is no recombination) in this cell. That is done by Miles by subtracting the thermoneutral voltage (1.52V) from the cell voltage to calculate the net Pin. Any T rise producing more Pout that that is an ‘excess’, i.e.Pxs = Pout – Pin = “k”(T2-T3) - Pin,net (using Miles’ adjusted Pin) but this ASSUMES no recombination.


    The result is a very spiky Pxs time profile. Many (possibly all) of the spikes seem to be associated with transients, and thus are likely not ‘real’. If we ignore these, the maximal amount of Pxs we see is 50-60 mW. This Pxs level is roughly consistent with the paper referenced within the Cantwell description Jed points out. The maximum level of heat from recombination in this run is ~555 mW, so Pxs represents ~ 10% recombination.


    Miles does NOT measure the faradaic efficiency (i.e. % recombination). He just assumes it is zero. (There is one annotation from the notebook of a 5.2 cc D2O addition to refill the cell. 10% of that is ~0.5 cc, which is on the order of the error I have discussed here previously in the ‘2004 paper’. Of course there, Miles and Fleischmann (and Szpak and Mossier-Boss) measured an _excess_ of ~7% while there should be a deficit of ~10% or so, indicating another process is adding water to the effluent.)


    In any case, the computed excess heats are typical, and explainable by a small level of recombination.


    From another angle, studies of the methods Miles uses suggest it really is pretty good, giving ~1%RSDs (relative standard deviations, 1% = very good, 5% = OK, 10% = barely acceptable). The maximal input power is ~2.265 W. 1% = .02265W or 22.65 mW – which is the 1 sigma level. 3 sigma is then ~ 67mW, so the 50-60 mW Pxs we calculated falls within the overall estimated random noise level. That doesn’t take any CCS into account.


    Therefore I conclude that Miles has not eliminated ATER and/or a CCS as the problem (or even recombination in general). But primarily I conclude it is unlikely he ever will address the ATER/CCS problem, so it’s up to us to cross-check all his conclusions regarding excess heat. Claiming excess heat from this experiment is not really kosher.


    I will be taking the holidays off (partially), so responses to this may be slow.


    Merry Christmas!

  • Experiments producing milliwatts of excess heat at low COP are a waste of time.



    Not exactly. If the milliwatt signal is well above the noise level it can be quite significant. That is why properly evaluating the error level is very important, and why the CFers refusal to deal with the CCS issue I raised is also so significant. MY's "night and day' difference can occur at any power level under the right conditions.