Mizuno reports increased excess heat

  • RB - I'm going in the data in the paper, plus what I could get out of Jed here. I'm sure Jed can resolve the matter.

    I have however not claimed 50% error - originally, befoe looking at the matter more fully, I said the error might be as much as 50%, and soon after that I tightened things up a bit. My best estimate of the error, having crunched the number inasfar as I can ATM, is 80% (20% error). And I have linked the calculations (and given the Re calculations, which other agree with) that indicate this. I did that in the post I made - not my first one, but when I was prompted to investigate much more fully and find numbers - a long time back. I gave my reasons and linked the source material from which I derived equations. I was hoping people like you would follow up on that.

    You should also note the way this works. If something is not done in the paper to tie down an uncertainty - but COULD be inferred from the paper results with some additional assumptions and calculation, I will claim that the figure is uncertain. Not that it is wrong bit that it could be wrong, and give the bounds as understood. Further investigation might (usually does) tie this down further reducing uncertainty - especially because we have Jed here to clarify issues. I should point out that the paper should be doing this. In this case the paper mentions the problem, but gives anenometer results that do not show what the paper claims they show. Hence I tried to investigate further.

    Not knowing right is different from knowing wrong. However, with these results that if real would transform physics most including me will treat not knowing right by default as equivalent to wrong, simply because transformations of physics don't happen that often.

    I am always open to new information to make my assessment of this data change (in either direction - you will see that has happened - both ways - on this thread).

  • I'm taking the power out from temp dependence from a graph M provides (it is supralinear - sort of quadratic). That makes stability worse than my conservative assumption.

    I'm taking linear temp from power out dependence from what Jed says about the calorimeter losses being relatively small < 25%) he gives figures and the known fact that forced air cooling provides roughly linear dependence.

    In any case this is a calculation, rather like a sensitivity analysis, that shows there is a large danger of instability in the COP = 6 - 10 R20, if all the claims from M are correct.

    I agree the test appears to run for a long time without instability, nor does M mention this as a problem when testing under different conditions, which is why I highlight this as an anomaly.

    My comments here are reasoned, and proportionate.

  • I'm taking the power out from temp dependence from a graph M provides


    you make assumptions all over the place

    and then you claim that

    Higher COPs mean instability.

    Would it be more wise to say

    ""based on my assumptions from control theory

    which may or may not apply to a novel nuclear reactor

    "Higher COPS may mean instability"""?

    that would in my view be reasoned and proportionate.:)

  • no , you should have read the figure more closely

    that is all

    I read it and I understood it

    the traverse width was greater than 5cm.

    This is actually quite annoying.

    The numbers given are 0,1,2,3 cm. That is a total traverse of 3cm. Or possibly D/2, 1, 2, 3cm

    My first thought on reading that was the same as yours, that the radius must be > 3cm because these were taken from the centre. I noted the inconsistency with the stated radius of 2.5cm. I asked Jed, who replied here at that time, that measurements were from the edge which of course makes it consistent if a bit strange.

    I concluded that these mean 2.5 (for centre) 1, 2, 3cm

    All that was done here, and you were reading it at the time?

    I agree that the paper is not clear, and I'd be happy for jed to clarify this matter.

  • THH

    from a graph M provides (it is supralinear - sort of quadratic

    One can explain the same graph with ten different theories

    using assumptions.

    you need to examine how many assumptions you have used in your theoretical derivation.

    please write it out not in this forum text

    but in algebraic notion like mathtext

    It is really hard to follow otherwise.

    It took me awhile to figure that this was assumption

    Let us take a simple example of F(T) = CT (linear)"

    You might also give a thought to how Mizuno's finding in 2017


    " The reaction activation energy Ea was calculated on the basis of the linear region

    between 100 and 523◦Cin Fig. 40 to be 0.165 eV/K/atom""

    is consistent with your control theory derivation

  • I noted the inconsistency with the stated radius of 2.5cm. I asked Jed,

    So you noted the 2.5 cm inconsistency and were so polite not to mention it to Jed.

    That is so polite... THHnew. Jed did mention your politeness.

    I didn't notice any inconsistency because I used the 0.0044m2 figure from 2017

    since Jed said not much of the setup changed except near the reactor... somewhere.

    Jed must be waking up soon.. sunup in Georgia.

    probably best for us to clear up any misunderstandings with him.. later today.

    • Official Post


    Further to your interest in low-cost experimentation I have one of these -used mostly not for LENR work but in the recycling part of our laboratory. Availiable via Amazon this are made from stainless steel with a very thick aromoured glass lid and chunky silicon sealing ring. They work suprisingly well, even the valves are good for the money, they will hold a vacuum without continuous pumping and only an almost imperceptible drop from Friday till Monday for sure. You can stand it on a hotplate or adapt it by putting in a pass-through for wiring. I think it is not beyond the bounds of belief you could use this for a LENR experiment.


  • I would like to put this in a new thread, but I can never remember how to start a new conversation. Anyway:

    I made some revisions to the Mizuno paper:


    Biggest changes:

    p. 1.

    Note: Readers have made valuable suggestions, so we have made some changes and corrections, mainly on p. 12. We appreciate the suggestions. This is the June 24, 2019 version.

    p. 12.

    Three meshes are prepared for one test. . . .

    Rubbing is done with a palladium rod, 100 mm long, diameter 5.0 mm, 99.95% purity. Before rubbing the mesh, weigh it with a precision scale. Then vigorously rub the entire surface, left and right and up and down. Turn the mesh over and rub the other side. Weigh the mesh again. Continue until the weight increases by 15 to 20 mg. The weight of the stack of three meshes should increase by about 50 mg.

    [NOTE: I was confused about this. I thought the 50 mg of Pd was for one mesh. My mistake, not Mizuno's.]

  • So what else is left? Platinum RTD's can react to electrical noise. But that should affect both reactors as well as calibrations.

    An RTD error large enough to explain this is out of the question. The first thing Mizuno does every morning is check the temperature readings. Anyone who does calorimetry does this several times a day. The inlet RTD should agree with the thermometers hanging on the wall, because the air is sucked in from the room. The two outlet RTDs are checked with the mercury thermometer or the Omega handheld TC. The two RTDs should agree with one another. If there was a problem, they would not agree. They do fluctuate and vary a little from one another, but these are real differences, not instrument artifacts. Example:

    26.42 26.50
    26.40 26.37
    26.42 26.40
    26.47 26.47
    26.47 26.47
    26.45 26.42
    26.40 26.40
    26.40 26.45
    26.42 26.58

    Average for 24 hours:

    29.94, 29.86

    The first one is closer to the fan, so it is ~0.1 deg C warmer. I think that is the reason. Before the test, they are removed from the calorimeter and are calibrated together in water over a range of temperatures, along with the mercury thermometer.

  • THHuxleynew wrote:


    Personally I most distrust the airflow measurements. Getting this wrong by 50% would explain these results.

    I do not understand this comment. If the airflow measurements are wrong by 50%, the excess heat would be ~125 W instead of 250 W. That would not "explain" the results.

    Also, as I have pointed out, if the airflow measurements were 50% wrong, the calibrations would not work; Fig. 2 would not be linear and it would be physically impossible; and the independent measurements of heat losses from the chamber walls would be completely different from the input minus output measured in the flow. A 50% error would be obvious. It would be impossible to miss. A 10% error would be impossible to miss. THH should address these issues, but he will not.

  • I would suggest changing the x axis of figures 5, 6 and 7 from hours to minutes.

  • There is some understandable confusion about the outlet. Here is some information from an unpublished paper. This is translated by me:

    "The blower exit is rectangular, 58 × 38 mm. The wind velocity in a square outlet is uneven. It is better to measure it some distance from the fan, so a 200-mm long cover pipe cylinder made of paper (Fig. 1-4) was attached in front of the blower outlet. One end is rectangular to fit the blower, and the other end is circular, 66 mm in diameter. This cylinder can be made of paper or plastic. 2-mm-thick urethane insulation is wrapped around the outside of the cylinder."

    The traverse test covers 6 cm, so that leaves out only 6 mm to the edge, 3 mm on each side. The wind speed 3 mm from the wall of the pipe is the same as the wind speed in the middle, to the limits of precision with this anemometer. It is not possible the wind speed is zero 2 mm from the wall, or 1 mm from it. Somewhere at a fraction of a millimeter, the speed is down to zero. That's basic physics. I wouldn't know where, but you can see from the calibrations that the error is not significant.

    The Reynolds number is 18,000, which is far about the textbook value of 2300 where turbulent flow begins.

  • Ah, ha. I see that I put the R-rating of the insulation into the final version of the ICCF21 paper. The Japanese supplier gave us these R-rating numbers. Similar insulation in the U.S. has similar numbers. So you don't need to go to the Home Depot website.

    I suggest you simplify the problem by pretending the acrylic does not insulate at all, and the insulating bricks under the reactors are perfect insulation. You can estimate losses with this information. You will see that THH is wrong. THH could do this estimate, but he won't.

  • Is the reaction stimulated/triggered by high voltage electric discharges ?

    In Fig. 1 of

    Excess Heat from Palladium Deposited on Nickel

    appears a "central discharge electrode"

    That is one of the old reactors. The Pd was deposited on the Ni with glow discharge during the test. That method is no longer used. With the newer reactor, the Pd is deposited by rubbing, before the test. That takes much less time. Weeks or months less time.

  • Regarding Kyukyutto detergent, I doubt it has any effect on the experiment. But, just in case it does, you should use something similar. The list of ingredients I translated came from here:


    The fourth item down in the box: 成分, which Google translates "component."

    Surfactant (37%, higher alcohol type (anion), sodium dialkyl sulfosuccinate), stabilizer, disinfectant

    More information can be found at the link at the bottom of item 4, "> Component information" This table appears:

    water Process agent
    Polyoxyalkylene alkyl ether sulfate sodium ester Surfactant
    Sodium toluene sulfonate Stabilizer
    Butyl carbitol Stabilizer
    Sodium dialkyl sulfosuccinate Surfactant
    Alkyl hydroxy sulfobetaine Surfactant
    Alkyl glycoside Surfactant
    Alkyl glyceryl ether Surfactant
    Alkylamine oxide Surfactant
    Magnesium chloride Stabilizer
    Propylene glycol Stabilizer
    Perfume Perfume
    Zinc sulfate Disinfectant
    Sulfuric acid soda Stabilizer

    Here is a cute advertisement for the stuff, with an animated puppet. A lot better than Rossi's. She says it cleans plastic ". . . ikki ni, HA!" meaning "at a stroke, HA!!" (like a karate chop).

    External Content www.youtube.com
    Content embedded from external sources will not be displayed without your consent.
    Through the activation of external content, you agree that personal data may be transferred to third party platforms. We have provided more information on this in our privacy policy.

  • Mizuno's results show hundreds of extra watts coming out of the reactor. One thermocouple (or several) would certainly show a significant teperature difference (tens of degrees) between a dummy and a loaded reactor. We are already having long discussions about calorimetry rights and wrongs... The setup I have suggested would confirm anomalous heat without any doubt, if the kind of COP Misuno has achieved is replicated.

    We only need a "dummy" i.e. control gas in the reactor. Because of the design of the experiment, the heat has to come out somewhere from the core heater or the reaction itself. If the reactor is first loaded with say helium or nitrogen and then run, and the temperature at the reactor cylinder thermocouple(s) and the Delta-T between the input and the output airflow is measured and recorded; and then the reactor is loaded with the D2 gas to near optimal parameters and the core heater is turned on to the same power setting; if the temperature is significantly higher for both the output airflow and the cylinder thermocouples, that is proof positive for excess heat from the rig. Simple -- no extra swapping of components, just hook up the inert gas cylinder first, run the control run; and then pump it out, load it with the D2 gas for the active run, and take the data. This would end any doubt that LENR works.

    (The above assumes as per Jed's discussion that the D2 valve is turned off when running the active experiment, and that the control gas valve is turned off when running the control experiment, and that there is no leakage of D2 gas from the pressure side of the valve during the active test that could by some weird coincidence oxidize the D2. I think that those items are easy to rule out for Mizuno or any replicator using for example pressure gage in the reactor and the pressure gage on the D2 or control gas rig. Simple alternative arrangements (i.e. two valves, one on the tank side and one on the rig side, with a pressure gage in between) can rule out any significant D2 being introduced.

    In short, with those assumptions, it seems to this anonymous observer that we have proof positive coming after a successful replication. Congratulations to the experimenters and the entire community conditional on the successful replication. It's almost too good to believe is true. Cheers!

Subscribe to our newsletter

It's sent once a month, you can unsubscribe at anytime!

View archive of previous newsletters

* indicates required

Your email address will be used to send you email newsletters only. See our Privacy Policy for more information.

Our Partners

Want To Advertise or Sponsor Us?
CLICK HERE to contact us.