Parkhomov Explains Cut-and-Paste of Plot Lines in March Experiment [Update: Raw Date Published]

Thanks to Pelgrim for finding this:

There has been a lot of discussion regarding a repeated pattern found in the temperature traces in Alexander Parkhomov’s most recent report which indicated that certain parts of it had been cut and pasted.

Today on Facebook a reader named Stefano Marcellini reported a communication he has had with Parkhomov about this matter. He writes:

Since nobody dared to ask Parkhomov about his cut-and-paste plot, I did it. Here is his answer in attachment. I did not include the excel file with all the data, as it is a huge list of numbers, but they correspond to the ones that he provided, and that you can see in the attached file. I don’t want to make any personal comment. Read it and make your own judgment.

UPDATE: Raw date is available at this link:

The attached file can be read here:

In the document, Parkhomov admits the cut-and-pasting, and explains the reason:

Dear Stefano Marcellini, I admire your observation and I repent of the sin.
The matter is that the laptop on which there was a record of temperature, worked steadily only when was disconnected from the power supply network and was powered from the accumulator [I assume he means battery]. Therefore sometimes it was necessary to interrupt record for recharge of the accumulator. It occurred at temperatures about 460, 1020, 1120, 1160 and 1200 oC. At this time temperature was recorded on the paper recorder and measured by pointer indicator. These devices showed the values of temperature close to the specified. That the plot looked beautiful and I didn’t cause the questions distracting from the main point, such peculiar interpolations were made. It, of course a great sin and I sincerely repent. However it doesn’t influence results of research in any way. I assure you that in results of measurement of power consumption and pressure of any shifts it wasn’t made.
I send you the Excel file with the data obtained during experiment on which pauses in registration of temperature are designated by admissions of rows.
Once again I admire your sharp observation and high professionalism. I hope that this incident won’t make the attitude towards me and my researches hostile. Alexander Parkhomov

So this seems to explain the affair. Parkhomov admits doing the cut-and-paste, explaining the reason was the laptop not working correctly when it was unplugged from the power supply network. He admits his mistake, which seems to have been done only by reason of necessity from his perspective, and seems to be chastened by the affair. We don’t see the raw data, but Stefano Marcellini says they correspond to what Parkhomov reported.

This case is a good example, I think, of peer review by the crowd, which seems to have worked very well here.

  • Rui Germano

    I suspect this will affect his reputation more than he wished for.
    The sceptics will have a field day.

    • yes, but it can be corrected by a good experiment

      • Rui Germano

        Let’s hope he does it. It also doesn’t help that nobody was able to replicate his results thus far. So I think you see where I’m (fearfully) going with this.

    • Freethinker

      The pathoskeptics have a field day every day, as they don’t care what is right or wrong, true or false. They need no excuses. Parkhomov should not have done what he did, but the explanation, retraction and correction does cover it all nicely.

      Somehow I doubt he will do it again.

  • Sanjeev

    As expected, it was a “beautification”. It is still wrong, but I’m glad that he admitted it. Now the replications can proceed with more confidence.
    He could have avoided the whole thing by simply mentioning in the original report, that there is a manual “interpolation” of missing data at such and such points. It is ok to interpolate, not ok to not to mention it.

    • Ged

      Wish the raw data would be passed on to us, no worry about “beautification” shenanigans then.

      • Sanjeev

        It is released, it will appear out of the internet sooner or later. Perhaps someone can email Stefano Marcellini and get the file.

  • MWerner

    My father had similar trouble with his laptop, GPS,and autopilot on his boat. We eventually disconnected the grounding lead of the power supply and solved the issue.

  • Andre Blum

    This is not good. Not good at all. We don’t want beautification. This severely alters my opinion about the value of this replication.

    • mcloki

      A “do over” is in order to get all monitoring equipment in working order. I’ll give him the benefit of the doubt, He’s not only experimenting on the reactor but with the monitoring equipment. Consider this a foul ball, dust yourself off, step back into the batter’s box and take another swing.

  • fritz194

    Its quite awesome that we programmed the first computer using handwritten notices.
    There was science without word, excel, matlab, computers.
    Real people, real paper, pencil, instruments and a watch.
    This modifications were done to satisfy a generation which trusts digital domain only.
    Would you believe Mr. Parkhomov if he presents you a Xeroxed handwritten table with data ?
    No. We believe into this excle graphs – because they are so easy to manipulate.

  • Curbina

    This is the kind of stuff that happens and only people that has ever carried experimental work with very limited resources really understands. This is certainly to cause the wrath of the pseudoskeptics to rage, Dr. Parkhomov should not have done the copy paste, but he already did, and now we have to live with that.

    • I hope the colleague around Moscow who proposed their help during ICCF19, will manage better protocol in their labs.

    • Fisher

      Perhaps it will cause “pseudoskeptics” to rage as it will certainly cause the e-cat/parkhomov minions to rage in the other direction. I am neither of those and yet this news does cause me to raise an eyebrow of concern in question of the doctor’s claims. I’m no atheist but I do exercise good critical thinking skills and there is a lot at stake here. I’ve been following along here since 2011 and, no matter how much I want all of this to be true and the world to be saved, I will not become a mindless LENR zombie wildly flailing away at any who dare question this still inadequately proven “reality”. People would be correct to be cautious and even skeptical with these findings under the current circumstances. If it’s all true and good then good facts will come to light and we will all rejoice. If the oposite turns out to be true then those who have turned away from emotional zealotry will not be harmed be a harsh reality check. We should not blame the skeptics nor silence them but rather we should continue asking for the data.

      • Curbina

        I’m fully aware that’s there’s no definitive proof as yet, no matter how much we would wish for it, and therefore keeping my expectances in check. On the other hand, other people wishes all the LENR field to be burnt in a great pyre. We should always be wary of both extremes.

  • Patrick Ellul

    Data has been published as a spreadsheet here

  • Bob Matulis

    Can anyone confirm the repeating patterns occurring at the stated temps of “460, 1020, 1120, 1160 and 1200 oC”? If so it will build credibility in his explanation. Undoubtedly his interpolation would have been acceptable only if the raw data (with gaps was provided) and the chart was footnoted with an explanation of the interpolation. I have done this type of interpolation with daily weather data to fill in the gaps on my excel (otherwise formulas would blow up)

    • Ged

      Yeah, Obvious did that analysis and was the first to point it out. He made versions of the graphs that clearly show where this happened. Should all be back in the Parkhomov Moscow presentation thread.

  • Teemu Soilamo


    And still no evidence Parkhomov is not a charlatan. Back to square one, it is.

    • Ged

      Raw data is up so all is moot.

    • Omega Z

      It would have looked much better had AP reported this up front & said that the raw data would follow.

  • Ophelia Rump

    God, someone rich buy him a decent piece of equipment, someone fund him. You fund a man like him, you do not stand around tisk tisking at the quality of his equipment. How god awful sad this is to see. We do not deserve to survive as a species.

    • Sanjeev

      As we know, he has refused all donations and has said that he has some funding now.

  • Sanjeev


  • Matt Sevrens

    Finally someone has the balls to actually ask him

    • artefact

      MFMP asked three times.

      • Bob Greenyer

        Yes, once within 1 hour of the discovery – he answered other questions in same emails. He was obviously mortified.

  • Curbina

    Now, looking at the data, I see that the °C data has only one gap, at the end of the file, not at the intervals that were copy pasted.I am left in Whysky Tango Foxtrot mode.

    • Daniel Maris

      It’s a reasonable expectation but it makes a further replication by another team all the more pressing.

    • Sanjeev

      Look for breaks/jumps in time.

      • Curbina

        Duh! thanks!

      • Gerard McEk

        I admire that AGP bravely and deeply admitted his sins. What I do not understand is why he did not say this when he published the data. Everybody would have accepted that and may have helped him to avoid that problem next time.
        Honesty shall be the prime focus of a scientist because if dishonest acts are done, then trust is compromized. This is also why this community has been fulminating against the treatment of Pons and Fleischmann which was an act of dishonesty in the n-th degree and without any apology!

        • bachcole

          To confess and repent of such a dreadful thing takes a lot of heart and inner strength and goodness, and I admire him for it. I still believe him, but for most people, his duplication doesn’t mean squat until such time as someone duplicates him or Rossi. He is essentially on probation with almost everyone until such time as we get another duplication.

  • Obvious

    I am glad that Dr. Parkhomov has admitted to, and explained the graph alterations. The reasons are essentially as I had suspected. I greatly respect his admission.

    • Bob Matulis

      BTW, hat tip to all of you who identified the doctored data and exposed it in a respectful way. This builds the credibility of the Ecat community in my opinion.

    • Andreas Moraitis

      Absolutely. Everybody can make a mistake and deserves a second chance if the mistake is frankly admitted. We should also not forget that the graph had originally been created for a PPT presentation at a conference. In such a case it might be tolerable to ‘simplify’ a diagram as long as the overall information content is not affected.

      However, it is important to find out if the latter is the case (especially regarding the HAD phase). Anyway, a successful replication is urgently needed.

  • pelgrim108

    I am credited with finding it, but in fact it was Ecco who brought it up first on The Always Open Thread.

  • Why isn’t he building a reactor and driving to one of the many universities in Russia and is conducting his experiment there, with high tech meassurement equipment? I guess a lot universities would offer him a lab for a few days.

    Why isn’t he doing that?

    • Bob Greenyer

      He likes working from home, he has everything there, things he has built and trusts. I left him with the new, good multimeter and scales I brought to verify things, these were paid for by donations.

      He has a wife and dog, by working from home, he can work on his own schedule and be around the things he loves.

      • I can take your points, but I guess he has friends and colleagues at some of the universities who would support him, and monitor the experiment the whole time so he could go home and do other things.

        Such a step could simultaneously be a big step forward to another independent replication. If these guys at the university measure positive results, they will want to know where this energy is coming from, in a scientific way.

    • GreenWin

      Great news Cmndr. NASA has been given many opportunities to come clean, get honest, make amends. Lacking those moral and ethical choices, the program will again be working on its own. We wish them good luck.

  • Obvious

    Looking at the raw data, one can see pulses well into the 600 W range within the low wattage values in the 1200°C heat zone. He seems to have an average Watt value, possibly averaged over 10 readings, that was used for the plot.
    Maybe this is SSM-like, or it could be aliasing. But at least the real values are there to look at and consider. I was very interested in those W steps, and how they relate to temperature at the RH side of the graph.

    • Obvious

      There is a weird formula in the data that does something to the date(?) and gives a wattage figure(?) (Bt10 column). The final W for the plot is derived from averaging the results of this formula.
      I can’t duplicate the formula by typing it out (bad results, nothing but ######), but it copy-pastes fine and gives the same results.
      I’m fiddling around the 9001 row area.
      Anyone have any idea what is going on?

      • Obvious

        With Bt plotted, not time. Oops, cut off by line freeze. See if I can do this in one piece…

        • Obvious


          • Obvious

            With averaging raw Bt (Watts) over 11 points, 5 on either side of time stamp and at the time stamp. Looks not so bad, and the steps are gone.

          • Bob Matulis

            Obvious, now that you have processed the data does interpolation by Parkhomov seem to be a plausible explanation? I am hoping so.

          • Obvious

            It looks pretty reasonable, minus the lack of actual temperature data for a huge stretch of time. I think the steps might come back, too, now that I understand what the meter was doing, compared to the spreadsheet.
            I am re-processing the W data again and see if the steps are really there. The Watt data is tricky, since it gets reported every 11250 Joules, whatever the time it takes from the last increment of 11250 J, rather than by using VA. The averaging method will really influence the shape of the Power trace. Mine is probably technically incorrectly displayed (above).

          • Obvious

            Taking a Watts (Bt) calculation, and the 9 prior Watts calculations (starting 10 time periods down), averaging them, and plotting only the average of these ten-Watts calculation averages using the last time sample period in the group of ten… The steps mostly go away, and the line looks like a line…
            This gets rid of the problem of using future values (IE: 5 time stamps after the plot point) being sequentially incorporated into the average for each time point, which seems to cause the steps.
            This is my best stab at this, if anyone wants do better, go for it. I am interested in any other solutions to the Watts distribution.

          • Bob Greenyer

            That looks more like it – good work Obvious.

          • Obvious

            The mystery of how the controller “read the future” in order to reduce power in steps and maintain temperature is solved. The original plot actually did read the future for data points.
            I have also extended the line from the end of the 1200 area with the below addition. The spaces need to be there, and the spreadsheet chart adjusted to connect gaps with a line (setting). (I used 1195 only so the line was visible).

          • Obvious

            I made this also.

          • Rheulan

            Exactly my
            thoughts. Today, Axil posted this same link you provided at vortex-l,
            and it made me think the same as you did. A second generation EM drive coupled
            with an e-cat is the way to go.

          • Obvious

            Amazing. Using the harmonic mean of the Watts per second over 10 sample periods prior, and plotting these at the last time period, the plot looks nearly the same as the original, including the steps. The Watt values are slightly lower, especially at the high end, due to longer periods of low power than high. So I have to say that Dr. Parkhomov’s original plot is both representative and conservative, using the data provided. His method of averaging is effective.
            Also, using a calibrated heat source, and reflecting the heat back with an aluminum block 4 mm thick, 2 mm form the source, I could not exceed the original heat by more than 15%, let alone three times. So the heat shield theory (see photos in his power point) cannot work to increase the heat to the degree seen in the data. Whatever he measured, even considering the blank areas for temperature, is very hard to explain by mundane effects.

          • Obvious

            And so…

          • Obvious

            Original again for easy comparison.

          • Jimr

            I am sure NASA,s announcement is important however those familiar with the TR-3B may find that it is not that significant.

          • Obvious

            Here is the original plot for easy comparison. If nobody replies to this post, it shouldn’t get shifted out of sight.

          • Bob Greenyer

            These things will not matter, or matter a lot, depending on the outcome of replication attempts.

      • Bob Greenyer

        He works out his power by a fixed joule pulse from the domestic power meter divided by time – gives watts

        • Obvious

          There is something I’m not quite getting. It makes sense to work out the time period, and average out the power over the period in a weighted manner (which is not what I did three plots down below) but I am having trouble seeing how that translates into Watts. The formula doesn’t seem to poll Watts directly, yet the values still come pretty close. I am still fiddling with it. The steps might come back, yet. It’s like there is a piece of formula hidden from view somewhere.

          • Bob Greenyer

            You may be getting closer.

          • Obvious

            W [Bt] = 11250J / (Time Interval between 11250 increments in seconds)*86400. Time stamp is the last of the two sequential time stamps, from which the prior time stamp is subtracted to arrive at Time Interval.

            However Bt10 = 11250J / Time Interval between 11250 increments using five increments in future minus five increments in past, in seconds *86400, using time stamp in center of group of ten

            And Сред [Average] = average of five Bt10 values, from two previous to two in future, inclusive, using time stamp in center of the group of five (which is a smoothed group of ten, see above).
            The last two operations seem to make the steps occur.

        • Obvious

          This is a bit strange, but why make up such a weird story? Sometimes the truth is stranger than fiction… I was thinking that the USB port goes dead if the laptop goes to sleep or hibernates.
          Many laptops will not supply power to the power sharing USB port (power available when asleep) unless the battery is above 50%, which will also cause problems where the USB port is allowed to keep the computer on (or wake it up). A momentary attempt by the laptop to go to sleep, with the battery below 50%, could easily result in the “always hot” USB port power being cut off, and then the “allow to wake by USB activity” function will not occur to cancel the sleep command. This shouldn’t occur plugged in, but often an additional BIOS command is needed to override the sleep cancelling functions in addition to unchecking the setting boxes that by default force the USB ports to go dead. If the laptop goes into hibernation instead of sleep, then these functions will not work. So his laptop may be set to hibernate after a certain period plugged in with no keyboard or mouse input, but go to sleep after a certain period when in battery mode (or the other way around?).
          Just a wild guess… I had to fiddle with these settings once to keep data rolling in when unattended, a few years ago on company laptop. Lucky one of the other guys knew what to do, and I only vaguely remember the settings, since I fooled around with them for quite a while before going for help.

      • Josh G

        When numbers are too long to display in a cell, Excel displays them as ###########. The value is there but it is just displayed like that. If you want to see the number, just widen the column or change the cell format to ‘scientific’ and it will use exponents.

        • Obvious

          I actually got an error message that said when a Date is negative(?), it gives that ###, (or something to that effect). I first expanded the column really wide, and it wouldn’t stop making more #. I changed it to Number, after seeing the error message, and it was fine then.

          • Josh G

            Ah, I see. So it was defined as a date. Well anyway, great work!!!

          • Obvious

            Thanks. I think I got the negative date while fiddling with the formula, trying to work out what it was doing, how the heck it got Watts out of time… Then Bob G got me going on the right track with the meter.

          • Alan DeAngelis

            Or use the 8.6 MeV alphas that are produced in the Hot-Cat as the propellant.

  • Bob Greenyer

    Alan Goldwater is back in California and following some discussion, here is the plan for the coming weeks.

    1. Reheat the fueled reactor to the higher temperature used in Padova. Compare pressure and temperature data where possible. After cooling, extract the fuel for analysis.

    2. New reactor with type B TC in core. Run calibrations to 1250 C outside core (1350+ inside). Then load with LiH only and run with Parkhomov temperature profile. Compare pressure data to that seen in Padua. Inspect filler rods and core for residue. This will give an understanding of if the reactor is leaking or defusing Hydrogen through its structure or if the Hydrogen really was being absorbed in the Padua Nickel, or a combination of both. Probably run this 2nd week of May.

    3. New reactor with fuel TBD and Parkhomov temperature profile. If we can get the legally approved LiAlH4 shipping technique sorted, this will be full Parkhomov fuel. Planned to run week of May 18.

    • EEStorFanFibb

      “If we can get the legally approved LiAlH4 shipping technique sorted”

      thank you. you can with help from a university I would think. good luck with everything. Cheers.

      • Bob Greenyer

        University of Missouri / SKINNR is working with us to establish that and they will ship us all the materials etc to do it right.

        • Omega Z

          Glad to hear about the SKINNR collaboration.
          By any chance did you meet or have any discussions with Robert Duncan?

          • Bob Greenyer

            Yes – at length, mostly about material transfer agreements for P&F palladium – of course he is not at SKINNR.

            I like Rob Duncan, a lot.

    • radvar

      So, uh, er, am I correct in perceiving that there seems to be a focus on demonstrating transmutation (via the ash) as much as on demonstrating excess energy?

      Not saying that’s a bad thing; just didn’t notice when the addition of focus occurred, if there was one. Transmutation by itself would be a paradigm shaker of the first order, and would seem likely to intensify pursuit of excess energy.

      And, except for the equipment requirements, perhaps would be easier to measure, in terms of before and after comparisons. No squirrely caliometry to debate.

      • Bob Greenyer

        If an experiment looks like it has produced excess heat – then get the ash (cake) compared to fuel, if the isotope/element shifts exist as claimed and the path to them is net energy yield then there is a strong argument that excess heat was seen and it was of nuclear origin.

    • Private Citizen

      Good steps all.

      Am wondering, if Parkhomov so openly supplies the fuel, why doesn’t he just lend an entire working reactor and dispel doubts about accuracy of duplication?

      • Bob Greenyer

        Risk, legal complications – plus, there is no guarantee the recipient would operate it correctly.

        • Private Citizen

          Such a shame that legal complications could pose a huge impediment to vitally important science. Heck, sign a waiver of liability and test the thing on a boat in international waters.

  • Bob Greenyer

    Everything that he has published to date that could be verified has and checked out. This didn’t check out – and now we know why. We will have to wait for a promising Parkhomov fuelled run from one of the replicators to see if his isotopic / elemental shift/analysis bares up to proper scientific scrutiny.

    The truth will out whatever, regardless of peoples conjecture or opinions, we are talking about nature.

  • Bob Greenyer

    It is 0.7g of fuel, it is claimed that it effectively produced 100Ws for a long period of time.

    Of course – longer runs are better.

  • bachcole

    There is a limiting calculation that can be done with say hydrogen and fluorine, which I assume has the maximum potential energy for any chemical reaction. If the alleged LENR reactor is putting out orders of magnitude more than the same amount of a hydrogen-fluorine reaction, then the reactor is no longer “alleged”, it is now an actual and verified LENR.

  • Ged

    Fair enough (well, depending on the program–A lot of them spit out comma delineated files that are supported by Excel and viewed as a “spreadsheet” and then save as xlsx, thus the spreadsheet is indeed the raw data in such cases; I’ve seen and used a variety of program scripts that did this very thing on raw data collection). But even an excel spreadsheet is vastly superior to a graph.

  • Bill

    Was the waveform in Padua identical to the waveform in Russia? If there is a lattice in the substance of the experiment, then lattice mean eigenfrequencies. Maybe, if the waveform was different, in Russia there were more harmonics in high frequencies with sufficient amplitude .

    • Bob Greenyer

      More likely we had too little dead volume and not enough time because we had not ground the LiAlH4 from 100s um to 5-20um

  • GreenWin

    Simulations of life are just that – attempts to artificially “simulate” that which programmers are incapable of. Why do they usually fall short? Because artifice will never replace life. This particular sim, IMO, struts and frets like a poor programmer! 🙂

  • Gerrit

    I think his explanation makes sense and I am happy that AP responded to clarify the issue.

    What I don’t understand is why the laptop can’t take measurements correctly when connected to the power supply. What happens exactly when it is connected to the power supply ? Readings go up, or down, or all over the place ? What does that tell us about the confidence of the readings when there is no connection to the power supply ?

    Let’s see what the upcoming batch of replications will show us. When some of the other experiments will show similar readings and the ashes indicate nuclear processes took place, we will have what we are after.

  • Bob Greenyer

    It is QuantumHeat CICs next project after The New Fire, we already recorded a major part of the Kickstarter pitch last year with a leading inventor of technology that has been talking about this kind of effect for a very long time.

    We made the recording in a planetarium!

  • Chris, Italy

    It is unfortunate that Parkhomov fell into such a silly blunder instead of overcoming the problem in a better way. His apologies are welcome.

    The main thing though that folks like Marcellini should get straight is that Parkhomov keeps no secret and does not stand in the way of replication. Marcellini is a researcher; although perhaps he can’t be bothered to, he certainly doesn’t lack the resources for attempting replication (at least with colleagues). His comment on that same post, asking Rossi, Parkhomov etc. to please present their results better (if they truly have discovered something great) somewhat neglects the fact that Parkhomov doesn’t prevent replication. After having had a quick look through the last few posta on that group, I’m not so sure if I’ll join it to say this there.