More on spectacular claims of Andrea Rossi
Ludwik Kowalski; 3/12/2011
Department of Mathematical Sciences
Montclair State University, Montclair, NJ, USA
1) The reactor invented by Andrea Rossi, about which I wrote a year ago (see unit 388 at this website), is again in the center of international
attention, as one can see by going to
The recent public demonstration, at the Bologna University, could have been more effective than it was, without revealing the nature of the secret catalyst. Rossi could have provided the blueprint of the apparatus to a trusted authority, for example, Italian government laboratory, asking them to manufacture his simple device. They would bring it to the University of Bologna and allowed Rossi to place the secret fuel (nickel powder mixed with something else) into the cylinder. He would not be allowed to do anything else to the apparatus. That would eliminate a possible suspicion of a hidden energy source, somewhere within the apparatus.
This, however, would not eliminate another possible suspicion--that some chemical fuel was mixed with nickel. But suppose the powder supplied by Rossi is weighted, both before and after the experiment. Suppose the change in weight is negligible, in comparison with what it would be expected for the most effective chemical fuel. That would be a sufficient indicator that a non-chemical process was responsible for the released thermal energy.
2) Unit 388 reminded me that in the patent application Rossi claimed that "a practical embodiment of the inventive apparatus, installed on October 16, 2007, is at present perfectly operating 24 hours per day, and provides an amount of heat sufficient to heat [his small] factory. That was nearly four years ago. Suppose the rate of heating was 50 kW; suppose the heating time was 4 months per year (totaling 4*3=12 months, or 8640 hours). How much thermal energy was generated? The answer is
50*8640=4.32 * 10^5 kWh, or 4.32 * 10^5 * 3.6*10^6 = 1.555* 10^12 J
This is the same as from burning 5.29 tons of coal.
3) Was the thermal energy, reported by Rossi, released from a nuclear reaction? This would produce a measurable amount of reaction products. Addressing this issue, I posted the following message on the website for CMNS researchers, on January 24, 2011.
Andrea Rossi and Sergion Focardi, have a reactor capable of producing 12,400 watts of heat power from an electrical input of just 400 watts." In other words, the net heat generation rate is 12,400-400=12,000 W =12,000 J/sec = 7.7*10^16 MeV/sec
Suppose a typical fusion reaction contributes only 2 MeV of heat (escaping neutrinos, neutrons, and gamma take away the rest of the released energy). That would mean that nearly (7.7/2)*10^16, or about 4*10^16 fusion reactions is taking place each second.
Suppose that the atomic mass of a detectable combustion product is about 60 amu (1 amu=1.66*10^-27 kg).
What is the total mass of all products created during one hour of operation? The mass production rate is: 60*4*10^16=2.4*10^18 amu/sec=4*10^-9 kg/sec=4*10^-3 mg/sec
The mass produced in 36,000 second (10 hours) would be 36,000* 4*10^-3=144 mg. That would be detectable with a standard analytical balance.
The mass would be considerably larger for non-radioactive products (because no energy is removed from the reactor by neutrinos, neutrons and gammas).
The bottom line is obvious; detection of heat, at the rate of 12.4 kW, correlated with detection of nuclear combustion products, should be trivial, for qualified chemists.
The reactor described by Rossi (see above) operated for about 12 months (not only 10 hrs, as in my illustration). That is 864 times longer. The accumulated mass of reaction products (and their daughters, if any) would thus be 0.144*864=124 grams. For a reactor operating at the rate of ~40 kW (rather than 10 kW) the mass of the accumulated reaction products would be four times larger, or about 500 grams. In other words a sizable fraction of the nickel fuel would be consumed, producing 500 grams of new products. The isotopic composition of the spent fuel would be drastically changed. What prevents Rossi from announcing existence of such change? This can easily be done without revealing the nature of the catalyst. Both scientists and potential investors would be convinced that something totally new has indeed been invented by Rossi.
He is certainly aware of this. Here is what Rossi wrote in 'Journal of Nuclear Physics:'
Actually, this is a blog, not an on-line journal. But that is not my point.
(A) the Ni powder I utilized were pure Ni, no copper . At the end of the operations in the reactor the percentage of copper was integrally bound to the amount of energy produced. A charge which has worked for 6 months, 24 hours per day, at the end had a percentage of Cu superior to 30%
(B) About the Ni isotopes: the isotopes after the operations were substantially changed in percentage. We are preparing a campaign of analysys with a Secondary Ions Mass Spectrometer at the University of Padua (Italy), at the end of which the data will be published on the Journal Of Nuclear Physics.
Why were such spectacular results--conversion of at least 30% of nickel into copper and drastic changes in isotopic composition--were not published at a scientific conference, or in a real journal? That puzzles me. How can this be explained?
4) Here is how the most recent demonstration of the device was summarized in
'Focardi and Rossi demonstrated a novel boiler on 14 January 2011, which converted water at about 13 degr C to steam at 101 degr C. The device reportedly produced energy from nuclear reactions between nickel and hydrogen. An electrical heater in the device drew about 1000 W at startup. Later, once the reactions started and provided heat, the input power was reduced to about 400 W. Consumption of hydrogen gas was essentially negligible. The input water flow was about 150 grams each half minute. Given the measured input and output temperatures, that flow rate, and a measurement that the steam was dry, it is easy to compute that [in the steady state] the device delivered over 10 kW of thermal [power] to the water. '
The reactor, shown on available illustrations, was a small cylinder; the amount of nickel powder is probably less than two or three kilograms. Accepting this estimation, I conclude that the 30% of produced copper, reported by Rossi, translates into about 1 kg, after a prolonged operation. My very conservative estimate, 144 mg after 10 hours of operation (see point 3 above) is probably consistent with what he reported.
According to Rossi (see the link to The Journal of Nuclear Physics in point 3 above) production of copper takes place when hydrogen nuclei, which are protons, fuse with nickel nuclei. Fusion of protons with nickel nuclei has been studied by many physicists, including myself (as reported in The Physical Review, Vol .163, Nr. 4, 1074-1077, November 1967). But our protons had the energy of 14.3 MeV, Rossi's protons, by contrast, had low temperature energies, close to 0.04 eV.
The probability of nuclear fusion, expressed in terms of measurable cross sections, is known to decrease rapidly with energy. How can 0.04 eV protons fuse with nickel, whose atomic number is 28? Rossi does not have to answer this question; his role is to provide experimental evidence that such fusion does occur, under the influence of a catalyst. I am sure that a theoretical explanation will be found, soon after his experimental results are independently confirmed, and after information about the catalyst is published.
One of the well known facts is that the cross sections of a nuclear reaction does not depend on the chemical composition of the target bombarded by protons. The target can be a crystalline nickel, a compound containing nickel, a melted nickel, or vaporized nickel. No exception from this rule has been found. But the rule has not been tested for all conceivable solid compounds. As far as I know, it has never been tested for protons of very low energy. Rossi, like several other scientists, probably believes that the rule is not universal and that exceptions are possible, especially at very low energies. His experimental results, if confirmed, would validate such expectations.
The 'all swans are white' rule, I remember reading somewhere, was valid up to the time at which black swans were discovered. In science, unlike in mathematics, validations, in final analysis, are based on experimental results, not on logical proofs. On the other hand, predictions based on logic, have often been successful; in other words, they were later validated by experimental data. Disagreements between new experimental results and existing theories should not be ignored. Yes, the 'theoreticians often guide while experimentalists decide.' But theories are based on verified results from experiments and observations. The chicken and the egg dilemma? Not really. Why not? Because the process of accumulation of scientific knowledge is not circular; it is spiral, as illustrated below.
Added on May 13, 2011
The nature of the catalyst remains secret; independent replications are impossible. That is why most messages posted on the private list for CMNS researchars, have been theoretical. Here is a good summary, posted two days ago by Abd ul Rohman Lamax.
Many theoretical physicists worked on CF theory, including more than one Nobel laureate. It is obviously a very tough theoretical problem. The first problem was that far too many, who might be otherwise competent, took a shallow look, said "impossible," and then dismissed the growing body of experimental evidence as *necessarily artifact, error,* since, *impossible.*
That was obviously a systemic failure, or, as Huizenga called it, perhaps not realizing the irony, The Scientific Fiasco of the Century.
In normal problems with artifact, the research was done to demonstrate the artifact, cf. Polywater and N-rays. With both polywater and N-rays, there were *confirmations,* then demonstrations that the observations were artifact, not merely argument from impossibility. With cold fusion, there was only replication failure, and, by definition, replication failure cannot demonstrate artifact in the original work! It can only demonstrate one of two things: failure to replicate, conditions not the same or a chaotic effect, or, secondly, total error or fraud on the part of the original experimenter, or possibly some unidentified artifact that isn't reproducible. The latter possibility declines and rapidly vanishes with successful replication.
If it's replicable -- and cold fusion was -- then any artifact should be identifiable, by controlling it in and out.
Once it was known that He-4 was being generated at the right heat ratio, and that was confirmed, as it was -- it's a reproducible and reproduced experiment, properly described -- it should have been all over. Before the close of the century.
But what is happening to cause this effect to appear? As Dr. Storms has pointed out, there are "plausible explanations," but no theory has been validated in any strong sense. Normally, what we want to see is predictive value, for theory is needed for prediction, as well as understanding. (Non-predictive theories can have a value for organizing thought, but not for engineering, and these are of some interest, but it's not really "scientific." Some non-predictive theories can usefully suggest avenues of approach, a good example might be the general wave theory of Dr. Irving Dardik.)
From my sense of the situation, any cold fusion theory is going to involve one or two "leaps," possibly more, that is, assumptions that lead to some possible understanding of a mechanism, but, because more than one are probably necessary, not to a complete understanding, adequate to make accurate predictions and answer objections. What is needed in this situation, my sense, is far more experimental discovery of the effect of controlled conditions.
My sense is that this is precisely what Rossi did. My own idea of approach would be to move in almost the opposite direction from much of the research, to design experiments that are smaller and cheaper and simpler, for thus it would become easier to vary the conditions, and to run the same experiment many times, to determine how much variation is still uncontrolled. Thus the effort to make it "bigger" and thus "better," worked against what is needed for understanding, particularly when funding was limited.
Rossi also apparently had a lot of his own money to put into this.
(In all this, I'm assuming that Rossi's heat is really being generated and is not fraudulently simulated. "Artifact," i.e., simple error in measurement or interpretation, seems highly unlikely at this time. Because of how much is at stake, "fraud" should not be casually discarded as a possibility, but I hasten to add that this is not an accusation of any kind, it is simple prudence. If there are hundreds of millions of Euros at stake, or more, what would be preposterous in a small scientific matter becomes not-so-inconceivable. Only when we have truly and fully independent replication -- which should happen this year if the effect is real and they are not blowing smoke, will we be able to leave "fraud" completely behind. And fraud may still take place in other ways! I predict there will be people who will claim other "discovery" who won't have actually done it, and further attempts to fleece unwary investors, even if Rossi and Defkalion are completely honest.)
So, right now, if anyone wants to pursue the Rossi track, and I know that there are people doing it, I suggest Smaller is Better. Make the tests as small as you can, consistent with showing some clearly significant effect. Depending on your calorimetry, 105% might be enough, it certainly was enough to show cold fusion with skilled calorimetry. Then keep varying it, a series of experiments that change only one variable, the more the better within one series. All the data is valuable, there are no "failures" in an approach like this.
As data like this becomes available, from those who choose to publish it, finding adequate theories will become easier. It will still be one of the most difficult puzzles in theoretical physics to come along for a long time, I predict. There are probably two miracles involved: how fusion or other LENR happens in the first place, and then how is the energy converted into heat, what happens to the expected gammas, etc.? Maybe someone will get lucky and propose a mechanism that, all by itself, causes the great forehead slap, and numbers will fall out. But I'm not counting on it.
Added on May 14, 2011
Intense theoretical discussion, involving several people, has been going on at the private Internet website for CMNS researchers. Let me show two interesting generalizations formulated by Ed Storms and John Fisher. They are addressed to those who speculate about mechanisms of nuclear processes responsible for the reported experimental data.
In one of his messages, Ed Storms wrote: We have two entirely different kinds of reactions to explain; fusion and transmutation. The questions is, "Is one or more than one mechanism in play"? Regardless of how many mechanisms are operating, the same consequence is produced, i.e. the barrier is overcome by a process other than brute force and the energy is released in small increments, generally too small to detect.
I think we have good evidence that fusion between two D takes place and produces the expected 4He. Tritium is a problem because it can not result simply from d-d fusion and produce the observed consequences. Tritium has been produced when deuterium is present but also in H2O. The evidence suggests both d and p must be present. In both cases, radiation is seldom detected. So, we need a mechanism that addresses these conditions and consequences.
Transmutation has been shown to involve clusters as well as single d or p. In both cases, radioactive isotopes are very seldom produced. In addition, radiation is seldom produced. The rate of the reaction does not seem to be sensitive to the charge on the target nuclei. This means the mechanism hides the charge on the p or d, even to the extent when a cluster contains 6d, but does not affect the charge on the target. When transmutation occurs in Pd+D, the Pd atoms forming the lattice are not the main target. The situation in the hydrogen system, according to Rossi, is different. Does this mean that several mechanisms are operating, which are sensitive to whether d or p is used?
I suggest the first approach is to see if a mechanism can be found that can produce all the observations without making ad hoc assumptions. None of the propose mechanisms can do this although some make that false claim. The erzion and polyneutron ideas do not fit the bill on many levels even though they are clever and address limited observations very well. The deflated electron idea of Horace, although possible as a basic concept, is much too complicated and filled with arbitrary assumptions and vocabulary to be plausible. The BEC ideas have no observational support and have limited application. If a BEC were found to form in a lattice at RT, this observation alone would warrant a serious prize, regardless of CF. I have addressed the W/L theory before and shown that it is not worth discussing. The idea suggested by Andrew is getting close to what is required, but it also needs to be explained better. So, I hope people who have made theoretical suggestions will take these comments seriously and try to fix the problems so that we can get something to guide experiment.
In one of his messages, John Fisher wrote: The comment by Ed Storms regarding X-rays associated with LENR reminded me again of the beauty and utility of the experiments done by Iwamura and his associates. Their reactor is conceptually simple: a specially prepared palladium foil with deuterium gas on one side and vacuum on the other. The Pd foil is special by having very thin layers of CaO interleaved with Pd layers near the surface facing the gas. With this reactor they have performed a number of critical experiments. They observed excess heat and X-ray emissions when using Pd foils containing CaO, but not when using plain Pd foils. They observed transmutation of target elements deposited on the deuterium-facing surfaces of foils, but there was no transmutation when H was substituted for D or when MgO was substituted for CaO. These experiments raise clear-cut questions for theory:
(1) How does theory account for the X-rays?
(2) Why is there excess heat with D gas but not with H gas?
(3) What is the role of CaO in this reactor?
(4) Why doesn't MgO support nuclear reaction like CaO?
(5) How does theory explain the observed transmutations?
We theoreticians owe a substantial debt of gratitude to the Iwamura team for raising these challenging questions against which to test our theories.