*A dependent, 2nd party, untestable claim means squat*
This paper is yet another unrefereed, sub-par cold fusion claim to add to the pile of unrefereed sub-par cold fusion claims. Only this is is an unrefereed, sub-par cold fusion claim made with a black box that no one else has access to. Pitiful. I'm amazed that this has caused such excitement among true believers. It will be amusing if this all comes crashing down like the wet steam claims did. The 2011 demos got Rothwell and Storms and others to say Rossi was the be-all and end-all in cold fusion, but a year or so later, Rothwell was back to citing McKubre's 1994 paper as the best evidence for cold fusion. It all just shows the incredible bias among true believers for any kind of a scrap of good news that they can rally around to feed their euphoria a little longer. And Storms' support for the 2011 demos served to show him as gullible as the rest of the true believers. What has changed with this new report? Before we had Levi, Essen, Kullander, and Focardi writing unrefereed reports published on-line claiming that Rossi has demonstrated a source of energy beyond chemical, and Pettersson giving his verbal support, and Levi reporting a secret experiment that verified the ecat without steam. Now we have Levi, Essen, Pettersson, and a couple of new Swedish professors (Uppsala's shame!) writing an unrefereed report of a secret experiment published on-line claiming that Rossi has demonstrated a source of energy beyond chemical. Some believers argue that this is what skeptics asked for, but it's not even close. In early April, I said that it was not at all unlikely that allegedly independent 3rd parties would claim clear evidence of LENR, but then wrote: " *Allegedly independent 3rd parties could simply be some little-known academics. But we already know from the BLP history that statements from such academics mean squat. Especially if the academics appear to be recruited […] [E]ven if the statement is definite, a scientific revolution will need more to go on than trust in a few recruited academics. Individual academics, […] with unremarkable reputations, have essentially nothing to lose by being wrong. In fact, if Rossi pays consulting fees, they may have a lot to gain. [It's not, as some have said, P&F again, because no one is paying attention, and so no one will notice if they get it wrong.]* ** *Skeptics have been asking for independent validation, but reports about a black box from a few (or even a dozen) individuals is not that.* ** *For exactly the same reason, it is unlikely a prominent journal will publish claims of a new phenomenon that can't be tested by its audience. So, if a report is published, it will be in a small-time journal, and no matter how detailed, if it's a black-box test, it still relies on trust.* ** *So what is needed? The best would be if the details of the reactor were disclosed so others could test the claims, but of course that won't happen. Next best would be if the reactor were made available on request, so it could be tested by anyone. Also won't happen. Third best would be a public demonstration involving some experts who are on the record as being skeptical, and which is really transparent. With an energy density a million times higher than dynamite, this should not be difficult. A completely and obviously isolated device that produced unequivocal heat many times its total weight in chemical fuel would do the trick (and should be trivial).* ** *Failing those, the only sort of validation that would have an impact is one in which the independence is truly transparent, and the consequences of being wrong are significant. This would be the case, for example, if Rossi had an open invitation to major national laboratories to run the test, and one or more openly and voluntarily sent a team for the purpose. Any labs that offer to perform the test and are turned down would be in a position to report that publicly. National labs or even university endorsed teams have a lot more to lose than a few academics acting on their own. This is less than perfect, but it would certainly draw a lot of attention. But I'm pretty sure that's not gonna happen either. Alas.* ** *Whatever happens in the next month or two, it will almost certainly have no significant impact on mainstream science, and at most a ripple on the mainstream media. A year from now, LENR and the ecat will be in exactly the same place they are now, which is exactly where they were a year ago, and where they were 2 years ago. There will be an eager internet following expecting something big real soon now, but the rest of the world will remain more or less oblivious.* " So, what has happened falls short of every criteria I considered that would make it significant, and met precisely those criteria I said beforehand would fall short. The authors are little-known academics, acting without the backing of their institutions, and they certainly appear to have been hand-picked, 3 of them being previous supporters of obviously flawed demos. They published in an unrefereed journal, and of course, it's a black-box test. We are asked to accept a revolution in physics based on trust of these few people. Trust that they have not been fooled by Rossi, that they are not complicit in deception, and that they are competent to validate the technology. We have already seen that Levi, Essen, and Pettersson are not competent, and the present paper suggests none of the authors are. If you can think of a similar revolution that was widely accepted based on this sort of charade, please let me know what it was. *The paper* As for the paper itself, as already mentioned, it's a black box test performed by selected scientists in secret. We are asked to accept a revolution in physics based on trust of these few people. Trust that they have not been fooled by Rossi, that they are not complicit in deception, and that they are competent to validate the technology. We have already seen that Levi, Essen, and Pettersson are not competent, and the present paper suggests none of the authors are. If you can think of a similar revolution that was widely accepted based on this sort of charade, please let me know what it was. Of course, a black-box test is no impediment to proving the ecat is real, as many of us have argued. But it must be more than a secret experiment presented in a written report. A written report is enough for a completely disclosed experiment, because then *anyone* skilled in the art can check the claims independently, and of course, that's how most discoveries reach the mainstream. But no one can check this result, and that's why most prominent journals would reject it as a scientific paper. A black-box experiment must either be openly performed with obvious controls, or must be available to any qualified team for testing. From that point of view, the 2011 demos were better than this, because it was at least semi-open, constraining to some extent what they could get away with. The problem there was that the reported observations themselves (except Levi's secret experiment) did not support the claims. This above all, is key. Even if the report were impeccable in every way, it would still require trust, and no scientific revolution should rely on the trust of a few men, no matter how respectable. The report is just not enough. In that sense, criticism of the report itself is kind of superfluous, but it does make the idea of incompetence or Rossi-imposed constraints more likely than complicity (which would be accompanied by a more flawless effort). *1. Independence* This is clearly not an independent test in the sense that the testers did not appear to choose the protocol or the methods. It was already running when they arrived. That means Rossi picked the protocol, and he was clearly involved because when it came time to inspect the inside, he took it off premises to remove the powder and then brought it back. This is important, because I don't think serious scientists would do the experiment in the way it was done, as Storms has lamented. *2. Why not self-sustaining* As with many cold fusion claims, the biggest weakness is not apparent from blatant errors detectable in a written report (especially when it can't *in principle* be checked) but from what could be so easily shown if the claims were valid, but isn't. If it were real, I'm nearly certain this is not the way it would be revealed. I'm satisfied that if the claims were true, it could be self-sustaining, and even if not for some obscure reason, with the existing claims, a truly whiz-bang demo could be easily staged, and Rossi would be on tour showing it off from the mountain tops. It would *not* be demonstrated in seclusion using an IR camera. A truly isolated, self-sustaining device that heats enough water, or lifts a heavy object (like a truck) to prove that it generates at least a few times its own weight in chemical energy could not be disputed. The amount of energy claimed could heat a 1000L hot tub to boiling (twice in the first run). It should be especially easy to make a thermal-to-thermal device self-sustaining. If (e.g.) 360 W uniformly distributed from the outside of the reactor cylinder is enough to initiate the reaction, then 1.5 kW generated within the cylinder would surely keep it going. And if it couldn't, as was suggested by someone, a little insulation and controlled cooling could maintain whatever temperature inside is necessary to keep the reaction going. Like combustion, it would self-sustain, and I think most serious scientists would find it difficult to accept such a claim if they have to keep the thing plugged in. The ludicrous suggestion that the on-off cycles represent self-sustaining is one illustration of the incompetence of the authors. Enough has been said about thermal mass to make their comments on the subject look foolish. *3. The input side* But no, as in all limitless energy claims, there's *always* input, and, par for the course, the input power is comically vague. Between the need for 3 phase input, clamp meters (which don't detect dc bias, or zero net current in paired wires), industrial trade secret waveform, a power shaping box not available for inspection, and the completely superfluous on/off cycling, it all looks like excuses to slip one past them, and that's the most likely source of the biggest deception. What kind of power supply that consumes less than 1 kW needs a 3-phase input, especially if it generates single-phase output? (Claims that other devices were plugged in to the same socket are not consistent with the 3-phase requirement of the power supply.) Independent scientists would have insisted on more careful scrutiny of the input, and would have reported a more detailed examination of the question. Some have claimed the wires were not capable of supplying the necessary power, but that's clearly nonsense in the only run that involved the Swedes, because in that run, the cylinder reached the same temperature in the calibration phase. In the first run, the power was about twice as high, but even then, no credible proof of lenr should come down to guessing at the capacity of the wires to within a factor of 2. Some have claimed the calibration run excludes tricks on the input side. But the calibration was run with constant power, and the live run used the on/off cycling. They gave no justification for the use of the on/off cycling, or whose idea it was. They didn't report measuring the box output during the cycling, and in fact, I think that was verboten. It could have been cycling, but sitting on a dc bias, so the average power was much higher. It's entirely possible that when the cycling was initiated, the measurement of the input power was somehow tricked. Ways to do this have been suggested, and no controls to exclude them were discussed in the paper. There is clearly room for some deception here. Why wouldn't they use identical power conditions for the calibration and live runs? That seems like a no-brainer. If they do have to provide input power to run the ecat, then it should be in an obviously finite way. For example, the batteries in a Chevy volt could supply the power for half of the first run on one charge. *4. Where does the heat come from?* In the spirit of point # 2, another fairly obvious indication of the source of the heat could have been provided with thermocouples placed on the reactor cylinder and outside the resistor radius. If the heat was coming from the inner reactor, it would be substantially hotter than the resistors themselves. Thermocouples are available for temperatures up to the melting point of most metals. In fact, Levi claims the horizontal black lines in the Nov run are shadows formed by the resistors in front of the hotter reactor core. It doesn't seem to occur to him that the dark lines could be the spaces between the resistors, which suggests a rather serious case of confirmation bias on his part. *5. Stefan-Boltzmann* No competent independent scientist would agree to validate a claimed revolution in physics by measuring total heat using IR thermometry and the Stefan-Boltzmann equation. They've spent months on the experiment. How hard could it be to enclose the thing and do proper flow calorimetry on it? It's gonna have to be done to exploit the heat, unless you only use it as a space heater. The question of emissivity is not trivial, especially for metals. In the first experiment, they claim to take unity as a conservative value, and justify this (prove it, they say) by assigning 0.8 and 0.95 and finding the software gives a higher temperature reading. The camera actually measures power directly, and converts to temperature using the supplied emissivity, so when you convert back to power using emissivity, it should cancel. Unfortunately, the power is measured over a very limited part of the spectrum, and so it has to infer the total power by fitting to Planck's curve, and that depends on emissivity, and therefore, the actual dependence of the final power on the emissivity will depend on the emissivity itself. And even then, the instrument makes a grey body assumption (wavelength independent emissivity), which does not apply to metals. So it is far from clear what the measurements mean if the emissivity were 0.2 or less, as is common for metals, and even some paints. Rossi could well have chosen a paint that erred in his favor for the first test, where the COP is about 5. In the second test, they used different paint and made some attempt to determine the emissivity and to check the temperature with a thermocouple, but for some reason, only at temperatures (245C) below the operating temperature (300C). This is better, but it's still error-prone. And this run got a COP of less than 3, and used trickier input with the on/off cycling. As Storms has said: "These are not difficult or complicated things to do (using calorimetry e.g.). Why are half measures repeatedly used? Why must we have to debate details that are easy to eliminate as issues?" I think the obvious reason is that when full measures are used, the effect disappears, and no one hears about it. *6. Mass delusion* Are these independent scientists really too daft to insist on mass measurements that mean something? In the first run, they couldn't measure the ecat beforehand because it was already running (i.e. Rossi started it!), but they measured another "perfectly similar device" -- really it was; Rossi said so. Then they said they learned from the experience, but in the second run they don't report weighing it beforehand either. In the second run they weigh it after, and then let Rossi make off with it to remove the powder, and use the difference as the mass of the fuel. In neither case do we have any idea of the mass of the fuel, other than the fact that it had to fit inside the cylinder. That probably still doesn't explain the claimed energy with a chemical fuel, but it might take a little heat off any input misrepresentation Rossi slipped by them. *7. Power density* The power density claims seem implausible and suggest the authors didn't really think about them. Nuclear sources do not necessarily produce high power densities, and in fact, the power density in a fission plant is typically lower than in a fossil fuel plant. That's because the temperature in the core has to be kept below the melting point of the fuel. When you burn coal, there's no such constraint. A nuclear plant is designed to remove the heat from the core as rapidly as possible, and if I read it right, a 1 GW plant uses about 200t of Uranium, for a power density of 5 kW/kg. That's 100 time lower than they're claiming for Ni-H, which has a much lower melting point. The rate of heat removal depends on the temperature difference, so it seems implausible that such a power density is possible without vastly exceeding the melting point of the nickel. Furthermore, to suggest it exceeds conceivable chemical power densities is nonsense. The power density of TNT is 10^8 kW/kg, 200 times higher than they claim. This is an explosive release of course, but still. The power density is neither here nor there, when it comes to claiming nuclear reactions, and the Ragone plot is silly. The important claims are all about energy density. But the power density reveals that they have been careless, unless I've made some grievous error above. *8. Energy density* The total energy claimed, if true, certainly suggests a new source beyond known chemical sources. In such cases, long runs, as skeptics always ask for, can exclude hidden power sources (or at least limit them), but they don't exclude errors in the power measurement, either input or output. Once you have an apparent excess power, any excess energy is just a matter of time, and so beyond excluding hidden sources as the explanation, the actual value is entirely arbitrary. In other words, the long run places at least some suspicion on the power measurements. --- To sum up, like Motl, I don't think the paper is well-written at all. It's kind of a stream of consciousness style writing, and a little amateurish. But the worst part is that the experiments are very poorly conceived. That is probably because of constraints imposed by Rossi, but whatever the case, no objective validators would choose IR thermometry to measure excess heat, and they would be far more careful and explicit about the measurement of input power and the device and internal masses. But again, the quality of the paper is secondary to its context. It gives some impression of the authors, but the overriding criticism is that the claims cannot be checked. So, to my mind, it's too early to celebrate. If Rossi ever makes ecats publicly available for truly independent testing by anyone who wants to, and it stands up, the world will beat a path to his door. It will be 1989 all over again. But, I doubt that a secret experiment performed by a cadre of true believers on an undisclosed device is going to make much of an impact.