I have been nonplussed by Levi the last few months, because I was expecting he would provide more details about the February 2011 18-hour test with flowing water. I still wish he would do that. As I said, I wish he would tell me the made and model of the flow meter. I asked; he did not respond. In this video he did confirm there was one and it registered just under 1 L/s, which is what I reported.
Despite my previous complaints, I was impressed by Levi and I understand better what his goal in doing the test was, and why he does not want to publish it. There were several revealing statements, confirming what I have long suspected. For example, Levi described Rossi as being frightened by the ~130 kW heat burst. Only a fool would not be frightened by that, and Rossi is no fool. This confirms that the machine was not fully under control, despite what Rossi said. Krivit's long report gave the impression that Levi has disavowed these results and he will never publish more details for that reason. The video shows that is not at all what Levi meant, but I can understand why Krivit got that impression. I think he does not appreciate how European academic scientists express themselves. Anyway, Krivit deserves our thanks for uploading the video, giving us a chance to see original sources to form our own opinions. Krivit spent much of the interview trying to make it seem Levi had done something unethical by talking to reporters about his observations. I think the implication is that if the test is not good enough to publish, he should not talk about it to reporters. I think that is ridiculous. There is no such standard, and there should not be one. This is a preliminary 1-day evaluation. It is serious work, done carefully and properly. As Levi said, it is essential to do this kind of test before you commit to a full-scale project. A full-scale project takes months and hundreds of thousands of dollars. A full-scale test would include calibration with an electric heater, and test runs with blank cells, such as nickel powder and air instead of hydrogen. It would call for measuring particles, radiation, helium and so on, to show correlations (assuming there are any). It would call for many repeated test runs, some lasting for weeks. You have to do that much work for a typical journal paper. Such papers do not describe one-off experiments or events. They describe months or years of effort. It is reasonable that journals demand this. It is also reasonable that NyTeknik and other serious mass media accept brief reports of preliminary work and one-off tests. Both have value. Imagine a computer software expert spends a day working with a new program, and forms a careful evaluation of its strengths and weaknesses. It would be perfectly reasonable for that expert to tell a trade-journal reporter her opinion of the product. It would not be reasonable to demand that the expert present a fully-worked out plan for the next release of the product. There is a difference between the expert forming an opinion (something you can do in a day) and doing a complete redesign (which takes months). In off-line discussions someone suggested to me that Levi should done a better version of the 18-hour test, with improved instrumentation, rather than a full-scale experiment. I wish he would. But this test was fine. It was 100% convincing. Sure, RTDs are better than thermocouples but when you are measuring 5°C temperature difference the improved quality makes no difference. For the purposes of this test, a conventional $50 analog flow meter is every bit as good as a $10,000 ultra-precise meter. The criticism of the instruments and techniques here have been grossly exaggerated. As far as I recall, there has not been a single valid problem cited here or anywhere else. On the contrary the criticism has been notably unhinged, even by the standards of cold fusion. For example: I believe it was Beene suggested the measurements might be off by a factor of 1,000 for unstated reasons. Lomax claims that only time-sequenced data is valid even when the observers observed there were no significant changes in the temperature and flow rates for long periods. If all you see are minor fluctuations on the thermocouple digital output, that is exactly what a computer would see too. Most thermocouples have a slow response time. Trillions of individual reactions may be occurring inside the cell every second. When you measure heat from these reactions with a flow of ~1 L/s of water, individual reactions and temperature spikes they are blurred out, leaving only a stable temperature. At these flow rates I can't imagine how else it could work. In any case, if you want a graph of time-sequenced data go ahead and make one yourself. Draw a line at 5°C, add in random fluctuations of ~0.2°C, and label it "Temperature difference." You will see what a computer log of data from these thermocouples would produce. The best computer in the world will not capture more than this. If there was anything more, a person watching the thermocouple display would see it. Levi sounded offended when Krivit pressed him about credibility, and the propriety of talking to reporters about a result that you would not submit to Phys. Rev. It is ridiculous to question his credibility, or his competence to do this test. Far too much has been made here about the supposed difficulty of this test. Of course the test can be done better. Any important experiment should be improved, and improved, and improved again. Research is a never-ending open-ended process. There is no such thing as a perfect test. This test was intended to be a first-approximation, or a reality-check. It was intended to answer the questions: "Is there a real effect here? It is something we should look at more closely?" The answers are irrefutably yes, and yes. This test was as definitive and irrefutable as Orville Wright's 2-minute flight test in France, in August 1908, that instantly convinced every European expert that he had mastered controlled flight, as he claimed. There is not a single valid reason to reject this test, even though it was crude and it did not answer every question we have, or measure every parameter we want to measure, such as helium or deuterium production. The difference between Wright's test and this one is that a few days later, Wright flew again. He and his brother kept flying for weeks, until Wilbur crashed. Whereas Levi and Rossi stopped testing altogether. This is intensely frustrating. It seems stupid to me. It is counterproductive. However it has no bearing on the validity of this test, or the choice of instruments or techniques. Along similar lines, the tests with steam do not produce such easily understood, irrefutable results. But the weakness of these tests does not call into question the flowing water test. Rossi finds the steam tests faster and more convenient for his own purposes. I wish he would take the trouble to do a test to satisfy his audience rather than himself. I wish he would invite observers better qualified than Krivit, and I wish he would let observers bring their own instruments. I think he is stupid not to do this. Perhaps he has a reason but I do not know what it might be. However, whether he is stupid, whether he has hidden reasons, and all other issues relating to his personality and his behavior are irrelevant. They have no bearing on calorimetry. The results must be evaluated on their own merit without regard to personality. Each individual test must be evaluated on its own merits. - Jed
