The University of Maine testing is part of an elaborate physics experiment meant to simulate conditions that full-scale floating wind turbines could face at an installation being planned about 10 miles off the Maine coast in up to 360 feet of water near tiny Monhegan Island.

For nearly 18 months in 2013 and 2014, an operating version of the apparatus - one-eighth of scale - sat in the waters off Castine, Me., sending electricity to the grid. That proved the technology fundamentally worked and guided refinements to the design. Now, Dr. Dagher's team is using the data collected at the lab to confirm the final form, a crucial next step in bringing the technology to market.

The saga has given renewed ammunition to cold fusion's critics in the US and Europe. In Japan, however, things have been proceeding more quietly - initially with a rather different end in mind than generating energy. In 2002, researchers from Japan's multinational Mitsubishi Heavy Industries (MHI) claimed to have used LENR techniques to "transmute" toxic, radioactive elements, such as those produced in conventional nuclear fission reactors, into other, less dangerous elements. That work is still going on. "MHI wants to develop the technique to clean up nuclear waste," says Jirohta Kasagi of Tohoku University's Clean Energy Research Lab.

In 2013, researchers from Toyota Central Research and Development Laboratories reported successfully replicating the original experiment. In a technical review published in December last year, Mitsubishi claims that "transmutation from cesium (Cs) to praseodymium (Pr), from barium (Ba) to samarium (Sm), from strontium (Sr) to molybdenum (Mo), etc., has been observed". The processes have, of course, been patented. The Japanese government, keen to decontaminate the site of the Fukushima nuclear reactor meltdown, is now providing some funding for academic LENR research.

A decade ago, researchers at the NRL tried to reproduce the Mitsubishi results, and sent a team to Japan to learn how to transmute elements first hand. NRL's David Kidwell, who carried out tests on the Mitsubishi lab equipment, was not given permission to talk to New Scientist directly, but NRL documents authorised for public release suggest another explanation for the results: contamination. They declare that "environmental surveys at MHI by NRL and MHI found praseodymium in key areas of laboratory" and the "presence of praseodymium may have other explanations than transmutation of Cs". Yasuhiro Iwamura, who led the Mitsubishi team, rebuts the NRL explanation and sticks by his claims for the experiments.

Post-Fukushima, Japan has also seen a wave of interest in LENR for energy generation, with Mitsubishi, Toyota and Nissan all investing money. Last year, the Japanese government's New Energy and Industrial Technology Development Organization announced a programme called "Energy and the Environment New Leading Technology" that called, among other things, for research into technologies that induce heat reactions between metals and hydrogen. Hideki Yoshino, a language schools magnate, has set up a company called Clean Planet to research "cleaner, safer, and more abundant resources such as solar, geothermal, LENR (also known as cold fusion), and wind to supply our energy needs".

Clean Planet is the driving force behind the Tohuku Clean Energy Research Lab, where Kasagi works. Kasagi's aim is to bring a device producing anomalous heat to market by the Tokyo Olympics in 2020. "Expected heat output might be up to several tens of watts or more," he says. As yet, he doesn't know how reliably this heat will be produced, but he is working on theories that might help improve reproducibility. "My deep interest is to clarify how the nuclear reaction can occur," he says.

That remains perhaps the biggest stumbling block: explaining how LENR is supposed to work when physics says it can't. "I dismiss most of the theories out of hand," Hubler says. Nagel feels similarly. "When it comes to the crunch, there's no theory that overlaps sufficiently with experimental data," he says.

Instead, the anomalous heat generation comes about because, when infused with deuterium and possibly other contaminants, a palladium surface generates a varying electromagnetic field that shifts electrons about, in turn releasing neutrons. These are absorbed by other nearby atoms, transmuting them and causing them to release gamma-ray photons that are absorbed by other electrons, which radiate the extra energy as heat.

Joseph Zawodny at NASA's Langley Research Center in Virginia thinks the theory is a "rich concept" that could prove extremely fruitful. "LENR is only one of its applications," he says. It doesn't rely on new physics, and makes some very specific predictions - not that those predictions have been properly tested yet. Zawodny made his own attempts, but they were "brief and low budget", he admits. The ongoing controversy surrounding Rossi's E-Cat has made getting funding for further experiments difficult, he says.

Besides Zawodny's inconclusive results, Widom and Larsen have graphs that purport to show a match between their theoretical predictions and experimental observations of how quickly various transmutation products are created. But this isn't terribly convincing to critics, because it is "after-the-fact" fitting to data from controversial experiments carried out years ago.

George Miley of the University of Illinois at Urbana-Champaign did those experiments. Now emeritus, he is still active. He had a patent granted in 2012 for a process called "dislocation site formation", which describes loading and unloading of isotopes of hydrogen into thin films in order to provide, among other things, "nuclear reaction processes". Miley claims to have a working LENR technology that produces hundreds of watts of energy, but doesn't want to say any more than that. "It is premature to discuss this new work in any detail," he says.

No risk, no reward

Such reluctance to share doesn't help to dispel scepticism, Zawodny laments. "The cold fusion stigma still remains, albeit in a weakened form," he says. "We're frustrated," says Hubler. "If we had just one-thousandth of the money going into hot fusion..."

That comparison is somewhat problematic. It's true that hot fusion isn't going anywhere fast either. ITER is beset by delays and cost overruns, and won't be working fully until the 2030s, while other projects are hardly out of the starting blocks. But at least we know why and how hot fusion works - it's what powers the sun, after all. Making it work on Earth is simply a monumental engineering problem.

Hubler's unspoken hope is for a funded fundamental physics programme that would methodically dissect cold fusion experiments to work out what is going on in the interesting 5 per cent. At the moment, he admits, there is too much trial-and-error experimentation coupled with wild speculation by theorists who often tend to ignore the details of experiments carried out so far - and indeed sometimes the known laws of physics.

Zawodny sees three vital tasks ahead. First, independent validation of existing methods for producing anomalous heat. Second, theoretical work to explain as much of the body of observations as is possible with one theory. Third, experimental testing of what a suitable theory predicts. "I think you must have all three," he says. The sticking point, as he sees it, is that most people want hard, reproducible proof that the effect is real before they stump up any cash for research. "There are few rewards without risk," he says. "One theory makes an interesting statement about cold fusion: it isn't fusion"

Some think it is worth a punt. Despite Cherokee's aborted investment in Rossi's technology, Woodford Investment Management in Oxford, UK, has recently funnelled £35 million into Industrial Heat. The company acknowledges that this is a "high risk area", but says it has done two years of due diligence and wants to build a suite of LENR technologies from what it sees as the most highly regarded people in the field. "We analysed numerous reports from a variety of scientists as well as data on investigations that had been undertaken by several government departments around the world," a spokesman told New Scientist.

Woodford's strategy is to take candidates showing real evidence of success, develop and optimise them, and then gain independent third-party verification of their findings. "It is an area that has been met with much scepticism and we are certainly not blind to this," the spokesman says. "However, the evidence we have seen to date, coupled with the potential market opportunity, suggests to us that it is an area that is worthy of further investigation."

Clearly, the House Committee on Armed Services feels the same way, but the latest signs are that the Secretary of Defense's report is delayed and won't now be presented next week as planned. We must wait a little longer to hear how warm his words will be.

The nuclear family

Nuclear fission involves splitting up heavy atomic nuclei into smaller ones. The energy this releases powers all nuclear power stations in operation today Nuclear fusion releases energy by joining up light atomic nuclei such as hydrogen and helium, the process that at huge temperatures powers the sun. So far fusion has only been achieved on any scale on Earth in the uncontrolled environment of the hydrogen bomb. Cold fusion is the controversial idea that high temperatures are not required for nuclear fusion: it can be achieved at or close to room temperature