Jed, how about this: Enrico Billi tells us that they weighed the E-Cat before and after, but not why it mysteriously gained a kilogram of weight. I can offer a plausible explanation.
On the bottom of the E-Cat housing sits a relatively large volume enclosure, the reactor module, which we are told houses a small reactor core and large amounts of lead shielding. This volume was not opened so its contents were not revealed. In fact, neither were its dimensions given and must be inferred from a photograph and a few other measurements. It is safe to say that it is at least 10 liters and could be as much as 20 liters. Enrico says that there were no smells of anything burning, but one of the best candidates for a hidden fuel would be and alcohol like methanol or ethanol. These are very pure chemicals that burn to produce mostly steam and a small amount of carbon dioxide. Their combustion is odorless. Their combustion products could easily have been emitted through the reactor output hose and never be detected. CO2 is odorless. Of course the obvious question is how would it receive oxygen. The not so obvious answer is a relatively unknown, but actually ubiquitous technology called a chemical oxygen generator. Referred to in the industry as an oxygen candle, it consists of a mixture of a strong oxidizer and a powdered metal. When ignited at about 600C, it smolders slowly, giving off heat and copious amounts of excess oxygen. This is the same process that provides the emergency oxygen in commercial aircraft. Its used in mining, emergency operations, any place a very compact and stable form of oxygen is required. Its storage density, in the case of a Lithium Perchlorate formulation, equals that of liquid oxygen! About 2 liters of propanol, and 2 liters of a Li Perchlorate formulation could provide more enthalpy than was measured in the Oct. 6 demonstration. The propanol, which boils at 98C would have started to emit vapor just before the water came to a boil during its warm up phase. A resistance heater would ignite the oxy candle and the two gasses would meet at the top of the housing, which is the underside of the heat exchange fins. That surface would be plated with nickel or platinum to catalytically help combust the two gasses, just as occurs in an inexpensive camping heater. This would burn for several hours, at which time a covert signal would tell Rossi its time to shut down the reactor, hence his need to be present. During the time the reactor is allowed to cool, small openings would allow water to seep into the reactor module case and make up the weight of the lost fuel and oxidizer, possibly the same openings which vented the combustion products. This would not be an exact process, hence the requirement of weighing with inaccurate scales, and the need to overlook a 1 kilogram weight gain. This example accounts for all of the observations that were reported, as well as the electrical and plumbing connections that were seen. It explains the mysterious weight gain, the need for such a prolonged warm up phase, and the need to stop the demonstration after just 4 hours.
