Cold Fusion Delusion: The Magic “Energy-Catalyser”
If you’ve been listening to The Skeptic Zone over the last year or so, you might have heard some renewed discussion of the “Energy Catalyser”; a purported “cold fusion” power reactor of some kind being promoted and sold by its Italian supposed inventor Andrea Rossi and his colleagues, after they announced their great “discovery” to the world last year.
But does this device stand up to scientific scrutiny? As you might predict, I’m going to go with no.
The purported device being promoted by Rossi et. al. is basically claimed to consist of a vessel containing nickel powder (Ni with its nuclear composition as found in nature) into which a supply of liquid water and gaseous hydrogen is admitted with heat supplied from an electrical heating element(s) powered by an outside supply of electrical energy. (There’s always an outside supply of electrical energy conveniently needed for some reason with these sorts of inventions. Every. Single. Time.) It is claimed that when the nickel is heated, a nuclear reaction occurs between nickel nuclei and protons (hydrogen-1 nuclei) at the nickel surface and a great amount of thermal energy is generated which boils the water, generating hot steam which can be used to drive a heat engine to generate electricity or otherwise do useful work (in excess of the power input fed into the system).
Let’s have a look at some of these claims.
This paper provides a description of one of the supposed experimental tests of Rossi’s supposed device.
“The flow of the inlet water was calibrated in the following way. The time for filling up 0.5 liters of water in a carafe was measured to be 278 seconds. Visual checks showed that the water flow was free from bubbles. Scaled to flow per hour resulted in a flow of 6.47 kg/hour (Density 1 kg/liter assumed).”
A “carafe”? For goodness’ sake, this is claimed to be world-changing science, so let’s at least try and be taken seriously, with some sort of scientifically respectable instrumentation. They aren’t serving up coffee to their guests. They could at least use a measuring cylinder or something.
“The water temperature was 18 °C. The specific heat of water, 4.18 joule/gram/ °C which is equal to 1.16 Wh/kg/ °C is used to calculate the energy needed to bring 1 kg of water from 18 to 100 °C. The result is 1.16 (100 – 18)=95 Wh/kg. The heat of vaporization is 630 Wh/kg. Assuming that all water will be vaporized, the energy required to bring 1 kg water of 18 °C to vapor is 95+630=725 Wh/kg. To heat up the adjusted water flow of 6.47 kg/hour from 18 °C to vapor will require 725 6.47=4.69 kWh/hour. The power required for heating and vaporization is thus 4.69 kW.”
If we assume that the water flowing into the system is completely boiled and turned into steam, then the total power required (including power input into the system, and/or energy generated within the system) will be thus:
P = ρ * dV/dt * ((Cv * (373.15 K — Tin) + L)
P is the power input into the system, in watts, ρ is the density of water (992.2 kg/m3 at 40 °C), dV/dt (the volumetric flow rate) is claimed to be equal to 1.8 * 10–6 m3/s, Cv is the specific heat capacity of water, 4.18 * 103 J kg–1K–1, Tin is claimed to be 18 C (291.15 K), and L is the latent heat of vaporization of water, which is 2.26 * 106 J/kg.
Plug that all together and you get 4.65 kW, most of which is the latent heat. So, OK, at least their arithmetic is sound enough here.
“The electric heater was switched on at 10:25, and the meter reading was 1.5 amperes corresponding to 330 watts for the heating including the power for the instrumentation, about 30 watts. The electric heater thus provides a power of 300 watts to the nickel-hydrogen mixture. This corresponds also to the nominal power of the resistor.”
330 W / 1.5 A = 220 volts, right? So, if these numbers are to be trusted, exactly what kind of power supply was used to supply 220 V to the electronics? What type of instrumentation was used to make these measurements?
What is the actual voltage drop measured across the resistor? Is the power supply AC or DC? If this is AC, shall we just assume the values are RMS values? What is the justification for saying that the “power for the instrumentation is about 30 watts”? If this was just mains grid power, its voltage cannot just be “trusted” to be 220 V RMS without actually measuring it.
(The Australian Standard for the mains power grid, just as an example, specifies a line voltage of 230 V RMS +10% –6%. That is, the RMS voltage could be anywhere from 216 VAC to 253 VAC and still be within standard.)
Why has the power dissipation in the heating resistor not been directly and independently measured? If there are multiple heating elements in any particular experimental setup, then they must all be instrumented and measured in the same way, independently.
The manufacturer’s rated power dissipation in the resistive heater (specified at what supply voltage?) cannot simply be assumed without confirming it empirically.
One can not simply measure the cold resistance of the heaters and measure either the current through them or the voltage across them and then use Ohm’s law to back-calculate their power dissipation, since resistive heating elements have a non-linear voltage-current curve and their resistance changes when they are hot making such calculations inaccurate.
The heating elements should be powered from a DC power supply, not AC, in order to avoid any incorrect or inaccurate power measurements or power calculations due to factors such as back EMF, incorrect measurement of apparent power as opposed to real power in an inductive load, incorrect calculation of true RMS power in the measuring instrument, etc, which all need to be factored in correctly and accurately when making power measurements on a reactive AC system.
If these heating elements are powered from the mains grid, then the voltage or frequency of the mains supply cannot simply be assumed without checking it.
A hydrogen-oxygen chemical reaction with any residual oxygen within the system, and/or a chemical reaction between pressurized hydrogen and some other organic material which may be present within the system, catalyses by the finely divided nickel, could also potentially account for any “excess” energy apparently generated within the system, as could the energy stored within compressed hydrogen.
Of course, like most scientific claims, these claims should not be particularly difficult to falsify. What would you need to do to falsify these claims empirically?
- Put a flow rate meter on the water inlet pipe to provide a continuous and accurate measurement of the input flow rate of liquid water.
- Put a thermocouple on the water inlet pipe to provide a continuous and accurate measurement of the temperature of the inlet water.
- Put a flow rate meter on the steam outlet, along with a thermocouple. If the output working fluid is a mixture of water and steam (or steam and hydrogen) then measuring the flow rate might be tricky… but you might just assume that the mass of working fluid is conserved and it doesn’t just disappear. For the sake of accurate measurement, the flow rate of water input and/or the system’s power output should be adjusted such that the output steam is completely vaporized, and it is not a mixture of gaseous steam and liquid water.
- Put a good needle valve or leak valve with a known, controlled flow rate on the hydrogen line — or, preferably, a mass flow controller.
- Put a pressure transducer in the reactor vessel to measure the hydrogen (and/or steam) pressure.
- Directly measure the current through and the voltage drop across each of the resistive heaters and log this data continuously and in real time, and hence directly measure the power dissipation in all of the heating elements attached to the reactor vessel. Measure the power dissipation in these heaters independently, independent of any other electronics, and power the heaters from DC to avoid the complexities of accurate AC power metrology.
- Plug all the above sensors into a data-logging acquisition interface and record their data continuously, record the raw data continuously prior to operation of the device and during operation of the device, publish this raw data, and show us.
These tests and experiments which would falsify the claims being made do not require any special nuclear physics expertise. For example, Adam Savage and Jamie Hyneman would be able to pull off a set of reproducible experiments, as described above, quite easily.
The discussions between Ian Bryce and Richard Saunders have discussed at length how you might have to investigate or modify the demonstration system demonstrated by Rossi et. al. in order to avoid “cheating” — for example, using the earth wire inside the mains cable to feed power to the load whilst not actually measuring that power — but I think it is impractical to scrutinize a demonstration device provided by Rossi in order to detect all possible “mistakes” or ways that energy could be supplied to the system without properly being accounted for whether through incompetence or malice. Furthermore, this is inconsistent with the usual scientific method by which these things are done.
What should be done is this:
- Extract drawings, documentation and details of the system from Rossi, including all the electrical connections and instrumentation, of sufficient detail to allow the experiment to be independently reproduced.
- Reproduce the experiment and see if the results can be reproduced.
Let’s just say that they’re not the first crackpots in the world who claim to have invented a magical machine that will solve all the world’s energy problems forever.
They’re not the first to try and pull the “we can’t show you any technical details about how it works at all and we can’t publish meaningful details that would allow the experiment to be reproduced… because then Big Oil (or whatever) will steal it” rubbish.
We really don’t need to, but anyway, just for the sake of completeness, why don’t we just look a little at the (claimed) nuclear physics characteristics of the system as well?
II: Nuclear physics
The supposed physics underpinning Rossi’s supposed device has been described primarily in several publications published in the Journal of Nuclear Physics.
That all sounds very exciting, until you realise that the Journal of Nuclear Physics is actually just the name they’ve given to their own self-published Web-based
What nuclear reactions are claimed to occur within this system?
Let’s take it straight from the horse’s mouth:
“Following Focardi Rossi . a Ni58 nucleus produces a Copper nucleus according to the reaction
Ni58 + p → Cu59
Copper nucleus Cu59 decays with positron (e+) and neutrino (ν) emission in Ni59 nucleus according to
Cu59 → Ni59 + ν + e+
Then (e+) annichilates [sic] with (e-) in two gamma-rays
e– + e+ → γ + γ
Starting  from Ni58 which is the more abundant isotope, we can obtain as described in the two above processes Copper formation and its successive decay in Nickel, producing Ni59, Ni60, Ni61 and Ni62. Because Cu63, which can be formed starting by Ni62, is stable and does not decay in Ni63, the chain stops at Ni62 (i.e. Cu63).”
So, what they’re saying here is that supposing we start with abundant natural 58Ni, a proton is captured to yield 59Cu which decays by positive beta decay to yield 59Ni, and then the process is repeated through a chain of successive proton captures and β+ decays to yield successively heavier Ni and Cu nuclei, until we get up to 63Cu which is stable. If you start off with one of those heavier Ni nuclides in the natural Ni sample, then you just “skip a step” or two, ostensibly. (One notes that 62Ni has the highest binding energy per nucleon of any known nuclide, so you do wonder how small its cross-section for this sort of transmutation would actually be.)
How do the protons overcome the Coulomb barrier, in order to interact with the nucleus? Why, simple! We just draw little cartoon masks on them, and then they can sneak right through the Coulomb barrier, right?
It’s funny how they claim to be using nickel as a nuclear fuel, an element whose natural, stable nuclides are essentially right near the peak of nuclear-binding stability, where it will be essentially impossible to get any significant amount of nuclear binding energy liberated in a nuclear reaction.
58Ni is the predominant natural nuclide, making up 68% of natural Ni (It is essentially stable, although theoretically unstable by double-beta decay). So, yes, it is the plausible place to look for the start of this supposed chain of “cold fusion” reactions. Proton capture on 58Ni yields 59Cu, which decays by electron capture or e+ emission with a half-life of 81.5 seconds. In the case of positron emission, the emitted positron will then annihilate into a pair of photons.
This is all acknowledged in the above-quoted paragraph… and yet at the same time they make this bizarrely inconsistent claim that the system contains no radioactivity and emits no ionizing radiations. Obviously this is nonsense intended solely to sell this snakeoil to a non-scientific audience.
Granted, the unstable copper nuclides produced do have very short half-lives — the longest-lived Cu radionuclide present is 61Cu with a half-life of only 3.33 hours. But nickel-59 has a half-life of 76,000 years.
Rossi et. al. claims that this fantastic “reactor” produces no radioactivity or ionizing radiation. This is physically completely implausible, but it’s an attractive “good” claim to try and use when they’re trying to sell this product to politicians, the media, and the public, because of the widespread fear and pseudoscience associated with radioactivity or ionizing radiation in the public. These two factors combined — that it’s totally implausible physics, but “good” marketing to the public — should immediately set off skeptical alarm bells that there’s snake oil here.
“The conclusions from the papers  to  are that nickel and hydrogen provide the fuel for nuclear processes inside a small container in a radiation shielded setup and that in the room outside, no radiation different from the ambient one is found.”
I frankly don’t give a damn about the room outside… are you detecting any kind of radiation emitted from the device itself, right next to the device, with no shielding? If so, what kind of particles, and what count rate? Neutrons, photons or what? Any bonafide nuclear reaction must be emitting X-rays or ɣ-rays, so what does the gamma spectroscopy look like? If what they claim is true, then 511 keV photons from positron annihilation must be clearly visible in the gamma-ray spectrum, just to give one specific example.
The detection of emitted ɣ-rays or particles of some kind is the classic signature of a nuclear reaction and the means by which we experimentally observe and characterize and identify nuclear reactions. What they’re claiming here is that fundamentally new physics exists, inconsistent with all established physics — so if they are to be taken seriously by physicists, then clearly probing the underlying physics is important.
[Creative Commons licensed Flickr photo by simonmonk]