Rossi: Natural Gas Powered E-Cat Not a Matter of Simple Heat

Andrea Rossi has frequently mentioned that ‘driving’ the E-Cat with a natural gas source, rather than electricity, is a goal that Industrial Heat is working towards, because in many cases it would make the E-Cat more economical (natural gas being much less expensive than electricity in many parts of the world).

If it were simply a matter of replacing one source of heat (a gas flame for an electric heating element) one might think that this was a relatively simple task to perform — but a comment from Andrea Rossi today on the Journal of Nuclear Physics indicates that things are not this simple.

Steven Karels asked Rossi:

Can you discuss what some of the challenges are in going to a gas-fueled eCat compared to an electricity heated eCat?

For example:

1. The difference in time constant between the application and removal of heat between electric heating and flame?
2. The difference in heat transfer for gas-fired versus direct electric windings?
3. The difficulty in providing adequate ventilation for gas-fired system (incoming air)?
4. The difficulty in exhausting the exhaust products?
5. The energy efficiency of gas-fired (how much energy goes up the chimney)?

Rossi responded:

None of them.
The problem is deeper and has its roots in the core of the know how. It is not a problem of heat exchange or of heat conservation. Otherwise, it could have been already resolved.

To me, this response suggests that there is some other kinds of important stimulation given to the reactor from the electrical heating element, than simply heat up the fuel in the reactor. In connection with this idea is the comment that Rossi made some months ago about the E-Cat requiring Alternating Current (AC), and that it could not run on Direct Current (DC).]

What this means is unclear, but it does make me wonder if there is some kind of radio frequency or magnetic pulsation involved in producing an E-Cat reaction. This could be useful information for replicators as they plan for various configurations and experiments.

Perhaps Rossi and his team at IH are trying a type of hybrid approach, where they combine heat from natural gas simultaneously with AC electrical stimulation — that could get complicated, and it seems from Rossi’s response here that they still haven’t sorted this problem out.

  • Andreas Moraitis

    Rossi mentioned some time ago that they were carrying out R&D on a hybrid Cat. It seems indeed to be logical, but maybe there are problems that cannot easily be foreseen from an outsider’s perspective.

  • Steve H


    I submitted a post a few days ago on this very subject – however it seems to have gone astray.

    This is a copy of the comments I sent to Bob Greenyer of MFMP, regarding the importance of magnetostriction in Nickel:-


    do you think the following suggestion may be useful when talking to Alexander:-

    This is an idea based on the magnetostrictive properties of Nickel and the fact that Andrea Rossi uses 3-phase power for his reactor heater. I believe Alexander and MFMP are using single-phase power for theirs.

    It seems likely that the magnetostrictive behaviour of Nickel is an important part of the Rossi Effect and a high C.O.P.
    Otherwise – why use such a complicated set-up when single-phase is readily available!

    With regard to magnetostriction – imagine if you will, the continual pumping, stirring action this would create within the Nickel lattice if the applied magnetic field was oscillating.
    Also the presence of Lithium and Hydrogen atoms in an excited state (above 1000 ºC), within a pulsating lattice may present the perfect conditions for interaction.

    I propose that the heating coils in the Rossi reactor are the source of a pulsing and variable magnetic field – simply by the current flowing through the coils and their geometry.
    In addition – a three-phase heating element may substantially improve the effect.

    At 50 Hz single-phase it generates a maximum amplitude, reversing, magnetic pulse; 100 times every second. (At 60 Hz = 120 times every second).
    With 3-phase power it generates the same magnetic pulse three times as often (300/second – 50Hz, 360/second – 60Hz).



    • Maybe it is not heat that needs to be added to keep E-Cats going, but strong electromagnetic stimulation, which costs electricity. Solar Hydrogen Trends uses magnetic stimulation plus ultrasonic stimulation. Their reactor does not produce heat, but rips water apart into hydrogen gas. Defkalion uses spark stimulation plus some type of RF stimulation. I hope we gt full disclosure from a successful LENR company soon.

      • Steve H

        I concur.
        The answer to stable control will eventually come from modulated waveform or electro-magnetic stimulation.

    • US_Citizen71

      It may also be required to excite the hydrogen atoms to a higher energy level. Some may remember the device created by John S. Kanzius that was able to excite hydrogen atoms in salt water to the point that they broke their bonds with oxygen and made it appear that salt water was burning on a wick.

      • Steve H

        I see your point.
        Something has to invoke a change in the atomic charge and structure of atoms – if the Coulomb barrier is to be overcome.

  • Steve H

    Robert Godes of Brillouin also intimated that they were using RF or similar frequencies to modulate and control their process.

  • Obvious

    Rossi said once that a Bunsen burner could initiate the effect, and that electricity per se wasn’t needed. There seems to be some other problem, besides heat by itself. That could be magnetic, etc., but might also be something like a gradient in heat, or the heat feedback that is harder to maintain. One can turn heat on and off quite easily electrically. Fire isn’t quite as easy to switch off and on, and control the temperature of while maintaining a clean combustion area. (Or worse, causing potentially explosive or un-ignitable over-rich gas-air mixtures). This might limit the device to a “constant on” system only, without SSM, cancelling much of the potential savings using gas compared to electricity.

    • Pekka Janhunen

      One possibility might be that the reaction is possible to start with heat alone, but that to stop it and/or to regulate it effectively and safely one also needs some EM stimulation. The increased number and partially redundant nature of control parameters (heat plus EM) might make the R&D work time-consuming. Rossi was anyway convinced that they will succeed.

      • Steve H

        I feel confident Rossi will succeed – with an expert team.
        The parameters seem endless however:-
        * Frequencies
        * Amplitudes
        * Waveforms
        * Limitations of the signal/power generators
        * Time constants
        and last but not least – the number of years in R&D to test every permutation.

      • Obvious

        I can imagine that controlling flame heat more or less “digitally” could be done, but it could be dangerous in the early stages. Gas fitters and other experts would need to be involved, which might (in Rossi’s opinion) add too many potential extra sources of information leaks at this time in the development process. An entire new experimental area might need to be made, with work on the combustion control systems without reactors present until refined enough to add multiple reactors to the system. Until a “positive, not negative” result is obtained with the “simpler” electrical system (The 1 MW plant) IH may not want to stretch their financial resources too thin, but only perform basic research in gas-fired E-Cats, probably with single reactors.

      • Andreas Moraitis

        I agree. Rossi has often
        mentioned that an electric input is required for safety reasons. He has also
        confirmed that the gas Cat needs as well electricity. I could imagine that
        thermal and EM (or whatever) regulation parameters are not clearly separated in
        the purely electric version (as you indicate), so that the control algorithm
        has to be completely rewritten and retested. But actually we simply do not know
        enough about the structure and function of Rossi’s reactors, so that all
        assumptions about this issue remain highly speculative.

    • Steve H

      Heat alone may exhibit over-unity, with the correct recipe.
      But high C.O.P. probably requires some additional input.

  • Steve H

    I like your style!

  • builditnow

    Magnetic pulses as an internal reactor stirrer:
    Rossi has mentioned in his blog that “strong” magnetic fields are generated in the reactor when it is running. Magnetic fields interact the same way that permanent magnets interact. Magnetic fields apply force to each other and where the magnets can align they will align south pole to North pole. Sending strong magnetic pulses into the reactor could act as a stirrer. One use of a stirrer could be to disperse hot spots, basically mixing and the active material (that is making the magnetic fields coming out of the reactor). This could help to provide a more even spread of the active material through out the reactor. I’ll post an experimental setup that is gas heated and has magnetic stimulation below.

    • builditnow

      Experimental setup for a gas heated, magnetic stimulated, air controlled reactor:

      A hot air system heating and control system with magnetic stimulation:

      It’s lower tech (no high tech high temperature windings), controllable, no pressure issues in the air control system and you can heat with a gas heater to easily achieve the desired temperature.
      It also has the ability to quickly dump lots of heat.

      Since the Lugano Report indicated kilowatts of heat, and the Russian replication indicated kilowatts of heat, measurements don’t need to be very accurate to get started with and refine the test reactors.

      I’m envisioning, a MFMP dog bone like reactor constructed of the same materials as the Lugano test and the Russian replication. It would be a tube sealed at both ends with no power source, no electric wires, just some high temperature thermocouples attached to the outside (and inside if you can).

      The reactor is placed in an air circulating system made of high temperature materials, built like a flattened circle mounted vertically. In the top part of the circle sits the reactor. In the bottom is a fan assisted gas heater to heat up the air and designed to not vent heat when the gas heater is not running. A circulating fan constructed of the same high temperature materials is placed in the cooler side of the lower tube driven by a motor outside the reactor (protected by insulation / cooling). If possible a similar circulating fan is also in the top tube to make sure that the reactor receives consistent moving air over it’s length.

      A cold air dump system is added with an air pump (could be a shop vac running in reverse) and control valves far enough from the reactor that they are not hot so they don’t need to be high tech (could be a used car automatic throttle control valve). The cold air dump system would consist of two down pipes attached to the bottom of the circle on both sides of the gas heater. The cold air in would be on the output of the heater. The air dump out would be on the in side of the gas heater.

      Explosion vents to be provided liberally. These are constructed as very thin areas that are designed to blow out very easily and can be replaced very easily.

      The whole flattened circle is insulated with a high temperature material.
      Around the top reactor section, multi turn fine copper wire coils are placed over the insulation and can be cooled by a fan with cold air. The electric stimulation of the copper coils can be provided by a standard chopping type dimmer switch or motor controller, preferably set to low (maximum chopped current) and designed to give a strong magnetic pulse.

      A programmable controller to automatically activate the cold air dump would be very nice to have.

      Data logging equipment is very desirable.

      The system operation:
      1. Load the reactor, seal it, test the seal,
      2. Place the reactor in the top part of flattened circle,
      3. Attach recording devices, thermocouples, IR camera via a suitable port with a high temperature window or small amount of cold air input to protect the camera,
      4. Start the cold dump fan with the control valve off so no cold air is provided (or it might be wise to dump small amounts of air to clear any leaked hydrogen out of the circulating air).
      5. Start the circulating fans that move the air inside the reactor circle,
      6. Place the safety gear in place, blast shield and fume extraction fans,
      7. Start the gas heater and bring the air temperature up the the first test point,
      8. Watch the reactor heat up,
      9. Look for any case where the reactor is hotter than the air,
      A REACTOR HOTTER THAN THE AIR = EXOTHERMIC, not necessarily LENR exothermic proof at this point,
      10. Very slowly increase the reactor temperature. Note if it is becoming increasingly exothermic (having an increasingly higher temperature than the hot air circulating).

      Expected results:
      A. Nothing happens, keep the experiment running for a week, try another reactor, talk to other researchers.
      B. There is a slight temperature increase. There might be a hydrogen leak, evaluate options.
      C. There is a very significant temperature increase. In this case the cold air dump system may need to be operated to keep the reactor from over heating.

      If you get option C. the reactor should easily maintain hot air in the entire flattened circle (which is well insulated) and some heat dump via the cold air dump system may be necessary to prevent reactor overheating. The external gas heat should not be required. The more exothermic the reactor becomes, the lower the temperature of the circulating air required to keep the reactor from over heating.

      If you get option C. working, enclose the reactor in a blast shield insulated box (with blast safety vents) and measure the exhaust air volume and temperature differential from the room temperature (a used auto mass air sensor and thermo couple could do). Measure the power input of the fans and remove it from the heat calculation. Now you have power output accurate enough for a multi kW reactor.

      When you have your reactors running reliably at multiple kW, obtain a small jet turbine (could be a model aircraft turbine) and adapt the turbine to be the heat dump part of the system. Get a turbine with shaft output power and a gearbox suitable to drive a generator, could be a car alternator charging a battery. Electronics can be powered by an inverter attached to the battery. In this case the system has to be much more robust with few leaks at 200psi. If all is good, you could have plenty of excess electrical power to drive the control electronics with power to spare. Then you have a self running system with no external connections. Enjoy watching the perceptual motion skeptics will go nuts.

      Keep me in the loop.

      • builditnow

        Here is a diagram of an air heated and controlled test setup

        • builditnow

          Robert, I like your thinking. Some kind of working fluid with a lot more thermal capacity than air would likely be desirable. One thing air could remain good for is a converted jet turbine. As you mentioned above, the magnetic stimulation only requires a small amount of power, so, COP has to go very high once control over the heat loss as well as heat added. The Lugano experiment only controlled heat added and left heat loss as unassisted radiation and convection loss.

          With an exothermic reaction, once it’s going, all that should be required is to control the heat loss and keep up the magnetic stimulation and that can be done with small amounts of power. The heat loss on a 10kW air controlled reactor could be achieved with less than 500 Watt for control circuits, air circulators and fans. That is a COP of 20. Additionally, most of this energy use can be identified as not adding heat to the reactor. Control electronics, control valves and cold air input (as the temperature of the incoming cold air can be measured). This could reduce the total energy added to 50 Watts for magnetic stimulation and air circulators, this is a COP of 200.
          For a converted jet engine, the air circulatory become part of the engine, so then it’s just the magnetic stimulation (say 25 Watts), the COP of the reactor goes even higher to 400. A jet turbine conservatively will produce 20% efficiency as shaft power. The resultant COP electrical = 200/5 = 40.

    • Agaricus

      I don’t think the magnetic ‘stirring’ need be particularly strong – Rossi probably noticed the effect almost from the outset when using 50/60Hz heating, and refined this by adding the ‘band heater’ in the tinfoil prototype to supply a field that could be regulated separately from the actual heater element. The stainless steel and copper of the reactor would have allowed a good part of the field through to the nickel. Parkhamov’s experiment probably also makes unintended(?) use of the field around his heater element.

      By the time of the ‘hot pipe’ it seems that AR was able to combine the heater and EM source (only the two sets of windings were visible in the ‘dissection’), most probably by overlaying RF on the heater current. I would hazard a guess that MFMP will not have much success until they begin to look at this factor.

  • Sanjeev

    I’m glad that you are starting your own experiments. Where can I follow your progress ? Any links/blog etc?
    What is your source for hydrogen?

  • Obvious

    I am a fan also of the kiln method. It does simply several things at once.
    How do you plan on protecting the wiring?

  • Ophelia Rump

    It also suggests strongly that whatever that effect is, it has a strong correlation to the heat generated by the electrical heating coils. I believe that there has been some speculation that part of the reaction takes place in the heating coils. If that is true then a substantial portion of the reaction may be eliminated when you use a gas-fired heat source.

    • Steve H

      My understanding is that there is a critical thermal zone after which magnetic properties become

      • Ophelia Rump

        If there were no correlation between heat and whatever the effect is which makes the E-Cat efficient, then an electromagnetic effect could simply be modeled without the heat and require a low electrical power source to be added with a gas heat system.

        • Steve H


        • Gerard McEk

          There is a strong indication that LENR is being influenced by EM radiation. Maybe the magnetic component (dominant at low frequencies e.g.50 Hz) is decisive. If I could do tests then I would certainly make a wide and long coil with copper wire with wich you can investigate the influence of the magnetic field at different fieldstrengths and frequencies. I am sure Rossi has done that, because it is quite obvious. So it must be more difficult than that.

        • Omega Z

          If the Rossi Effect requires heat to jump start the Rossi effect & electromagnetic field for control or “Stimulation”, An electromagnetic field could supply both using the same coils.

          MRI’s produce a lot of heat & they use Helium to cool them. Most MRI’s produce a 1.5T EMF, but can range from 0.2T to 7T.

          Recall that DGT said they produced 1.5T during some part of their start up process & Rossi confirmed the the E-cat displayed a similar effect tho Rossi didn’t provide a T-scale.

          Also, If I recall, Rossi has 3 lines & only 2 have power at any given time. When switching the line power say from lines 1,2 to Lines 1,3, Rossi could be reversing the polarity of the AC current or maybe uses 2 sets of coils..

          NOTE: I could be a little off track here, But who knows.
          Check out MRI’s

          Go to Section: How MRI works-
          protons (hydrogen atoms), oscillating magnetic field, appropriate resonant frequency, excited hydrogen atoms emit a radio frequency signal.

          • Steve H

            The reversing magnetic field occurs naturally, due to the alternating current sine-wave supply. Some have commented on RF frequencies etc. but the mains frequency of 60Hz USA (50Hz Europe) is probably all that is required.
            It basically acts as the primary coil of a transformer.
            At 60 Hz single-phase the magnetic field polarity changes from +ve to -ve 120 times per second. With 3-phase this occurs 3 times as much.
            The noise emitted from a large transformer when it is energised is mainly due to this constant magnetic reversal. They are designed and built in a particular way so that the components don’t rip themselves apart with the magnetic fields and eddy currents.

          • Obvious

            The three phase is far less “bumpy”. The reversals are almost eliminated, due to the overlapping sines. Unless the sines are heavy chopped.

      • Do you think it could be possible (in the future when the details of the reaction is better understood) to use the heat just for ignition, and then keep it up with more or less “simple” magnetic stimulation?

        The COP would be impressive!

        • Steve H

          It seems likely with enough R&D.

          • If so, LENR would be unbeatable.

    • Steve H

      Heat at first – then a finely controlled magnetic field.

  • Steve H

    Gentlemen – WE ARE ON FIRE


  • Steve H

    Stick with it man. We’re all rooting for you. Bob greener of MFMP may be able to give you some tips.

  • Obvious

    I had intended to do kiln tests without any wiring at all, other than temp sensors. I suppose you will do some like that anyway as baseline tests. It should be interesting to see if the extra stimulation, …, actually stimulates the reaction.

  • Zizzle

    To avoid moisture absorbance and reactions from LAH, a simple glove box can be made. This link might be helpful:

    Chances are good that if you’re a tinkerer, you have some of the supplies on hand.

  • Gerrit

    can it be that the heat has to be delivered as heat radiation with a certain frequency, or frequency band. The thermal energy out comes with a different frequency, or is thermalized further away from the actual reaction ?

    • Obvious

      I am thinking something along these lines. Instead of a black body, there is a resonant grey body series, stoked by kinetic energy supplied by the heater.

  • Sanjeev

    All the best !
    Perhaps putting the LAH into a tiny plastic capsule which can melt at low temperature and release the LAH may work ?

    • Robert Ellefson

      Thanks, Sanjeev, I appreciate your support and suggestions. I’ll put your capsule idea on the list of things to try; I’m sure I’ll be making more than one approach before finding something that works.

  • Axil Axil

    Surface coils are designed to provide a very high RF sensitivity over a small region of interest. These coils are often single or multi-turn loops which are placed directly over the anatomy of interest. The size of these coils can be optimized for the specific region of interest.

    Surface coils make poor transmit coils because they have poor RF homogeneity, even over their region of interest. Their small field of view makes them ideal for receive, as they only detect noise from the region of interest.,d.aWw&psig=AFQjCNEQGJ52PqTs4iPgNDGJsIUArB9n6w&ust=1425246069189924

    The production of an EF signal may be why there were three coils in the heater setup.

    • Steve H

      Because freedom of speech and an education allow us to have an opinion.

  • builditnow

    Yes, magnetic stimulation could be used to help disburse hot spots like a stirrer and help even out the temperature in the reactor. This is particularly important if the reaction become more exothermic as the temperature rises, as all indications are that it does. Hot spots could develop that get hotter and hotter, thus generating more heat which could lead to a runaway heat bubble and destroy the reactor. I favor that pulsed alternating magnetic fields do not create increased reactions, but rather smooth out the hot spots.

    I’m purely guessing on this one.

  • Robert Ellefson

    Keith, when you compare the fuel-vs-ash photographs of the nickel grains, what emerges in the ash has transformed from a typical carbonyl-process nickel surface morphology (ie spiky and rough, not smooth) to a smooth, sintered-appearing surface. This suggests to me that during reactor operation, at least the surface layers of that nickel grain are in a liquid state, as evidenced by the ash photograph (page 43, particle 1 of figure 2). Have you considered this possibility?

  • Steve H

    What he said!

    Nice one Keith.



  • KeithT


    I have doubts that the analysis of the fuel and ash was fully representative, the sample size is reported as being small, also from appendix 3, page41, there is a statement “Thus, all spectrum presented henceforth is acquired from sputter cleaned areas.”

    The samples were surface cleaned to get the bulk analysis.

    Keith Thomson

  • Nicholas Cafarelli

    Thanks Robert. I am very interested in aspects of your apparatus. Please contact me through my amateur CMNS blog at

    Your idea to implement RTD in particular is worthy of note.