Some Thoughts on the Design and Operation of Rossi’s LENR Reactors (Michael Lammert)

The following article has been submitted by Michael Lammert

Some Thoughts on the Design and Operation of Rossi’s LENR Reactors

 Michael Lammert (AKA Dr. Mike)


    It is quite an accomplishment that Andrea Rossi has already built a LENR reactor that was able to deliver 1MW of heat for a period of almost one year, whereas many scientists attempting to replicate his results are still trying to demonstrate excess heat in their reactors.  Although everyone following Rossi’s work is certainly disappointed that a report has not yet been released detailing the operation of the 1MW reactor, it appears that Rossi has designed and built a system that has accomplished two fundamental goals:

  • Extracting significant useful heat out of a LENR reactor
  • Operating a LENR reactor in a manner that achieves a high COP

This post examines some of the issues of LENR reactor design relative to useful heat extraction and possible implications on LENR reactor operation based on the achievement of a very high COP.

Heat Transfer in the Ni-H Reactor and Comparisons to Other LENR Reactors

Rossi’s earliest reactors used a fuel of Ni powder, hydrogen gas, plus some other “additives” or “catalysts”.  It is still not known exactly how these early reactors worked and exactly what fusion reactions were taking place within the reactor.  Perhaps the dominant fusion reactions were proton-proton, proton-deuterium, and proton-tritium as claimed to be the fusion reactions in Brillouin Energy’s Q-Pulse reactor, or perhaps the fusion reactions were proton-nickel reactions.  (What happened to the claim that the Ni was being converted to Cu in Rossi’s early reactors?)  Whatever fusion reaction took place in the early reactors, the heat was generated primarily within the lattice of the Ni particles. However, for this heat to be useful it has to be transferred to a reactor wall that is in contact with some medium to be heated.   One of the initial engineering problems to be solved in the design of the LENR reactor is how to extract useful heat (after first learning how to generate that heat).  Conductive heat transfer would be quite small in a Ni powder system because there is minimal contact between Ni particles, and few Ni particles would be in contact with the reactor wall.  Most of the fusion heat generated within the Ni powder will be transferred to the reactor walls through convection, although at very high temperatures radiative heat will also become a factor in heat transfer out of the Ni powder.  The question that needs to be answered is what would the temperature rise be in the center of the Ni powder (or the powder most distant from the reactor wall) relative to the temperature of the reactor outer wall when heat was being generated within the Ni powder at rates of several kilowatts per gram of Ni powder?   Although convective heat transfer would be aided by both the large surface area of the Ni particles and a hydrogen atmosphere, the hottest portion of the Ni powder seemingly could be several hundred degrees Centigrade above the temperature of the outer reactor wall.  Therefore, a basic Ni powder-hydrogen reactor is likely limited in power density and operating temperature by the ability to get heat transferred out of the Ni powder.  What about other LENR reactors?  Do they have a similar heat transfer limitation?

In the Pons-Fleischmann LENR reactor heat was generated by the fusion of deuterium within a palladium electrode that was part of a Pt-Pd heavy water electrolytic cell.  Most likely, this reactor was never commercially developed because of the cost of the materials, the time it took to load the Pd with deuterium, and the effect of trace contamination on keeping the cell operating.  However, extraction of useful heat also would have been an issue with the maximum output temperature being the boiling point of the heavy water.

In Bob Greenyer’s recent post “New Fires 100+ Year Gestation Part 1”, he discusses the patent of a LENR reactor that might be thought of as the plasma version of the Pons-Fleischmann device.  This reactor was developed by Canon Kabushiki Kaisha and is described in the European patent application #93111362.5.  In one application of this patent deuterium ions are generated in a plasma and then are attracted to a biased palladium coated substrate.  At a sufficiently high concentration of deuterium within the Pd coated substrate, deuterium ions will fuse to form He, releasing the heat to the Pd lattice.  The generated heat can be extracted from the backside of the substrate.  Since heat transfer in this reactor design would be primarily through conduction, which would be very efficient for a good thermally conducting substrate, this reactor could theoretically be ran at a temperature close to the melting point of Pd.  However, the fact that this invention has not been commercialized in the 25+ years since it was originally developed, probably indicates that it was difficult to achieve high rates of deuterium fusion within the Pd coated substrate.

One final LENR reactor that is very relevant to Rossi’s more advanced LENR reactors is the Unified Gravity Corporation’s lithium-proton reactor that is discussed on their website and in their patent # WO 2014 / 189799 A9.  In this reactor protons are accelerated to an energy of slightly more than 200eV, then collide with a plasma of lithium ions, resulting in the fusion of the proton with Li7 ions to produce two energetic He particles.  The ~16MeV of energy released in the fusion reaction can be captured by the walls of the reactor via kinetic energy transfer from the He particles.  The work on the Unified Gravity reactor resulted in a couple of important factors adding to the knowledge of LENR, the most important showing that a 200eV proton can fuse with Li7, whereas calculations using classical physics based on Coulomb repulsion would require the proton energy to be about 300KeV before it would fuse with Li7.  This system also demonstrates the potential for a large energy gain achievable in LENR reactors.  Supply a little more than 200eV to a proton and recover 16MeV of thermal energy!  Finally, this system demonstrates a new way to transmit heat to the reactor walls by having those walls directly capture the kinetic energy of energetic He particles.


Adding Li to the Fuel

     It is not known when Rossi first added Li to his reactor fuel, but it most likely occurred when he first made claims that only a “trace” of hydrogen was required to run his e-cat reactor.  (It is also possible that Li was one of the “additives” in fairly early reactors.)  With lithium coating the surface of the Ni particles the primary fusion reaction within the reactor becomes a proton fusing with Li7 to form two energetic He particles, releasing about 16MeV of energy per fusion.  This fusion reaction not only releases more energy per reaction than proton-proton fusion reactions or fusion reactions of protons with the Ni lattice atoms, a significant amount of the total energy generated can be directly transmitted to the outer vessel wall via the kinetic energy from the He particles similar to the Unified Gravity Corporation’s reactor.  A LENR reactor having Li as part of its fuel should be able to operate at a much higher output power density than a reactor without Li, since some portion of the generated heat is not generated within the Ni particles.  (Some portion of the He kinetic energy will be absorbed by the Ni particles.)  However, if the “ash” data from the 1MW plant data is real and is believed, there are still some of the fusion reactions converting the Ni lattice atoms to the Ni62 isotope.  The energy from these reactions directly heats the Ni lattice and therefore would still be a limiting factor in the maximum operating temperature of the reactor.

How can the maximum operating temperature of the reactor be improved?  The obvious answer is to start with a Ni fuel that is nearly 100% Ni62!  Using Ni62 in the fuel, rather than nickel with naturally occurring isotopes, could be the secret of achieving the high operating temperatures in the “hot-cat” reactor tested at Lugano.  (The reported temperatures were actually higher than actually achieved.)  When the “ash” data in the Lugano report came out showing that the ash contained much less Li7 (and a low Li7:Li6 ratio) and almost all of the Ni had been converted to Ni62, most followers of LENR were convinced that fusion had taken place within the Lugano reactor.  However, there are two factors, the number of available protons and the total energy produced by the reactor, that indicate that it was not possible to convert all of the Ni to Ni62 during the Lugano experimental runs.  The first argument for the Ni not being converted to Ni62 during the operation of the Lugano reactor is that there just weren’t enough hydrogen atoms loaded into the reactor to provide the protons necessary to convert Ni with naturally occurring isotopes all to the Ni62 isotope.  The analysis of the fuel data indicates that Li was added to the fuel in the form of LiAlH4.  Assuming this is the only source of hydrogen loaded into the reactor, there would be one hydrogen atom available to convert all of the Li to He, but the 3 remaining hydrogen atoms per LiAlH4 molecule would only be able to provide protons to convert a small percentage of the Ni to the Ni62 isotope.  Of course, it would appear to be a weak argument to claim that the fuel mixture was known by analysis to only contain LiAlH4 as a source of hydrogen when that argument is being used to claim that this fuel really wasn’t used in the reactor.  It is also possible that the Ni was pre-loaded with hydrogen, which would not have shown up in the “fuel” analysis.

An analysis of the total energy output by the Lugano “hot-cat” reactor is a much stronger argument to indicate that the Lugano reactor was loaded with Ni62 rather than the naturally occurring Ni isotopes.  Those reviewing the Lugano report had concerns about the measurement of both the input power and the output power.  The problem with the input power was that calculations of Joule heating in the copper wires connected to the reactor indicated quite high currents were supplied to the reactor heater coils during the active runs.  My belief is that the input power was probably measured correctly, even though the active runs had high heater coil currents.  However, it is not likely that the high output currents were caused by a decrease in the resistance of the Inconel wire, but rather an increase in the conductivity at high temperatures of the Al2O3 paste which covered the heater coils.  (This could easily be experimentally verified.)   An analysis of the output power form the Lugano reactor can be found on line in Thomas Clarke’s paper “Comment on the report ”Observation of abundant heat production from a reactor device and of isotropic changes in the fuel” by Levi et al”.  His output power analysis (which is certainly much more accurate than the original output power calculations by the Lugano report authors) shows that not only was the output power much less than claimed by the authors due to incorrect temperature measurements, the output power should have been many times greater than what was measured if all of the Ni had been converted to the Ni62 isotope during the Lugano experiment.   The simplest explanation that fits both the observed output power and the “ash” analysis from the Lugano reactor is that the reactor actually was loaded with Ni62, rather than a mixture of the naturally occurring isotopes.  It also makes sense from heat transfer considerations that to run a Ni-Li-H LENR reactor at the highest possible operating temperature that Ni62 should be used in the reactor to limit lattice heating from isotropic changes to the Ni in the lattice.


Implications of the Data from the 1MW system on Its Operating Parameters.

The COP of >50 for the 1MW system claimed by Rossi in his lawsuit against IH is the most important data available for this system.  Although the >50 COP can not be substantiated until the measurement procedures that were used are validated by peer review of a technical report on the system operation, it is surely true that Rossi believes in the >50 COP number since he has too much riding on winning the lawsuit to put forth a COP number that could be challenged in the lawsuit.

Attaining a COP of >50 is quite a remarkable accomplishment, especially considering that this is the first attempt to run a large scale system for an extended period of time.  What information about the operation of the 4 by 250KW reactor system might be inferred knowing that it operated at a COP of >50?  If each of the 250KW reactors required 70KW or less of external heat to bring them up to operating temperature, then it is possible that all 4 reactors could have been brought up simultaneously (assuming the data is correct that the facility is supplied with a maximum of just over 300KW).  It is more likely that each of the 250KW reactors was brought up to their operating temperature separately with 100KW to 200KW of external heat.  This is based on the assumption that the reactor might have a COP of 1.3-2.5 when external heat is being supplied.  As the 250KW reactor temperature is raised to the operating temperature, the water flow to the reactor would have to be raised until it reaches ¼ of the total flow for the system. At this point the reactor would supply its full 250KW of heat to the water flowing through the reactor (less any heat losses in the reactor itself).  Before the second reactor could be brought up to operating temperature the first reactor would probably have to be put into a self-sustaining mode (SSM), otherwise there would not be enough available power to provide heat for the second reactor.  (Perhaps if the external heaters only needed to supply 100KW-140KW, two reactors could be brought up at the same time.)  The third and fourth reactor would be brought on-line in a manner similar first reactor until all four reactors were operating in a self-sustaining mode with each reactor delivering its designed 250KW of heat to the inflowing water.

There are a couple of factors that indicate the reactors are operated only in the self sustaining mode after each reactor reaches its operating temperature with the proper water flow rate (with the external heaters used only if a reactor has to be shut down and re-started).  First, if a COP of >50 was really achieved, power to the external heaters could only be supplied 1-2% of the total operating time.  Why would external heat be needed for such a short fraction of the operating time?  A second argument for the reactors operating in a continuous SSM is that even if the external heater power was only turned on for 1-2% of the operating cycle, it would be possible that whatever measurement triggered the requirement for external heat, this trigger could happen simultaneously on 2, 3, or even all 4 of the reactors.  It’s most likely the simultaneous need for external heating power to two or more reactors would exceed the power capability of the facility.  (Note that each reactor could have the external power turned on for one minute of each hour of operation, at 15 minute intervals and still possibly achieve a COP of >50, but again what would be accomplished by such a short period of applied external heat?)

Did the reactors just run in SSM for the entire 350 day test with no external power applied?   Probably not!  It seems more likely that the reactors were actually run with a continuous, relatively low power (2KW or less) waveform having the frequency, shape and amplitude necessary to promote the fusion reactions within the reactors, perhaps at a resonance frequency of the Ni lattice.  It makes a lot more sense to control the rate of fusion reactions within the reactors with a temperature feedback loop controlling the amplitude of a low power, high frequency waveform, rather than turning on a high power heater for a small fraction of the operating cycle.  (Alternatively, the feedback loop on the temperature control system could be controlling the rate of applied pulses.)  Had the COP only been in the 5-10 range, it would have been more probable that the reactor temperature could have been controlled by turning on the full heating power for about 10-15% of the operating cycle.

Another big hint that Rossi’s LENR reactors are actually controlled by a secondary waveform is the statement from the “Introduction” of the Lugano report: “In addition, the resistor coils are fed with some specific electromagnetic pulses”.  This statement seems to have been ignored by most, if not all, of the scientists attempting to replicate Rossi’s results.




Based on the heat transfer considerations within a high temperature LENR reactor, such as the “hot-cat” used in the Lugano, it seems that the only way to limit fusion reactions within the Ni lattice, thereby permitting operation at a temperature closer to the melting point of Ni, is to use Ni that is nearly 100% Ni62, rather than use Ni with naturally occurring isotopes.  The low total output energy of the Lugano “hot-cat”, relative to the output energy expected had Ni with naturally occurring isotopes all been converted to the Ni62 isotope during the experimental runs indicates that the Lugano reactor was loaded with Ni that was close to 100% Ni62.

To achieve a COP of >50 in the 1MW system it is not likely that high power external heat was supplied to the reactors a any time other than the system start up.  It is more likely that the reactor temperature is controlled by adjusting the amplitude of a waveform that is optimized in frequency and shape to promote fusion reactions within the reactor.  If this conclusion is correct, scientists trying to replicate Rossi’s work would be more likely to achieve success by using one power source to supply heat to a reactor during start up and a separate waveform generator to supply the reactor with a range of high frequency signals that can be optimized to promote fusion reactions within the reactor.  After optimizing the waveform, the fusion reaction rate within the reactor can be controlled by feedback to the waveform generator, and the heater power supply can be turned off.