Dr. Randall Mills reports multiple observations of excess heat in the Ni-H₂O-K₂CO₃ system on the Blacklight Power web page at http://www.blacklightpower.com. In the section entitled Detail on Technology and Representative Technical Support several examples are mentioned:

- 24.6 watts out for 4.73 watts in during pulsed current electrolysis. Source unclear, perhaps Mills & Good

- pulsed and continuous current electrolysis with Pout/Pin > 37. Mills and Kneizys

- pulsed and continuous current electrolysis with 41 watts out and Pout/Pin > 8. Thermacore, Inc.

- pulsed and continuous current electrolysis with Pout/Pin > 16. HydroCatalysis Power Corporation.

- reproduction of Mills & Kneizys experiment with Pout/Pin ~ 1.67. Noninski

-continuous electrolysis with Pout/Pin ~ 2. Noninski w/ Weismann observing at Brookhaven.

- electrolysis with Pout/Pin ranging from 2.7 to 3.4. Notoya & Enyo of Catalysis Research Center, Hokkaido U

- electrolysis similar to Thermacore with Pout/Pin up to 1.68. NASA Lewis

- electrolysis with Pout/Pin ~ 10. MIT Lincoln Laboratories

- electrolysis with Pout/Pin ~ 2. Westinghouse Electric Corporation

- electrolysis with Pout/Pin 1.28-1.38. Atomic Energy Canada Limited, Chalk River Labs

- electrolysis with Pout = .75 watts and Pin = .3 watt (V*I). Moscow Power Engineering Institute

- pulsed electrolysis with Pout/Pin = 8.5. Idaho National Engineering Laboratory

After reading this list we decided to build a Ni-H₂O-K₂CO₃ cell of our own design and run it in the water-flow calorimeter originally intended for the Ragland triode cell work.

We utilized the cell structure from our Ragland triode experiments. It features a plastic frame that holds the cathode and two anodes (one for each side of the cathode). This frame seals into a 100 ml beaker and has provisions to hold Pd-coated alumina pellets which serve to recombine the electrolysis gasses.

We mounted a 1 cm² piece of Ni Fibrex sheet (a felt-like material made of Ni with a high surface area) in the cathode holder and crimped a Ni lead wire to it.

We used 1 in² pieces of Pt mesh for the anodes (one anode on each side of the Ni Fibrex sheet).

We did employ the recombiner pellets.

We filled the cell with a 0.3M solution of K₂CO₃ in H₂O.

Using constant current power supplies (we always drove each anode at approximately the same current as the other anode) we operated the cell at a variety of current levels over a 164 hour period. We employed a liquid displacement apparatus to measure the volume of gas emerging from the cell during operation.

This plot shows Pout and Pin plotted vs Time. The vertical scale is 0-10 watts. The horizontal scale is 164 hours and the vertical dotted lines are spaced every 10 hours. As you can see, there was never any obvious indication of excess heat.

The next plot allows us to look closer:

This plot shows Pout-Pin as a function of time. The vertical scale is -0.1 watts to +0.1 watts. Each of the 7 plateaus in the first plot are represented as mostly horizontal zones in this plot. It is important to notice that the very last plateau was conducted with Pin=0. A systematic error (zero error) caused the output power to read slightly positive during this plateau. The magnitude of this error is approximately equal to the apparent excess heat signals at several of the other power plateaus. In other words, the small excess heat indications above are probably due to a systematic zero error.

We did not observe any significant emission of gas from our cell. This means that the recombiner worked properly during the run. It also means that there was no significant absorption of H by the Ni cathode. When Pd cathodes are run in this cell we see 10's of cc's of O₂ gas coming out because it is "orphaned" by the H that goes into the Pd.

Our Mills experiment shows no detectable sign of excess heat. A 10% excess…less than any of the reported confirmations of Mills' experiment showed…would have produced a prominent positive signal in our experiment.

We would like to make this experiment work. Any suggestions?