Replication of Mills Light Water Calorimetry Experiment - Run 5 - 9MAR01
(Explorer users:  please set your text size (under View) to medium) 

An introduction to this experiment, the report on Run 1, the report on Run 3, and the report on Run 4 should be reviewed before reading this report.

Run 5 was an active run which employed the pulsed electrolysis power described by Mills on page 474 of the September 1996 edition of "The Grand Unified Theory of Classical Quantum Mechanics" as follows:

"...periodic square-wave having an offset voltage of 1.60 volts; a peak voltage of 1.90 volts; a peak constant current of 47.3 mA; a 36% duty cycle; and a frequency of 600 Hz."

Actually, this recipe is over-specified so we assumed that Mills was just telling us the typical peak voltage (1.90 volts) that occurred during the 47.3 mA constant current periods.  On this assumption we devised a circuit that delivered a 36% duty cycle square wave to the cell with the current during the on periods supplied from a constant current generator set at 47 mA and the voltage supplied during the off periods supplied from a DC power supply set to 1.6 volts.

This photo shows the overall appearance of the experiment during Run 5.  The equipment involved in generating the pulsed electrolysis power is grouped on the right.  At the bottom of the stack is our Clarke-Hess 2330 Power Analyzer which was employed to determine the true electrical power being delivered to the cell under these unusual electrical conditions.

This photo shows the circuit that generates the pulsed electrolysis power.  An LM317 was configured as the constant current generator and a transistor driven by a signal generator (set to 36% duty cycle and 600 Hz) provided the switching function.

This scope image, captured during Run 5, shows how closely we managed to replicate Mills conditions.  On the right, the scope's built in measurement functions indicate a frequency of 600.4 Hz, a duty cycle of 36.3%, and a peak current for the current trace (lower trace) of 47.2 mA.  The upper trace is the cell voltage (500 mV/div) and you can see that it varies from about 1.6 volts during the current-off periods to about 1.7 volts during the current-on periods.  The difference between our current-on voltage of 1.7 volts and Mills reported 1.9 volts appears to be the only significant electrical difference between our system and his.  Since we chose to match the current Mills specified during the current-on periods, there was nothing to be done about the lower voltage.  It simply indicates that our cell had a lower overall impedance than Mills' cell.

GAS PRODUCTION:

The most surprising part of Run 5 was the nearly complete absence of electrolysis gas production from the cell.  Apparently, the 36% duty cycle electrolysis power provides sufficient "relaxation" periods between gas production periods to permit essentially 100% recombination within the cell.  On most of the days of Run 5 we were unable to observe any sign of gas production at all.  Once or twice, the cell would produce a very low emission of gas amounting to ~2% of the expected flow rate.  At this rate it would take the cell 3 hours to generate 1cc of gas!

This observation is in sharp contrast to Mills' assumption (p. 474 and 479 in the above referenced book) that 100% of the electrolysis gases were escaping from the cell. 

This plot shows the entire calorimetric record for Run 5.  The horizontal axis is time (0-600 hours) with 50 hr/div.  The color-coded power traces Pout and Pin are plotted on a vertical scale that runs from -10 mW to +90 mW (10 mW/div).

The other traces are temperatures and are all plotted on a scale that runs from 10°C to 60°C (5°C/div).  The temperature of the electrolyte in the active cell is Tcell.  The temperature of the water in the reference cell is Tref.  Room air temperature is Troom.

Per Mills, the run necessarily starts with electrolysis power on.  His protocol calls for immersing the Ni cathode into the electrolyte with electrolysis power applied.  By the way, we used the same Ni cathode from Run 4 but it was given the complete cleaning procedure including soaking in 0.57M K₂CO₃ - 3%H₂O₂ solution prior to usage in Run 5.

For some unknown reason, the cell voltage wandered around 2.3-2.5 volts during the first 20 hours of the run and would not fall down to the 1.6 volt offset voltage during the current-off periods. During this time, the input power was up around 43 mW and somewhat variable.  At about 20 hours into the run, this condition corrected itself over a 1-2 hour period resulting the waveforms shown above and an average input power of about 30 milliwatts.

As you can see from the plot, there is a first equilibrium at about 150 hours, an equilibrium with a 30 mW standard addition at 300 hours, another equilibrium with electrolysis power only at 450 hours, and finally a zero-check at the very end of the run.  The table below summarizes the raw results from these four periods.

Using the standard addition to adjust the calorimeter sensitivity so that it is reported correctly produces the following adjusted results:

Because of the absence of gas production in this run, there are no further adjustments to be made to these numbers.  In other words, our Run 5 shows that the observed heat output power from the cell was nearly identical to the measured electrical input to the cell.

For the prototype experiment (see p. 479 of the above referenced book), Mills reported a V*I electrical input power of 32 mW, very similar to our 30 mW value.  Using the assumption that 100% of the electrolysis gases escape from the cell, he computes a net electrical input power of only 7 mW.  However, Mills reports a heat output power of 114 mW for this experiment, much higher than our 30 mW observation.  Even if Mills were to abandon his gas assumption, the experiment would still show a large apparent excess heat...i.e. 32 mW input and 114 mW output.

CONCLUSIONS:

Once again, our results do not match Mills' results.   Our gas flow measurements are in sharp contrast with Mills' assumptions about the behavior of his cell.  Combined with the fact that he did not actually measure the gas flow from his cell, this at least raises the possibility that Mills' assumptions were wrong.

The large discrepancy between our heat output results remains a mystery.  Either there was something seriously wrong with Mills' calorimetry or our experiment is simply not producing any excess heat.  The latter is certainly a real possibility and, in that case, we are essentially at the mercy of Dr. Mills to correct that problem, having replicated the experiment to the best of our knowledge and ability.

Until we receive such assistance from Dr. Mills, we reluctantly conclude...with uniformly negative results...our efforts to replicate his light-water Ni electrolysis experiments.

Comments and suggestions are welcome:  little@earthtech.org