The primary purpose of Run 9 was to demonstrate the response of our calorimetry to a simulated excess heat signal. The run employed a used cathode (of our own fabrication) from a previous run and the used 0.2M K2CO3 electrolyte from Run 8.

For the simulated excess heat signal, an immersible Joule heater was constructed by covering five series-connected 10 ohm precision resistors with Teflon heat-shrink tubing. This heater was placed immediately under the quartz cell vessel in the water that couples the cell to the calorimeter's heat exchanger.

The heater was driven with regulated DC power from a bench power supply. Both voltage and current to the heater were monitored with Fluke 87 DVM's.

The major divisions on this ruler are centimeters.

Results:

The color legend and vertical scales for this plot are as follows:

Pin (0-200 watts)

Pout (0-200 watts)

Tcell (0-100° C)

Vcell (0-200V)

Icell (0-5A)

Tin (39-41° C)

The horizontal scale covers 4 hours. Also on this plot is room temperature (in white) on a vertical scale that runs from 25C to 35C.

After a 1 hour initial equilibration period, we applied warm-up electrolysis power until the cell temperature reached 80C and then quickly raised the voltage to 135 volts which promptly induced a bright orange cathode incandescence. By the way, it appears that cathode incandescence is more easily established in 0.2M K2CO3 than in the 0.1 M K2CO3 used in Runs 1-7.

At 2.0 hours, the Joule heater was energized. At 31.1 volts, it drew 0.622 amps, which indicates precisely 50 ohms of resistance. The power delivered to the heater was therefore 19.34 watts.

We left the heater on for 30 minutes, then turned it off, and allowed the system to recover and then ended the electrolysis power at 3.07 hours.

The Pin trace declines slowly and steadily during this run. This is due to a steadily declining cell current. Similar behavior can be seen in all of our incandescent W runs and is probably due to the steady erosion of the W cathode. The observed reduction in cathode size is consistent with the power and current decreases.

For this run we have implemented a more sophisticated correction for this non-steady input power condition. To first order, the effects of the system time constant on the reported Pout can be corrected with the following expression¹:

where C is a constant related to the system time constant.

This plot shows the corrected Pout trace. As you can see the correction is not perfect but most of the exponential growth and decay characteristics have been eliminated. More importantly, the corrected Pout trace now does not include the "cooling power" mentioned in the report from Run 8.

Averaging the Pout and Pin traces during 20-minute periods before and after the Joule heater's operation, we obtain:

Avg Pin: 89.36 Avg Pout: 88.17

Averaging Pout and Pin traces during the Joule heater's operation we obtain:

Avg Pin: 88.90 Avg Pout: 107.60

Operating the Joule heater at 19.34 watts (see above) made the Pout-Pin difference increase by 19.89 watts. In other words, the calorimeter correctly measured the contribution from the Joule heater to within 3% relative.

This simulated excess heat signal was only about 20% of the input power. Please note the clear and unmistakable impact it has upon the Pout trace. The excess heat signals observed by Mizuno, et al, are at least this large and reportedly range up to more than 100% of the input power. Such a signal would drive our Pout trace off the top of the graph!