RUN 1
3rd Series of Incandescent W Experiments (300 volt power supply)
10DEC99
After obtaining negative excess heat results in our first
two series of these experiments, we sent some W cathodes of our own manufacture
to Dr. Tadahiko Mizuno at Hokkaido University in Japan. Operated in his
cell our cathodes DO appear to generate excess heat! This fact eliminates
a number of questions about our cathode preparation methods, including the TIG
welding procedure we developed to attach the W sheet to the W lead wire.
In recent discussions with Dr. Mizuno, he indicated that results are significantly
better if the cell voltage is around 200 volts or more. The DC power supply
we used in our earlier experiments was only capable of 150 volts. In view
of this and the fact that both Mizuno and Ohmori continue to observe apparent
excess heat in their experiments we decided to purchase a new power supply and
continue this experimentation. We are also in possession of a great deal
of visual information (video and photos) obtained by Jed Rothwell during a recent
visit to Mizuno's lab in Japan. This information documents essentially
all the procedures that Mizuno employs in his experimentation. A review
of this material has yielded a number of ideas for changes in our technique
that will be employed in this next series of experiments.
Mizuno has recently sent us 5 more cathodes made with his
fabrication techniques. They arrived in good condition and we are anxious
to test them. However, for the first few runs we will use our own cathodes
to get the new system running properly.
On the left is the new cell layout. We have changed the anode from
the large Pt sheets used in previous experiments to a significantly smaller
ring of Pt mesh that sits down at the bottom of the cell, just under the W cathode.
This geometry more closely resembles that reported by Mizuno and Ohmori in "Nuclear
Transmutation Reaction Caused by Light Water Electrolysis on Tungsten Cathode
Under Incandescent Conditions", Infinite Energy #27, 1999. The cell
cap is made of G-10 and all penetrations are O-ring sealed. The cell is
equipped with a vent port that allows the steam and electrolysis gases to be
safely conducted away for collection and analysis. The cell vessel is
a standard 200 ml tall form Pyrex beaker without pour spout. Mizuno also
uses Pyrex vessels when he is only interested in excess heat results.
He switches to quartz vessels to eliminate contamination when searching for
transmutation products.
This closeup shows the hodge-podge construction of the Pt anode (adapted
from Mizuno's own construction techniques). It was assembled from two
scraps of Pt mesh and laced together with 0.5 mm diameter Pt wire which eventually
exits and forms the anode power lead. Note the appearance of the W cathode.
We have made hundreds of criss-crossed scratches in the surface (on both sides)
approximating a treatment given the cathode primarily by Ohmori and sometimes
by Mizuno. We tried using a quartz shard as recommended by M&O but
found it marginally hard enough to scratch the W reliably. We resorted
to a diamond stylus, which did the job very nicely. Both anode and cathode
leads are encased in special "super-shrink" TFE-FEP tubing obtained
from Small Parts, Inc. (part number SMDT-060) This material forms a thick,
tough coating that seals to the wire. Also visible in this photo is the
tip of the glass-jacketed thermistor temperature probe that senses electrolyte
temperature during the run.
Here is the entire experimental setup. On the far right (bottom)
is the new 300 volt 9 amp power supply that weighs 68 kg (150 lb) and delivers
DC power with 0.1% voltage regulation and 300 mV maximum ripple. Sitting
atop it is the Clarke-Hess 2330 Power Analyzer that monitors the electrical
power delivered to the cell. Atop that is a fan-cooled coil of Cu tubing
that serves to dump the majority of the heat generated by this experiment before
the water is returned to the temperature regulated bath (open ice-chest behind
the electronics). Also on top of the Clarke-Hess are two Omega CN76000
temperature controllers. One of these is devoted to the fan-cooled Cu
tubing and the other is devoted to the bath. Just to the left of the bath,
behind the laptop computer that controls everything, is the main calorimetry
unit, which serves to circulate the water (via FMI positive-displacement pump)
and regulate its temperature to within +/- 0.01C. On the far left is the
insulated experiment enclosure. Inside this enclosure is the cell, which
is surround by a heat exchanger made from coiled Cu tubing. A "double-boiler"
arrangement couples the cell thermally to the Cu tubing using water that does
not mix with either the electrolyte or the flowing calorimetry water.
For completeness, this photo shows as much of the surrounding lab
as possible. This corner of the room has a sink (right in the corner),
a fume hood, and another bench (foreground) where cell assembly is performed.
The door leads up to the office area of our facility. Behind the camera
is our machine shop, welding area, and large experiment bay (i.e. junk-filled
area). Fortunately, the activity level in this shop is quite limited so it is
possible to keep this end of the room reasonably clean.
Calibration Verification
Calibration of this type of calorimeter consists of adjusting for
zero offset and heat collection efficiency. The fundamental relationship
between temperature rise in the cooling water and heat absorbed by the water
(i.e. 4.186 joules/gm-K) provides an absolute foundation for the measurement.
Heat collection efficiency typically runs ~97% and, in this case, was observed
to be 97.3%. The zero offset takes the form A + B*(Tin - Troom) where
the second term compensates for heat lost to the surroundings due to the fact
that the calorimeter water temperature (Tin) is higher than room temperature
(Troom). In this case, A = 0.089 and B = 0.00484, which makes the total
offset correction about +0.185 degrees (for Tin = 40C and Troom = 20C).
After applying these corrections, the calibration data appear as shown in the
graph above. These data were taken with a calibration cell that consists
of a Pyrex beaker (same type as used for the electrolysis experiments) filled
with oil in which a nichrome heating element is submerged. The plot shows
various parameters plotted against time over a 24 hour period. The electrical
input power Pin is
shown in green and the measured heat output power Pout is shown in purple. The run starts with about 16 hours
of steady ~100 watt input power. During this time the average input power
was 102.3 watts and the average output power was 102.1 watts. Then the
input power was increased to about 160 watts for a few hours. During the
last hour of that period, the average input power was 161.4 watts and the average
output power was 161.7 watts. Finally the input power was removed from
the cell, and during the last hour of zero input power, the observed heat output
power was +0.05 watts, a satisfactory zero reading for this power range.
Results of Run 1
This graph shows the results of Run 1. The horizontal scale
is 4 hours (1 hr/div). The run starts with about 1.3 hours of zero input
power to allow the calorimeter to equilibrate. Just before the input
power was applied, the average Pout reading was -0.10 watts. The first application of electrical
power was about 60 watts (~20 volts) for about 10 minutes. Then the voltage
was raised to 30 volts (~120 watts) for another 10 minutes. At that time
(about 1.5 hours into the run), the cell temperature (Tcell
is yellow) was about 80C and the cell was ready for the
transition to contact glow discharge electrolysis. As is evident from
the Vcell trace
(blue), the cell voltage was raised suddenly to about 165 volts (the voltage
scale runs 0-300 as noted after the Vcell numerical display above the plot). As you can see, in the ~20
minutes that followed, Pout rose and levelled off a few watts under Pin while the voltage was
at 165. During the flattest part of this period, the average Pin was 133.6
watts while the average Pout was 127.0 watts, 95% of the Pin. At 2.0 hours, we raised the cell voltage to about 220 volts.
Unfortunately, this caused an adverse reaction in the Clarke-Hess 2330, apparently
due to increased EM noise. The vertical Pin lines show the erratic behavior
of the Clarke-Hess during this period. We turned down the voltage to about
200 volts but the problem did not go away. We scrambled to find a suitable
filter capacitor (100 mfd - 250 VDC) and installed it across the Clarke-Hess
voltage input terminals by about 2.15 hours. The noise problem ceased
abruptly. During the next 20 minutes, we observed an average Pin of 153.8 while the Pout averaged only 136.3 watts,
89% of the input power. Where was the missing 17.5 watts?
During this high-power period of the run, we noticed a considerable
flow of steam issuing from the vent hose that leads into the calorimeter chamber
and connects to the sealed electrolysis cell. We set up an ice-cooled
condensor and, in 9.6 minutes, collected 3.93 grams of grayish liquid.
We measured the salt content of this liquid (by evaporating to dryness) and,
comparing the results with the original electrolyte concentration, we determined
that this liquid was apparently composed of 0.65 grams of electrolyte expelled
from the cell mechanically (e.g. as mist) with the balance being pure H2O evaporated
from the electrolyte in the cell plus a very small amount of finely divided
material which imparted the grayish look. The power required to evaporate
that amount of H2O in 9.6 minutes is 12.6 watts, which brings our observed shortfall
down to about 5 watts. Since the cell was drawing about 0.8 amps
during this period we can immediately dismiss 1.2 watts (1.48*0.8) as fuel gas
escaping from the cell. Furthermore, judging from previous runs in which
total gas measurements were performed, it is likely that about 3 times that
much fuel gas was being generated by the sum of Faradaic decomposition and the
effects of the plasma discharge (excess dissociation up to 8 times the Faradaic
efficiency has been reported in the literature for contact-glow-discharge-electrolysis
cells). Thus we can reasonably expect 3.6 watts worth of power escaping
as fuel gas and that brings us quite close to a power balance: corrected
output power of 152.5 watts for an input power of 153.8 watts.
We tentatively conclude that there was no excess heat generated
in this run.
The finely divided residue in the collected liquid was identified
via x-ray fluorescence as W powder. This is evidence of fairly violent
conditions in the cell since the vent tube is simply a small port in the cell
cap. The electrolyte must be sloshing all over the place.
This shot shows what you see when you look into the small viewport
on the calorimeter enclosure. The thick black bars across the view are
the 1/4" diameter Cu tubing of the heat exchanger that surrounds the cell.
However, you can clearly see the lower left corner of the W cathode during incandescence.
The concentration of discharge around the edges of the sheet is typical.
For future runs, we intend to improve the EM noise filtering
so that our setup can tolerate even higher voltages approaching the 300 volt
capacity of our new supply.
Comments and suggestions are welcome.
little@earthtech.org